id
int64
28.8k
36k
text
stringlengths
44
3.03k
translation
stringlengths
36
2.47k
29,672
We propose an alternative approach towards cost mitigation in volatility-managed portfolios based on smoothing the predictive density of an otherwise standard stochastic volatility model. Specifically, we develop a novel variational Bayes estimation method that flexibly encompasses different smoothness assumptions irrespective of the persistence of the underlying latent state. Using a large set of equity trading strategies, we show that smoothing volatility targeting helps to regularise the extreme leverage/turnover that results from commonly used realised variance estimates. This has important implications for both the risk-adjusted returns and the mean-variance efficiency of volatility-managed portfolios, once transaction costs are factored in. An extensive simulation study shows that our variational inference scheme compares favourably against existing state-of-the-art Bayesian estimation methods for stochastic volatility models.
ما یک روش جایگزین برای کاهش هزینه در پرتفوی مدیریت شده با نوسانات بر اساس صاف کردن چگالی پیش بینی کننده یک در غیر این صورت مدل نوسانات تصادفی استاندارد.به طور خاص ، ما توسعه می دهیم روش تخمین متغیر جدید Bayes که انعطاف پذیر متفاوت است فرضیات صافی صرف نظر از تداوم نهان زیربنایی حالت.با استفاده از مجموعه بزرگی از استراتژی های معاملات سهام عدالت ، ما نشان می دهیم که هموار سازی هدف قرار دادن نوسانات به منظم کردن اهرم/گردش مالی شدید کمک می کند نتایج حاصل از برآورد واریانس تحقق یافته معمولاً استفاده می شود.این مهم است پیامدهای مربوط به بازده تنظیم شده در معرض خطر و هم واریانس پس از هزینه های معامله ، کارآیی اوراق بهادار با نوسانات Factored in. یک مطالعه شبیه سازی گسترده نشان می دهد که استنتاج تغییر یافته ما این طرح به طرز مطلوبی در برابر پیشرفته ترین چیزهای بیزی مقایسه می شود روش های تخمین برای مدل های نوسانات تصادفی.
29,673
This paper investigates the finite sample performance of a range of parametric, semi-parametric, and non-parametric instrumental variable estimators when controlling for a fixed set of covariates to evaluate the local average treatment effect. Our simulation designs are based on empirical labor market data from the US and vary in several dimensions, including effect heterogeneity, instrument selectivity, instrument strength, outcome distribution, and sample size. Among the estimators and simulations considered, non-parametric estimation based on the random forest (a machine learner controlling for covariates in a data-driven way) performs competitive in terms of the average coverage rates of the (bootstrap-based) 95% confidence intervals, while also being relatively precise. Non-parametric kernel regression as well as certain versions of semi-parametric radius matching on the propensity score, pair matching on the covariates, and inverse probability weighting also have a decent coverage, but are less precise than the random forest-based method. In terms of the average root mean squared error of LATE estimation, kernel regression performs best, closely followed by the random forest method, which has the lowest average absolute bias.
در این مقاله به بررسی عملکرد نمونه محدود طیف وسیعی از متغیر ابزاری پارامتری ، نیمه پارامتری و غیر پارامتری برآوردگرها هنگام کنترل یک مجموعه ثابت از متغیرهای متغیر برای ارزیابی محلی اثر درمانی متوسط.طرح های شبیه سازی ما مبتنی بر کار تجربی است داده های بازار از ایالات متحده و در ابعاد مختلف از جمله اثر متفاوت است ناهمگونی ، انتخاب ابزار ، قدرت ابزار ، نتیجه توزیع و اندازه نمونه.در میان برآوردگرها و شبیه سازی های در نظر گرفته شده ، برآورد غیر پارامتری بر اساس جنگل تصادفی (یک یادگیرنده ماشین کنترل متغیرهای متغیر به صورت داده محور) از نظر شرایط رقابتی را انجام می دهد از میانگین نرخ پوشش اطمینان 95 ٪ (مبتنی بر بوت استرپ) فواصل ، در حالی که نسبتاً دقیق نیز هستند.هسته غیرعادی رگرسیون و همچنین نسخه های خاصی از شعاع نیمه پارامتری بر روی نمره گرایش ، تطبیق جفت بر روی متغیرهای متغیر و احتمال معکوس وزن نیز دارای پوشش مناسبی است ، اما از تصادفی دقیق تر است روش مبتنی بر جنگل.از نظر میانگین میانگین خطای مربع ریشه اواخر تخمین ، رگرسیون هسته بهترین عملکرد را دارد ، از نزدیک با تصادفی دنبال می شود روش جنگل ، که کمترین تعصب مطلق را دارد.
29,674
This paper considers the estimation of treatment assignment rules when the policy maker faces a general budget or resource constraint. Utilizing the PAC-Bayesian framework, we propose new treatment assignment rules that allow for flexible notions of treatment outcome, treatment cost, and a budget constraint. For example, the constraint setting allows for cost-savings, when the costs of non-treatment exceed those of treatment for a subpopulation, to be factored into the budget. It also accommodates simpler settings, such as quantity constraints, and doesn't require outcome responses and costs to have the same unit of measurement. Importantly, the approach accounts for settings where budget or resource limitations may preclude treating all that can benefit, where costs may vary with individual characteristics, and where there may be uncertainty regarding the cost of treatment rules of interest. Despite the nomenclature, our theoretical analysis examines frequentist properties of the proposed rules. For stochastic rules that typically approach budget-penalized empirical welfare maximizing policies in larger samples, we derive non-asymptotic generalization bounds for the target population costs and sharp oracle-type inequalities that compare the rules' welfare regret to that of optimal policies in relevant budget categories. A closely related, non-stochastic, model aggregation treatment assignment rule is shown to inherit desirable attributes.
در این مقاله تخمین قوانین واگذاری درمان در هنگام سیاست ساز با بودجه عمومی یا محدودیت منابع روبرو است.استفاده از چارچوب Pac-Bayesian ، ما قوانین جدید واگذاری درمان را پیشنهاد می کنیم که اجازه می دهد برای مفاهیم انعطاف پذیر نتیجه درمان ، هزینه درمان و بودجه محدودیتبه عنوان مثال ، تنظیمات محدودیت امکان صرفه جویی در هزینه را فراهم می کند. هزینه های عدم درمان بیش از هزینه های درمان برای یک زیرمجموع است. به بودجه.همچنین تنظیمات ساده تری مانند محدودیت های کمیت ، و نیاز به پاسخ و هزینه های نتیجه ندارد همان واحد اندازه گیری.مهمتر از همه ، رویکرد تنظیمات را به خود اختصاص می دهد جایی که محدودیت های بودجه یا منابع ممکن است مانع از درمان همه آنچه می شود سود ، در جایی که هزینه ها ممکن است با خصوصیات فردی متفاوت باشد و در آنجا وجود دارد ممکن است عدم اطمینان در مورد هزینه قوانین مربوط به درمان باشد.با وجود نامگذاری ، تجزیه و تحلیل نظری ما خصوصیات مکرر را بررسی می کند قوانین پیشنهادیبرای قوانین تصادفی که به طور معمول نزدیک می شوند به حداکثر رساندن سیاست های رفاه تجربی با حداکثر بودجه در نمونه های بزرگتر ، ما مرزهای تعمیم غیر غیر متعارف را برای هزینه های جمعیت هدف و نابرابری های تیز اوراکل که از رفاه قوانین با آن مقایسه می کنند سیاست های بهینه در دسته های بودجه مربوطه.از نزدیک مرتبط ، قانون واگذاری درمان با تجمع غیر تصادفی ، مدل نشان داده شده است ویژگی های مطلوب.
29,675
We develop a general framework for the identification of counterfactual parameters in a class of nonlinear semiparametric panel models with fixed effects and time effects. Our method applies to models for discrete outcomes (e.g., two-way fixed effects binary choice) or continuous outcomes (e.g., censored regression), with discrete or continuous regressors. Our results do not require parametric assumptions on the error terms or time-homogeneity on the outcome equation. Our main results focus on static models, with a set of results applying to models without any exogeneity conditions. We show that the survival distribution of counterfactual outcomes is identified (point or partial) in this class of models. This parameter is a building block for most partial and marginal effects of interest in applied practice that are based on the average structural function as defined by Blundell and Powell (2003, 2004). To the best of our knowledge, ours are the first results on average partial and marginal effects for binary choice and ordered choice models with two-way fixed effects and non-logistic errors.
ما یک چارچوب کلی برای شناسایی ضد عملی ایجاد می کنیم پارامترها در یک کلاس از مدلهای پانل نیمه خطی غیرخطی با ثابت اثرات و اثرات زمانی.روش ما در مورد مدل ها برای نتایج گسسته اعمال می شود (به عنوان مثال ، انتخاب باینری اثرات ثابت دو طرفه) یا نتایج مداوم (به عنوان مثال ، رگرسیون سانسور شده) ، با رگرسیون گسسته یا مداوم.نتایج ما انجام می شود نیازی به فرضیات پارامتری در مورد شرایط خطا یا همگن بودن زمان بر روی آن نیست معادله نتیجهنتایج اصلی ما روی مدلهای استاتیک متمرکز است ، با مجموعه ای از نتایج اعمال شده برای مدل ها بدون هیچ گونه شرایط اگزوژن.ما نشان می دهیم که توزیع بقا از نتایج متقابل مشخص شده است (نقطه یا جزئی) در این کلاس از مدل ها.این پارامتر برای اکثر آنها یک ساختمان ساختمانی است اثرات جزئی و حاشیه ای علاقه به عمل کاربردی که مبتنی بر آن است عملکرد ساختاری متوسط ​​همانطور که توسط بلوندل و پاول تعریف شده است (2003 ، 2004). به بهترین دانش ما ، ما اولین نتایج به طور متوسط ​​جزئی و اثرات حاشیه ای برای انتخاب باینری و مدل های انتخاب شده با دو طرفه ثابت اثرات و خطاهای غیرقانونی.
29,676
The analysis of discrimination has long interested economists and lawyers. In recent years, the literature in computer science and machine learning has become interested in the subject, offering an interesting re-reading of the topic. These questions are the consequences of numerous criticisms of algorithms used to translate texts or to identify people in images. With the arrival of massive data, and the use of increasingly opaque algorithms, it is not surprising to have discriminatory algorithms, because it has become easy to have a proxy of a sensitive variable, by enriching the data indefinitely. According to Kranzberg (1986), "technology is neither good nor bad, nor is it neutral", and therefore, "machine learning won't give you anything like gender neutrality `for free' that you didn't explicitely ask for", as claimed by Kearns et a. (2019). In this article, we will come back to the general context, for predictive models in classification. We will present the main concepts of fairness, called group fairness, based on independence between the sensitive variable and the prediction, possibly conditioned on this or that information. We will finish by going further, by presenting the concepts of individual fairness. Finally, we will see how to correct a potential discrimination, in order to guarantee that a model is more ethical
تجزیه و تحلیل تبعیض از مدتهاست که اقتصاددانان و وکلا علاقه مند هستند.که در سالهای اخیر ، ادبیات در علوم کامپیوتر و یادگیری ماشین به موضوع علاقه مند شوید و دوباره خواندن جالب از آن را ارائه دهید موضوع.این سؤالات عواقب انتقادات بیشماری از آن است الگوریتم های مورد استفاده برای ترجمه متون یا شناسایی افراد در تصاویر.با ورود داده های گسترده و استفاده از الگوریتم های فزاینده مات ، این است از الگوریتم های تبعیض آمیز تعجب آور نیست ، زیرا این کار آسان شده است با غنی سازی داده ها به طور نامحدود ، یک پروکسی از یک متغیر حساس داشته باشید. به گفته کرانزبرگ (1986) ، "فناوری نه خوب و نه بد است و نه خنثی "، و بنابراین ،" یادگیری ماشین چیزی مانند جنسیت به شما نمی دهد بی طرفی "به صورت رایگان" که شما صریحاً درخواست نکرده اید "، همانطور که ادعا شده است Kearns et a.(2019).در این مقاله ، ما به متن عمومی باز خواهیم گشت برای مدل های پیش بینی در طبقه بندی.ما مفاهیم اصلی را ارائه خواهیم داد انصاف ، به نام انصاف گروه ، بر اساس استقلال بین حساس متغیر و پیش بینی ، احتمالاً در مورد این یا آن اطلاعات مشروط است. ما با ارائه بیشتر ، با ارائه مفاهیم فردی به پایان خواهیم رسید انصافسرانجام ، ما خواهیم دید که چگونه یک تبعیض احتمالی را اصلاح کنیم ، در به منظور تضمین اینکه یک مدل اخلاقی تر است
29,677
It is well-known that generalized method of moments (GMM) estimators of dynamic panel data models can have asymptotic bias if the number of time periods (T) and the number of cross-sectional units (n) are both large (Alvarez and Arellano, 2003). This conclusion, however, follows when all available instruments are used. This paper provides results supporting a more optimistic conclusion when less than all available instrumental variables are used. If the number of instruments used per period increases with T sufficiently slowly, the bias of GMM estimators based on the forward orthogonal deviations transformation (FOD-GMM) disappears as T and n increase, regardless of the relative rate of increase in T and n. Monte Carlo evidence is provided that corroborates this claim. Moreover, a large-n, large-T distribution result is provided for FOD-GMM.
کاملاً مشهور است که برآوردگرهای روش کلی از لحظات (GMM) از در صورت تعداد زمان ، مدل های داده پانل پویا می توانند تعصب بدون علامت داشته باشند دوره (T) و تعداد واحدهای مقطعی (N) هر دو بزرگ هستند (آلوارز و آرلانو ، 2003).این نتیجه گیری ، با این حال ، در صورت وجود همه در دسترس است از ابزارها استفاده می شود.در این مقاله نتایج حاصل از خوش بینانه تر ارائه شده است نتیجه گیری هنگامی که کمتر از همه متغیرهای ابزاری موجود استفاده می شود.اگر تعداد ابزارهای مورد استفاده در هر دوره با T به اندازه کافی آهسته افزایش می یابد ، تعصب برآوردگرهای GMM بر اساس انحرافات متعامد رو به جلو تحول (FOD-GMM) بدون در نظر گرفتن افزایش T و N از بین می رود میزان نسبی افزایش در t و n.شواهد مونت کارلو ارائه شده است که این ادعا را تأیید می کند.علاوه بر این ، یک نتیجه توزیع بزرگ و بزرگ و بزرگ است برای FOD-GMM فراهم شده است.
29,849
We reassess the use of linear models to approximate response probabilities of binary outcomes, focusing on average partial effects (APE). We confirm that linear projection parameters coincide with APEs in certain scenarios. Through simulations, we identify other cases where OLS does or does not approximate APEs and find that having large fraction of fitted values in [0, 1] is neither necessary nor sufficient. We also show nonlinear least squares estimation of the ramp model is consistent and asymptotically normal and is equivalent to using OLS on an iteratively trimmed sample to reduce bias. Our findings offer practical guidance for empirical research.
ما استفاده از مدلهای خطی را برای تقریبی احتمالات پاسخ ارزیابی می کنیم نتایج باینری ، با تمرکز بر اثرات جزئی جزئی (APE).ما آن را تأیید می کنیم پارامترهای طرح ریزی خطی با میمون ها در سناریوهای خاص همزمان هستند.از طریق شبیه سازی ، ما موارد دیگری را شناسایی می کنیم که OLS تقریباً انجام نمی شود میمون ها و دریافتند که داشتن بخش بزرگی از مقادیر متناسب در [0 ، 1] هیچ یک نیست لازم و نه کافی.ما همچنین تخمین حداقل مربعات غیرخطی را نشان می دهیم مدل رمپ سازگار و بدون علامت طبیعی است و معادل آن است با استفاده از OLS بر روی یک نمونه مکرر برش خورده برای کاهش تعصب.یافته های ما ارائه می دهد راهنمایی های عملی برای تحقیقات تجربی.
29,679
Survey questions often elicit responses on ordered scales for which the definitions of the categories are subjective, possibly varying by individual. This paper clarifies what is learned when these subjective reports are used as an outcome in regression-based causal inference. When a continuous treatment variable is statistically independent of both i) potential outcomes; and ii) heterogeneity in reporting styles, a nonparametric regression of integer category numbers on that variable uncovers a positively-weighted linear combination of causal responses among individuals who are on the margin between adjacent response categories. Though the weights do not integrate to one, the ratio of local regression derivatives with respect to two such explanatory variables identifies the relative magnitudes of convex averages of their effects. When results are extended to discrete treatment variables, different weighting schemes apply to different regressors, making comparisons of magnitude less informative. I obtain a partial identification result for comparing the effects of a discrete treatment variable to those of another treatment variable when there are many categories and individual reporting functions are linear. I also provide results for identification using instrumental variables.
سوالات نظرسنجی اغلب پاسخ هایی را در مقیاس های سفارش داده شده که برای آنها ایجاد می کنند تعاریف دسته ها ذهنی است ، احتمالاً توسط فرد متفاوت است. در این مقاله آنچه آموخته می شود هنگام استفاده از این گزارش های ذهنی به عنوان نتیجه در استنتاج علیت مبتنی بر رگرسیون.هنگامی که یک درمان مداوم است متغیر از نظر آماری مستقل از هر دو) نتایج بالقوه است.و ii) ناهمگونی در سبک های گزارش ، یک رگرسیون غیرپارامتری عدد صحیح اعداد دسته بندی در آن متغیر ، یک خطی با وزن مثبت را کشف می کند ترکیبی از پاسخ های علّی در بین افرادی که در حاشیه هستند مقولات پاسخ مجاور.گرچه وزنه ها با یک ادغام نمی شوند ، اما نسبت مشتقات رگرسیون محلی با توجه به دو مورد توضیحی متغیرها بزرگی نسبی میانگین محدب آنها را مشخص می کنند اثراتهنگامی که نتایج به متغیرهای درمانی گسسته گسترش می یابد ، متفاوت است طرح های وزنه برداری در مورد رگرسیون های مختلف اعمال می شود و مقایسه ای از آنها بزرگی کمتر آموزنده است.من یک نتیجه شناسایی جزئی برای مقایسه اثرات متغیر درمانی گسسته با موارد دیگری متغیر درمانی در هنگام وجود دسته های مختلف و گزارش های فردی توابع خطی هستند.من همچنین با استفاده از شناسایی نتایج ارائه می دهم متغیرهای ابزاری.
29,680
In this paper, we propose Forest-PLS, a feature selection method for analyzing policy effect heterogeneity in a more flexible and comprehensive manner than is typically available with conventional methods. In particular, our method is able to capture policy effect heterogeneity both within and across subgroups of the population defined by observable characteristics. To achieve this, we employ partial least squares to identify target components of the population and causal forests to estimate personalized policy effects across these components. We show that the method is consistent and leads to asymptotically normally distributed policy effects. To demonstrate the efficacy of our approach, we apply it to the data from the Pennsylvania Reemployment Bonus Experiments, which were conducted in 1988-1989. The analysis reveals that financial incentives can motivate some young non-white individuals to enter the labor market. However, these incentives may also provide a temporary financial cushion for others, dissuading them from actively seeking employment. Our findings highlight the need for targeted, personalized measures for young non-white male participants.
در این مقاله ، ما Forest-PLS را پیشنهاد می کنیم ، یک روش انتخاب ویژگی برای تجزیه و تحلیل ناهمگونی اثر سیاست در یک انعطاف پذیرتر و جامع تر شیوه ای که معمولاً با روشهای معمولی در دسترس است.به خصوص، روش ما قادر به ضبط ناهمگونی اثر سیاست هم در داخل و هم است در زیر گروه های جمعیت تعریف شده توسط خصوصیات قابل مشاهده.به به این هدف ، ما از مربع های حداقل جزئی برای شناسایی اجزای هدف استفاده می کنیم جمعیت و جنگل های علّی برای برآورد اثرات سیاسی شخصی در سراسر این مؤلفه هاما نشان می دهیم که این روش سازگار است و به آن منتهی می شود اثرات سیاست به طور غیر عادی توزیع شده است.برای نشان دادن اثربخشی از رویکرد ما ، ما آن را در داده های استفاده مجدد از پنسیلوانیا اعمال می کنیم آزمایش های جایزه ، که در سالهای 1988-1989 انجام شد.تجزیه و تحلیل نشان می دهد که مشوق های مالی می تواند برخی از افراد جوان غیر سفید پوست را برای ورود به آن انگیزه دهد بازار کار.با این حال ، این انگیزه ها همچنین ممکن است یک مالی موقت را فراهم کنند کوسن برای دیگران ، آنها را از جستجوی فعال اشتغال منصرف کرد.ما یافته ها نیاز به اقدامات هدفمند و شخصی برای جوانان را برجسته می کند شرکت کنندگان مرد غیر سفید پوست.
29,681
This paper proposes a novel testing procedure for selecting a sparse set of covariates that explains a large dimensional panel. Our selection method provides correct false detection control while having higher power than existing approaches. We develop the inferential theory for large panels with many covariates by combining post-selection inference with a novel multiple testing adjustment. Our data-driven hypotheses are conditional on the sparse covariate selection. We control for family-wise error rates for covariate discovery for large cross-sections. As an easy-to-use and practically relevant procedure, we propose Panel-PoSI, which combines the data-driven adjustment for panel multiple testing with valid post-selection p-values of a generalized LASSO, that allows us to incorporate priors. In an empirical study, we select a small number of asset pricing factors that explain a large cross-section of investment strategies. Our method dominates the benchmarks out-of-sample due to its better size and power.
در این مقاله یک روش آزمایش جدید برای انتخاب مجموعه ای پراکنده ارائه شده است متغیرهای متغیر که یک صفحه بزرگ بعدی را توضیح می دهد.روش انتخاب ما کنترل صحیح تشخیص کاذب را در حالی که قدرت بالاتری نسبت به دارد فراهم می کند رویکردهای موجودما تئوری استنباطی را برای پانل های بزرگ توسعه می دهیم بسیاری تنظیم آزمایش.فرضیه های داده محور ما مشروط به پراکنده هستند انتخاب متغیر.ما برای نرخ خطای خانواده عاقلانه برای همبستگی کنترل می کنیم کشف برای مقاطع بزرگ.به عنوان یک کاربرد آسان و عملاً مرتبط است رویه ، ما پانل-پوسنی را پیشنهاد می کنیم ، که ترکیب داده محور را برای آن ترکیب می کند تست چند پانل با مقادیر P معتبر پس از انتخاب یک عمومی Lasso ، که به ما امکان می دهد Priors را بگنجانیم.در یک مطالعه تجربی ، ما را انتخاب می کنیم تعداد کمی از فاکتورهای قیمت گذاری دارایی که بخش بزرگی از آن را توضیح می دهد استراتژی های سرمایه گذاری.روش ما به دلیل معیارهای خارج از نمونه حاکم است اندازه و قدرت بهتر آن است.
29,682
This paper examines the dynamics of Tether, the stablecoin with the largest market capitalization. We show that the distributional and dynamic properties of Tether/USD rates have been evolving from 2017 to 2021. We use local analysis methods to detect and describe the local patterns, such as short-lived trends, time-varying volatility and persistence. To accommodate these patterns, we consider a time varying parameter Double Autoregressive tvDAR(1) model under the assumption of local stationarity of Tether/USD rates. We estimate the tvDAR model non-parametrically and test hypotheses on the functional parameters. In the application to Tether, the model provides a good fit and reliable out-of-sample forecasts at short horizons, while being robust to time-varying persistence and volatility. In addition, the model yields a simple plug-in measure of stability for Tether and other stablecoins for assessing and comparing their stability.
در این مقاله به پویایی Tether ، Stablecoin با بزرگترین سرمایه گذاری در بازار.ما نشان می دهیم که خصوصیات توزیع و پویا نرخ Tether/USD از سال 2017 به 2021 در حال تحول است. ما از تجزیه و تحلیل محلی استفاده می کنیم روشهای تشخیص و توصیف الگوهای محلی مانند روندهای کوتاه مدت ، نوسانات و تداوم متغیر زمان.برای قرار دادن این الگوها ، ما یک پارامتر متغیر را در نظر بگیرید مدل TVDAR (1) Autoregressive Double Autoregressive Double (1) فرض ثابت بودن محلی نرخ تتریک/USD.ما TVDAR را تخمین می زنیم فرضیه های غیر پارامتری را در مورد پارامترهای عملکردی مدل کنید.که در برنامه Tether ، مدل مناسب و قابل اعتماد را ارائه می دهد پیش بینی های خارج از نمونه در افق های کوتاه ، در حالی که برای متغیر بودن زمان قوی است پایداری و نوسانات.علاوه بر این ، مدل یک افزونه ساده به دست می آورد اندازه گیری ثبات برای تتر و سایر استبکونها برای ارزیابی و مقایسه ثبات آنها.
29,683
Non-causal processes have been drawing attention recently in Macroeconomics and Finance for their ability to display nonlinear behaviors such as asymmetric dynamics, clustering volatility, and local explosiveness. In this paper, we investigate the statistical properties of empirical conditional quantiles of non-causal processes. Specifically, we show that the quantile autoregression (QAR) estimates for non-causal processes do not remain constant across different quantiles in contrast to their causal counterparts. Furthermore, we demonstrate that non-causal autoregressive processes admit nonlinear representations for conditional quantiles given past observations. Exploiting these properties, we propose three novel testing strategies of non-causality for non-Gaussian processes within the QAR framework. The tests are constructed either by verifying the constancy of the slope coefficients or by applying a misspecification test of the linear QAR model over different quantiles of the process. Some numerical experiments are included to examine the finite sample performance of the testing strategies, where we compare different specification tests for dynamic quantiles with the Kolmogorov-Smirnov constancy test. The new methodology is applied to some time series from financial markets to investigate the presence of speculative bubbles. The extension of the approach based on the specification tests to AR processes driven by innovations with heteroskedasticity is studied through simulations. The performance of QAR estimates of non-causal processes at extreme quantiles is also explored.
فرآیندهای غیر مقطعی اخیراً در اقتصاد کلان مورد توجه قرار گرفته است و برای توانایی آنها در نمایش رفتارهای غیرخطی مانند نامتقارن پویایی ، نوسانات خوشه بندی و انفجار محلی.در این مقاله ، ما خصوصیات آماری مقادیر مشروط تجربی را بررسی کنید فرآیندهای غیر مقطعی.به طور خاص ، ما نشان می دهیم که Autoregression Quantile (QAR) تخمین برای فرآیندهای غیر مقطعی در سراسر ثابت نمی ماند مقادیر مختلف بر خلاف همتایان علیت آنها.علاوه بر این ، ما نشان می دهد که فرآیندهای اتورگرایی غیر مقطعی غیرخطی را می پذیرند بازنمایی برای مقادیر مشروط با مشاهدات گذشته.استثمار این خصوصیات ، ما سه استراتژی تست جدید از عدم نیاز را پیشنهاد می کنیم برای فرآیندهای غیر گایزیایی در چارچوب QAR.تست ها ساخته شده اند یا با تأیید ثبات ضرایب شیب یا با استفاده از آزمون شناسایی خطی مدل QAR خطی بر روی مقادیر مختلف روند.برخی از آزمایشات عددی برای بررسی نمونه محدود درج شده است عملکرد استراتژی های آزمایش ، جایی که ما مشخصات مختلف را با آن مقایسه می کنیم تست های مقادیر پویا با تست ثابت Kolmogorov-Smirnov.جدید روش شناسی برای برخی از سری های زمانی از بازارهای مالی تا وجود حباب های سوداگرانه را بررسی کنید.گسترش رویکرد بر اساس تست های مشخصات فرآیندهای AR که توسط نوآوری ها هدایت می شوند ناهمگونی از طریق شبیه سازی مورد مطالعه قرار می گیرد.عملکرد قار برآورد فرآیندهای غیر مقطعی در مقادیر شدید نیز مورد بررسی قرار گرفته است.
29,697
The present paper proposes a new treatment effects estimator that is valid when the number of time periods is small, and the parallel trends condition holds conditional on covariates and unobserved heterogeneity in the form of interactive fixed effects. The estimator also allow the control variables to be affected by treatment and it enables estimation of the resulting indirect effect on the outcome variable. The asymptotic properties of the estimator are established and their accuracy in small samples is investigated using Monte Carlo simulations. The empirical usefulness of the estimator is illustrated using as an example the effect of increased trade competition on firm markups in China.
مقاله حاضر یک برآوردگر اثرات درمانی جدید را ارائه می دهد که معتبر است هنگامی که تعداد دوره های زمانی اندک است ، و شرایط موازی مشروط به متغیرهای متغیر و ناهمگونی بدون نظارت در قالب اثرات ثابت تعاملی.برآوردگر همچنین به متغیرهای کنترل اجازه می دهد تحت تأثیر درمان قرار می گیرد و تخمین غیرمستقیم حاصل را امکان پذیر می کند تأثیر بر متغیر نتیجه.خصوصیات بدون علامت برآوردگر است تأسیس شده و دقت آنها در نمونه های کوچک با استفاده از مونت بررسی شده است شبیه سازی های کارلو.سودمندی تجربی برآوردگر نشان داده شده است استفاده به عنوان نمونه تأثیر افزایش رقابت تجاری در نشانه های شرکت در چین.
29,684
This paper proves a new central limit theorem for a sample that exhibits multi-way dependence and heterogeneity across clusters. Statistical inference for situations where there is both multi-way dependence and cluster heterogeneity has thus far been an open issue. Existing theory for multi-way clustering inference requires identical distributions across clusters (implied by the so-called separate exchangeability assumption). Yet no such homogeneity requirement is needed in the existing theory for one-way clustering. The new result therefore theoretically justifies the view that multi-way clustering is a more robust version of one-way clustering, consistent with applied practice. The result is applied to linear regression, where it is shown that a standard plug-in variance estimator is valid for inference.
در این مقاله یک قضیه محدودیت مرکزی جدید برای نمونه ای که نمایش داده می شود اثبات می کند وابستگی چند طرفه و ناهمگونی در خوشه ها.استنباط آماری برای موقعیت هایی که هم وابستگی چند طرفه و خوشه ای وجود دارد ناهمگونی تاکنون مسئله ای باز بوده است.نظریه موجود برای چند راه استنتاج خوشه بندی نیاز به توزیع یکسان در بین خوشه ها دارد (ضمنی با فرض به اصطلاح مبادله جداگانه).با این حال چنین همگن نیست در تئوری موجود برای خوشه بندی یک طرفه مورد نیاز است.جدید بنابراین نتیجه تئوری این دیدگاه را توجیه می کند که خوشه بندی چند طرفه است یک نسخه قوی تر از خوشه بندی یک طرفه ، مطابق با تمرین کاربردی. نتیجه برای رگرسیون خطی اعمال می شود ، جایی که نشان داده شده است که یک استاندارد برآوردگر واریانس افزونه برای استنتاج معتبر است.
29,685
It is customary to estimate error-in-variables models using higher-order moments of observables. This moments-based estimator is consistent only when the coefficient of the latent regressor is assumed to be non-zero. We develop a new estimator based on the divide-and-conquer principle that is consistent for any value of the coefficient of the latent regressor. In an application on the relation between investment, (mismeasured) Tobin's $q$ and cash flow, we find time periods in which the effect of Tobin's $q$ is not statistically different from zero. The implausibly large higher-order moment estimates in these periods disappear when using the proposed estimator.
تخمین مدل های متغیر خطا با استفاده از مرتبه بالاتر است لحظات مشاهده.این برآوردگر مبتنی بر لحظات فقط زمانی سازگار است ضریب رگرسور نهفته فرض می شود که غیر صفر است.ما توسعه می دهیم برآوردگر جدید بر اساس اصل تقسیم و تسخیر که برای آن سازگار است هر مقدار از ضریب رگرسور نهفته.در یک برنامه در رابطه بین سرمایه گذاری ، (mismeasured) $ q $ توبین و جریان نقدی ، ما می یابیم دوره زمانی که در آن تأثیر $ q $ توبین از نظر آماری متفاوت نیست از صفربرآورد لحظه ای مرتبه بالاتر به طور غیرقانونی در این دوره ها هنگام استفاده از برآوردگر پیشنهادی ناپدید می شوند.
29,686
The overwhelming majority of empirical research that uses cluster-robust inference assumes that the clustering structure is known, even though there are often several possible ways in which a dataset could be clustered. We propose two tests for the correct level of clustering in regression models. One test focuses on inference about a single coefficient, and the other on inference about two or more coefficients. We provide both asymptotic and wild bootstrap implementations. The proposed tests work for a null hypothesis of either no clustering or ``fine'' clustering against alternatives of ``coarser'' clustering. We also propose a sequential testing procedure to determine the appropriate level of clustering. Simulations suggest that the bootstrap tests perform very well under the null hypothesis and can have excellent power. An empirical example suggests that using the tests leads to sensible inferences.
اکثریت قریب به اتفاق تحقیقات تجربی که از خوشه های خوشه ای استفاده می کند استنباط فرض می کند که ساختار خوشه بندی شناخته شده است ، حتی اگر وجود داشته باشد غالباً چندین روش ممکن که می توان یک مجموعه داده را خوشه بندی کرد.ما پیشنهاد می کنیم دو آزمایش برای سطح صحیح خوشه بندی در مدلهای رگرسیون.یک آزمون تمرکز بر استنباط در مورد یک ضریب واحد و دیگری بر استنتاج حدود دو یا چند ضرایب.ما هر دو بوت استستراپ بدون علامت و وحشی را ارائه می دهیم پیاده سازی هاآزمون های پیشنهادی برای یک فرضیه تهی از هر یک کار می کنند خوشه بندی یا "ریز" "خوشه بندی" در برابر گزینه های "درشت" خوشه بندیما همچنین یک روش آزمایش پی در پی برای تعیین سطح مناسب خوشه بندی.شبیه سازی ها نشان می دهد که آزمایش بوت استرپ تحت فرضیه تهی بسیار خوب عمل کنید و می تواند قدرت عالی داشته باشد.در مثال تجربی نشان می دهد که استفاده از تست ها منجر به استنتاج های معقول می شود.
29,687
We provide computationally attractive methods to obtain jackknife-based cluster-robust variance matrix estimators (CRVEs) for linear regression models estimated by least squares. We also propose several new variants of the wild cluster bootstrap, which involve these CRVEs, jackknife-based bootstrap data-generating processes, or both. Extensive simulation experiments suggest that the new methods can provide much more reliable inferences than existing ones in cases where the latter are not trustworthy, such as when the number of clusters is small and/or cluster sizes vary substantially. Three empirical examples illustrate the new methods.
ما روشهای محاسباتی جذاب برای به دست آوردن jackknife را ارائه می دهیم برآوردگرهای ماتریس واریانس خوشه ای (CRVE) برای مدل های رگرسیون خطی حداقل مربعات تخمین زده می شود.ما همچنین چندین نوع جدید از وحشی را پیشنهاد می کنیم Bootstrap خوشه ای ، که شامل این CRVE ، Bootstrap مبتنی بر Jackknife است فرآیندهای تولید داده یا هر دو.آزمایش های شبیه سازی گسترده نشان می دهد اینکه روشهای جدید می توانند استنتاج های بسیار قابل اطمینان تری نسبت به موجود ارائه دهند مواردی که دومی قابل اعتماد نیستند ، مانند تعداد تعداد خوشه ها کوچک هستند و/یا اندازه خوشه ای متفاوت هستند.سه تجربی مثالها روشهای جدید را نشان می دهد.
29,688
I introduce a generic method for inference on entire quantile and regression quantile processes in the presence of a finite number of large and arbitrarily heterogeneous clusters. The method asymptotically controls size by generating statistics that exhibit enough distributional symmetry such that randomization tests can be applied. The randomization test does not require ex-ante matching of clusters, is free of user-chosen parameters, and performs well at conventional significance levels with as few as five clusters. The method tests standard (non-sharp) hypotheses and can even be asymptotically similar in empirically relevant situations. The main focus of the paper is inference on quantile treatment effects but the method applies more broadly. Numerical and empirical examples are provided.
من یک روش عمومی برای استنباط در کل کمی و رگرسیون معرفی می کنم فرآیندهای کمی در حضور تعداد محدودی از بزرگ و خودسرانه خوشه های ناهمگن.این روش به صورت مجانبی اندازه را با تولید کنترل می کند آماری که تقارن توزیع کافی را نشان می دهند به گونه ای که تصادفی سازی آزمایشات قابل استفاده است.آزمون تصادفی سازی نیازی به تطبیق آن ندارد از خوشه ها ، عاری از پارامترهای انتخاب شده کاربر است و عملکرد خوبی دارد سطح اهمیت متعارف با حداقل پنج خوشه.تست های روش فرضیه های استاندارد (غیر Sharp) و حتی می تواند به صورت بدون علامت در هماهنگ باشد شرایط تجربی مرتبط.تمرکز اصلی مقاله استنباط است اثرات درمانی کمی اما این روش به طور گسترده تری اعمال می شود.عددی و نمونه های تجربی ارائه شده است.
29,689
We study causal inference in a setting in which units consisting of pairs of individuals (such as married couples) are assigned randomly to one of four categories: a treatment targeted at pair member A, a potentially different treatment targeted at pair member B, joint treatment, or no treatment. The setup includes the important special case in which the pair members are the same individual targeted by two different treatments A and B. Allowing for endogenous non-compliance, including coordinated treatment takeup, as well as interference across treatments, we derive the causal interpretation of various instrumental variable estimands using weaker monotonicity conditions than in the literature. In general, coordinated treatment takeup makes it difficult to separate treatment interaction from treatment effect heterogeneity. We provide auxiliary conditions and various bounding strategies that may help zero in on causally interesting parameters. As an empirical illustration, we apply our results to a program randomly offering two different treatments, namely tutoring and financial incentives, to first year college students, in order to assess the treatments' effects on academic performance.
ما استنتاج علی را در شرایطی مطالعه می کنیم که در آن واحدهای متشکل از جفت افراد (مانند زوج های متاهل) به طور تصادفی به یکی از چهار نفر اختصاص می یابند دسته بندی ها: درمانی که در عضو جفت A قرار دارد ، به طور بالقوه متفاوت است درمان هدفمند در جفت B ، درمان مشترک یا بدون درمان.در راه اندازی شامل مورد ویژه مهمی است که در آن اعضای زوج هستند همان فردی که توسط دو درمان مختلف A و B انجام می شود. عدم رعایت درون زا ، از جمله تصاحب درمان هماهنگ ، و همچنین تداخل در درمان ها ، ما تفسیر علی مختلف را به دست می آوریم برآوردهای متغیر ابزاری با استفاده از شرایط یکنواختی ضعیف تر از ادبیات.به طور کلی ، برداشتن درمان هماهنگ ، کار را دشوار می کند تعامل درمانی جداگانه از ناهمگونی اثر درمانی.ما فراهم می کنیم شرایط کمکی و استراتژی های مختلف محدودیت که ممکن است به صفر کمک کند پارامترهای جالب توجه.به عنوان یک تصویر تجربی ، ما خود را اعمال می کنیم نتایج به یک برنامه که به طور تصادفی دو درمان مختلف را ارائه می دهد ، یعنی آموزش و مشوق های مالی ، به دانشجویان سال اول ، به منظور اینکه اثرات درمان بر عملکرد دانشگاهی را ارزیابی کنید.
29,696
Many economic and causal parameters of interest depend on generated regressors. Examples include structural parameters in models with endogenous variables estimated by control functions and in models with sample selection, treatment effect estimation with propensity score matching, and marginal treatment effects. Inference with generated regressors is complicated by the very complex expression for influence functions and asymptotic variances. To address this problem, we propose Automatic Locally Robust/debiased GMM estimators in a general setting with generated regressors. Importantly, we allow for the generated regressors to be generated from machine learners, such as Random Forest, Neural Nets, Boosting, and many others. We use our results to construct novel Doubly Robust and Locally Robust estimators for the Counterfactual Average Structural Function and Average Partial Effects in models with endogeneity and sample selection, respectively. We provide sufficient conditions for the asymptotic normality of our debiased GMM estimators and investigate their finite sample performance through Monte Carlo simulations.
بسیاری از پارامترهای اقتصادی و علّی مورد علاقه به تولید شده بستگی دارد رگرسیونمثالها شامل پارامترهای ساختاری در مدل هایی با درون زا است متغیرهای تخمین زده شده توسط توابع کنترل و در مدل هایی با انتخاب نمونه ، برآورد اثر درمانی با تطبیق نمره گرایش و حاشیه اثرات درمانی.استنتاج با رگرسیون تولید شده توسط بیان بسیار پیچیده ای برای توابع تأثیر و واریانس بدون علامت.به این مشکل را برطرف کنید ، ما GMM به صورت محلی قوی/debiased را پیشنهاد می کنیم برآوردگرها در یک محیط کلی با رگرسیون تولید شده.مهمتر ، ما اجازه می دهد تا رگرسیون های تولید شده از زبان آموزان ماشین تولید شوند ، از این دست به عنوان جنگل تصادفی ، شبکه های عصبی ، تقویت و بسیاری دیگر.ما از نتایج خود استفاده می کنیم تا برآوردگرهای دو برابر قوی و قوی برای عملکرد ساختاری متوسط ​​و میانگین اثرات جزئی در به ترتیب مدل هایی با درون زا و انتخاب نمونه.ما فراهم می کنیم شرایط کافی برای عادی بودن بدون علامت GMM ما برآوردگرها و عملکرد نمونه محدود آنها را از طریق مونت کارلو بررسی کنید شبیه سازی
29,690
Robust M-estimation uses loss functions, such as least absolute deviation (LAD), quantile loss and Huber's loss, to construct its objective function, in order to for example eschew the impact of outliers, whereas the difficulty in analysing the resultant estimators rests on the nonsmoothness of these losses. Generalized functions have advantages over ordinary functions in several aspects, especially generalized functions possess derivatives of any order. Generalized functions incorporate local integrable functions, the so-called regular generalized functions, while the so-called singular generalized functions (e.g. Dirac delta function) can be obtained as the limits of a sequence of sufficient smooth functions, so-called regular sequence in generalized function context. This makes it possible to use these singular generalized functions through approximation. Nevertheless, a significant contribution of this paper is to establish the convergence rate of regular sequence to nonsmooth loss that answers a call from the relevant literature. For parameter estimation where objective function may be nonsmooth, this paper first shows as a general paradigm that how generalized function approach can be used to tackle the nonsmooth loss functions in Section two using a very simple model. This approach is of general interest and applicability. We further use the approach in robust M-estimation for additive single-index cointegrating time series models; the asymptotic theory is established for the proposed estimators. We evaluate the finite-sample performance of the proposed estimation method and theory by both simulated data and an empirical analysis of predictive regression of stock returns.
تخمین M قوی از توابع ضرر استفاده می کند ، مانند حداقل انحراف مطلق (LAD) ، از دست دادن کمی و از دست دادن هوبر ، برای ساخت عملکرد عینی آن ، در به عنوان مثال از تأثیر دور از دسترس استفاده کنید ، در حالی که دشواری در تجزیه و تحلیل برآوردگرهای حاصل بر عدم استفاده از این ضررها استوار است. توابع عمومی در چندین مورد مزایایی دارند جنبه ها ، به ویژه توابع تعمیم یافته دارای مشتقات از هر نظم هستند. توابع تعمیم یافته شامل توابع یکپارچه محلی ، به اصطلاح عملکردهای عمومی منظم ، در حالی که به اصطلاح مفرد تعمیم یافته است توابع (به عنوان مثال عملکرد Dirac Delta) را می توان به عنوان محدودیت های a بدست آورد توالی توابع صاف کافی ، به اصطلاح توالی منظم در زمینه عملکرد تعمیم یافته.این امر امکان استفاده از این مفرد را فراهم می آورد توابع تعمیم یافته از طریق تقریب.با این وجود ، قابل توجه سهم این مقاله تعیین میزان همگرایی منظم است دنباله ای به از دست دادن غیرمستقیم که به یک تماس از ادبیات مربوطه پاسخ می دهد. برای برآورد پارامترها که در آن عملکرد هدف ممکن است غیر نرم باشد ، این مقاله ابتدا به عنوان یک الگوی کلی نشان می دهد که رویکرد عملکرد عمومی چگونه می تواند باشد برای مقابله با توابع از دست دادن غیر نرم در بخش دو با استفاده از یک بسیار ساده استفاده می شود مدل.این رویکرد از علاقه عمومی و کاربردی برخوردار است.ما بیشتر استفاده می کنیم رویکرد در تخمین قوی M برای ادغام تک شاخص افزودنی مدل های سری زمانی ؛نظریه بدون علامت برای پیشنهادی ایجاد شده است برآوردگرهاما عملکرد نمونه محدود پیشنهادی را ارزیابی می کنیم روش و نظریه تخمین توسط داده های شبیه سازی شده و یک تحلیل تجربی رگرسیون پیش بینی کننده بازده سهام.
29,691
We revisit conduct parameter estimation in homogeneous goods markets to resolve the conflict between Bresnahan (1982) and Perloff and Shen (2012) regarding the identification and the estimation of conduct parameters. We point out that Perloff and Shen's (2012) proof is incorrect and its simulation setting is invalid. Our simulation shows that estimation becomes accurate when demand shifters are properly added in supply estimation and sample sizes are increased, supporting Bresnahan (1982).
ما برآورد پارامتر را در بازارهای کالاهای همگن دوباره بررسی می کنیم حل و فصل درگیری بین برسناهان (1982) و پرلوف و شن (2012) با توجه به شناسایی و برآورد پارامترهای رفتار.ما اشاره می کنیم این اثبات Perloff و Shen (2012) نادرست و شبیه سازی آن است تنظیم نامعتبر است.شبیه سازی ما نشان می دهد که تخمین دقیق می شود شیفت های تقاضا به درستی در تخمین عرضه اضافه می شوند و اندازه نمونه ها هستند افزایش ، حمایت از برسناهان (1982).
29,692
Evaluating policy in imperfectly competitive markets requires understanding firm behavior. While researchers test conduct via model selection and assessment, we present advantages of Rivers and Vuong (2002) (RV) model selection under misspecification. However, degeneracy of RV invalidates inference. With a novel definition of weak instruments for testing, we connect degeneracy to instrument strength, derive weak instrument properties of RV, and provide a diagnostic for weak instruments by extending the framework of Stock and Yogo (2005) to model selection. We test vertical conduct (Villas-Boas, 2007) using common instrument sets. Some are weak, providing no power. Strong instruments support manufacturers setting retail prices.
ارزیابی سیاست در بازارهای ناقص رقابتی نیاز به درک دارد رفتار محکمدر حالی که محققان رفتار را از طریق انتخاب مدل آزمایش می کنند و ارزیابی ، ما مزایای مدل رودخانه ها و Vuong (2002) (RV) را ارائه می دهیم انتخاب تحت اشتباه.با این حال ، انحطاط RV بی اعتبار است استنتاجبا یک تعریف جدید از ابزارهای ضعیف برای آزمایش ، ما به هم وصل می شویم انحطاط به قدرت ابزار ، به دست آوردن خصوصیات ابزار ضعیف RV ، و با گسترش چارچوب سهام ، تشخیصی برای ابزارهای ضعیف ارائه دهید و یوگو (2005) برای انتخاب مدل.ما رفتار عمودی را آزمایش می کنیم (Villas-Boas ، 2007) با استفاده از مجموعه های ابزار مشترک.برخی ضعیف هستند و هیچ قدرتی ندارند.قوی ابزارها از تولید کنندگان پشتیبانی می کنند که قیمت های خرده فروشی را تعیین می کنند.
29,693
This paper develops a semi-parametric procedure for estimation of unconditional quantile partial effects using quantile regression coefficients. The estimator is based on an identification result showing that, for continuous covariates, unconditional quantile effects are a weighted average of conditional ones at particular quantile levels that depend on the covariates. We propose a two-step estimator for the unconditional effects where in the first step one estimates a structural quantile regression model, and in the second step a nonparametric regression is applied to the first step coefficients. We establish the asymptotic properties of the estimator, say consistency and asymptotic normality. Monte Carlo simulations show numerical evidence that the estimator has very good finite sample performance and is robust to the selection of bandwidth and kernel. To illustrate the proposed method, we study the canonical application of the Engel's curve, i.e. food expenditures as a share of income.
در این مقاله یک روش نیمه پارامتری برای برآورد تهیه شده است اثرات جزئی کمی بی قید و شرط با استفاده از ضرایب رگرسیون کمی. برآوردگر بر اساس نتیجه شناسایی است که نشان می دهد ، برای مداوم متغیرهای متغیر ، اثرات کمی بی قید و شرط میانگین وزنی است مشروط در سطوح کمی خاص که به متغیرهای متغیر بستگی دارند. ما یک برآوردگر دو مرحله ای را برای اثرات بی قید و شرط که در آن پیشنهاد می کنیم پیشنهاد می کنیم مرحله اول یک مدل رگرسیون کمی ساختاری ، و در مرحله دوم رگرسیون غیرپارامتری در مرحله اول اعمال می شود ضرایبما خصوصیات بدون علامت برآوردگر را ایجاد می کنیم. قوام و عادی بودن بدون علامت.شبیه سازی مونت کارلو عددی را نشان می دهد شواهدی مبنی بر اینکه برآوردگر عملکرد نمونه محدود بسیار خوبی دارد و هست قوی به انتخاب پهنای باند و هسته.برای نشان دادن پیشنهادی روش ، ما کاربرد متعارف منحنی انگل ، یعنی غذا را مطالعه می کنیم مخارج به عنوان سهمی از درآمد.
29,694
Many problems ask a question that can be formulated as a causal question: "what would have happened if...?" For example, "would the person have had surgery if he or she had been Black?" To address this kind of questions, calculating an average treatment effect (ATE) is often uninformative, because one would like to know how much impact a variable (such as skin color) has on a specific individual, characterized by certain covariates. Trying to calculate a conditional ATE (CATE) seems more appropriate. In causal inference, the propensity score approach assumes that the treatment is influenced by x, a collection of covariates. Here, we will have the dual view: doing an intervention, or changing the treatment (even just hypothetically, in a thought experiment, for example by asking what would have happened if a person had been Black) can have an impact on the values of x. We will see here that optimal transport allows us to change certain characteristics that are influenced by the variable we are trying to quantify the effect of. We propose here a mutatis mutandis version of the CATE, which will be done simply in dimension one by saying that the CATE must be computed relative to a level of probability, associated to the proportion of x (a single covariate) in the control population, and by looking for the equivalent quantile in the test population. In higher dimension, it will be necessary to go through transport, and an application will be proposed on the impact of some variables on the probability of having an unnatural birth (the fact that the mother smokes, or that the mother is Black).
بسیاری از مشکلات سوالی را مطرح می کنند که می تواند به عنوان یک سوال علّی تدوین شود: "چه اتفاقی می افتد اگر ...؟"به عنوان مثال ، "آیا شخص داشته است جراحی اگر او سیاه بوده است؟ "برای پرداختن به این نوع سؤالات ، محاسبه یک اثر درمانی متوسط ​​(ATE) اغلب ناآگاهانه است ، زیرا کسی می خواهد بداند که یک متغیر (مانند رنگ پوست) چه تاثیری در یک دارد فرد خاص ، که توسط متغیرهای خاص مشخص می شود.تلاش برای محاسبه a به نظر می رسد مشروط (CATE) مناسب تر است.در استنباط علی ، رویکرد نمره گرایش فرض می کند که درمان تحت تأثیر x ، a است مجموعه متغیرهای متغیر.در اینجا ، ما دید دوگانه خواهیم داشت: انجام مداخله ، یا تغییر درمان (حتی فقط فرضی ، در یک فکر به عنوان مثال با پرسیدن اینکه اگر شخصی بوده است چه اتفاقی می افتد سیاه) می تواند در مقادیر x تأثیر بگذارد.ما اینجا را بهینه خواهیم دید حمل و نقل به ما امکان می دهد ویژگی های خاصی را که تحت تأثیر قرار می گیرند تغییر دهیم متغیری که ما در تلاش هستیم تا اثر را تعیین کنیم.ما در اینجا یک mutatis پیشنهاد می کنیم نسخه mutandis از cate ، که به سادگی در ابعاد انجام می شود با بیان اینکه کیت باید نسبت به سطح احتمال محاسبه شود ، مرتبط با نسبت x (یک متغیر متغیر) در کنترل جمعیت و با جستجوی مقدار معادل در جمعیت آزمون. در ابعاد بالاتر ، لازم است از طریق حمل و نقل ، و برنامه در مورد تأثیر برخی متغیرها بر روی احتمال ارائه خواهد شد از تولد غیر طبیعی (این واقعیت که مادر سیگار می کشد ، یا آن مادر سیاه است).
29,695
This paper revisits the identification and estimation of a class of semiparametric (distribution-free) panel data binary choice models with lagged dependent variables, exogenous covariates, and entity fixed effects. We provide a novel identification strategy, using an "identification at infinity" argument. In contrast with the celebrated Honore and Kyriazidou (2000), our method permits time trends of any form and does not suffer from the "curse of dimensionality". We propose an easily implementable conditional maximum score estimator. The asymptotic properties of the proposed estimator are fully characterized. A small-scale Monte Carlo study demonstrates that our approach performs satisfactorily in finite samples. We illustrate the usefulness of our method by presenting an empirical application to enrollment in private hospital insurance using the Household, Income and Labour Dynamics in Australia (HILDA) Survey data.
در این مقاله ، شناسایی و تخمین یک کلاس از داده های صفحه نمایش باینری (بدون توزیع) داده های پانری با تاخیر متغیرهای وابسته ، متغیرهای اگزوژن و اثرات ثابت موجودیت.ما فراهم می کنیم یک استراتژی شناسایی جدید ، با استفاده از "شناسایی در Infinity" بحث و جدل.در مقابل با افتخار مشهور و کیریازیدو (2000) ، ما روش اجازه می دهد روند زمان به هر شکلی داشته باشد و از "نفرین" رنج نمی برد ابعاد ". ما حداکثر نمره مشروط به راحتی قابل اجرا را پیشنهاد می کنیم برآوردگرخصوصیات بدون علامت برآوردگر پیشنهادی کاملاً است مشخصهیک مطالعه مونت کارلو در مقیاس کوچک نشان می دهد که رویکرد ما در نمونه های محدود به طور رضایت بخش عمل می کند.ما سودمندی خود را نشان می دهیم روش با ارائه یک برنامه تجربی برای ثبت نام در بیمارستان خصوصی بیمه با استفاده از خانواده ، درآمد و پویایی کار در استرالیا (هیلدا) داده های نظرسنجی.
29,698
In this paper, we describe a computational implementation of the Synthetic difference-in-differences (SDID) estimator of Arkhangelsky et al. (2021) for Stata. Synthetic difference-in-differences can be used in a wide class of circumstances where treatment effects on some particular policy or event are desired, and repeated observations on treated and untreated units are available over time. We lay out the theory underlying SDID, both when there is a single treatment adoption date and when adoption is staggered over time, and discuss estimation and inference in each of these cases. We introduce the sdid command which implements these methods in Stata, and provide a number of examples of use, discussing estimation, inference, and visualization of results.
در این مقاله ، ما یک اجرای محاسباتی از مصنوعی را شرح می دهیم برآوردگر اختلاف در اختلافات (SDID) از Arkhangelsky و همکاران.(2021) برای استاتاتفاوتهای مصنوعی در تمایزها را می توان در یک کلاس گسترده از شرایطی که تأثیرات درمانی بر برخی از سیاست ها یا رویداد های خاص وجود دارد مشاهدات مورد نظر و مکرر در مورد واحدهای تحت درمان و درمان نشده در دسترس است در طول زمان.ما تئوری زیربنایی SDID را بیان می کنیم ، هر دو وقتی یک واحد وجود دارد تاریخ پذیرش درمانی و هنگامی که فرزندخواندگی با گذشت زمان مبهم است ، و بحث کنید برآورد و استنباط در هر یک از این موارد.ما دستور SDID را معرفی می کنیم که این روش ها را در Stata پیاده سازی می کند ، و چندین نمونه از آن را ارائه می دهد استفاده ، بحث در مورد تخمین ، استنباط و تجسم نتایج.
29,699
This paper introduces a maximum likelihood estimator of the value of job amenities and labor productivity in a single matching market based on the observation of equilibrium matches and wages. The estimation procedure simultaneously fits both the matching patterns and the wage curve. While our estimator is suited for a wide range of assignment problems, we provide an application to the estimation of the Value of a Statistical Life using compensating wage differentials for the risk of fatal injury on the job. Using US data for 2017, we estimate the Value of Statistical Life at \$ 6.3 million (\$2017).
در این مقاله یک برآوردگر حداکثر احتمال از ارزش شغل ارائه شده است امکانات و بهره وری نیروی کار در یک بازار تطبیق بر اساس مشاهده مسابقات تعادل و دستمزد.روش تخمین همزمان با الگوی تطبیق و منحنی دستمزد متناسب است.در حالی که ما برآوردگر برای طیف گسترده ای از مشکلات واگذاری مناسب است ، ما کاربرد برآورد ارزش یک زندگی آماری با استفاده از جبران دیفرانسیل دستمزد برای خطر آسیب دیدگی کشنده در کار.استفاده كردن داده های ایالات متحده برای سال 2017 ، ما ارزش زندگی آماری را با قیمت 6.3 میلیون دلار تخمین می زنیم (\ $ 2017).
29,700
Modeling and predicting extreme movements in GDP is notoriously difficult and the selection of appropriate covariates and/or possible forms of nonlinearities are key in obtaining precise forecasts. In this paper, our focus is on using large datasets in quantile regression models to forecast the conditional distribution of US GDP growth. To capture possible non-linearities, we include several nonlinear specifications. The resulting models will be huge dimensional and we thus rely on a set of shrinkage priors. Since Markov Chain Monte Carlo estimation becomes slow in these dimensions, we rely on fast variational Bayes approximations to the posterior distribution of the coefficients and the latent states. We find that our proposed set of models produces precise forecasts. These gains are especially pronounced in the tails. Using Gaussian processes to approximate the nonlinear component of the model further improves the good performance, in particular in the right tail.
مدل سازی و پیش بینی حرکات شدید در تولید ناخالص داخلی بسیار دشوار است انتخاب متغیرهای متغیر و/یا اشکال احتمالی غیرخطی در به دست آوردن پیش بینی های دقیق مهم هستند.در این مقاله ، تمرکز ما روی استفاده است مجموعه داده های بزرگ در مدل های رگرسیون کمی برای پیش بینی مشروط توزیع رشد تولید ناخالص داخلی ایالات متحده.برای گرفتن غیر خطی های احتمالی ، ما را شامل می شویم چندین مشخصات غیرخطی.مدل های حاصل ابعاد عظیم خواهند بود و ما به این ترتیب به مجموعه ای از مقدمات انقباض تکیه می کنیم.از آنجا که زنجیره مارکوف مونت کارلو برآورد در این ابعاد کند می شود ، ما به خلیج های تغییر سرعت متکی هستیم تقریب به توزیع خلفی ضرایب و نهفته ایالت ها.ما می دانیم که مجموعه مدل های پیشنهادی ما پیش بینی های دقیقی را تولید می کند. این دستاوردها به ویژه در دمها تلفظ می شوند.با استفاده از فرآیندهای گاوسی به تقریبی مؤلفه غیرخطی مدل باعث بهبود بیشتر می شود عملکرد ، به ویژه در دم مناسب.
29,701
Inference on common parameters in panel data models with individual-specific fixed effects is a classic example of Neyman and Scott's (1948) incidental parameter problem (IPP). One solution to this IPP is functional differencing (Bonhomme 2012), which works when the number of time periods T is fixed (and may be small), but this solution is not applicable to all panel data models of interest. Another solution, which applies to a larger class of models, is "large-T" bias correction (pioneered by Hahn and Kuersteiner 2002 and Hahn and Newey 2004), but this is only guaranteed to work well when T is sufficiently large. This paper provides a unified approach that connects those two seemingly disparate solutions to the IPP. In doing so, we provide an approximate version of functional differencing, that is, an approximate solution to the IPP that is applicable to a large class of panel data models even when T is relatively small.
استنتاج در پارامترهای مشترک در مدل های داده پانل با خاص فردی جلوه های ثابت یک نمونه کلاسیک از حادثه نیان و اسکات (1948) است مشکل پارامتر (IPP).یک راه حل برای این IPP متفاوت است (Bonhomme 2012) ، که هنگامی که تعداد دوره های زمانی t ثابت باشد ، کار می کند (و ممکن است کوچک باشد) ، اما این راه حل برای کلیه مدلهای داده پانل کاربردی نیست علاقه.راه حل دیگر ، که در مورد یک کلاس بزرگتر از مدل ها اعمال می شود ، این است تصحیح تعصب "بزرگ-T" (پیشگام هان و کویرستینر 2002 و هان و Newey 2004) ، اما این فقط تضمین شده است که وقتی T به اندازه کافی خوب کار کند بزرگاین مقاله یک رویکرد یکپارچه را ارائه می دهد که آن دو به ظاهر را به هم وصل می کند راه حل های متفاوت برای IPP.با انجام این کار ، ما یک نسخه تقریبی ارائه می دهیم از تفاوت عملکردی ، یعنی یک راه حل تقریبی برای IPP که هست قابل استفاده برای یک کلاس بزرگ از مدل های داده پانل حتی اگر T نسبتاً باشد کم اهمیت.
29,702
Thousands of papers have reported two-way cluster-robust (TWCR) standard errors. However, the recent econometrics literature points out the potential non-gaussianity of two-way cluster sample means, and thus invalidity of the inference based on the TWCR standard errors. Fortunately, simulation studies nonetheless show that the gaussianity is rather common than exceptional. This paper provides theoretical support for this encouraging observation. Specifically, we derive a novel central limit theorem for two-way clustered triangular arrays that justifies the use of the TWCR under very mild and interpretable conditions. We, therefore, hope that this paper will provide a theoretical justification for the legitimacy of most, if not all, of the thousands of those empirical papers that have used the TWCR standard errors. We provide a guide in practice as to when a researcher can employ the TWCR standard errors.
هزاران مقاله استاندارد دو طرفه خوشه-روبست (TWCR) را گزارش کرده اند خطابا این حال ، ادبیات اخیر اقتصاد سنجی به پتانسیل اشاره می کند غیر GEASSIANITY از نمونه خوشه دو طرفه ، و در نتیجه بی اعتبار بودن استنتاج بر اساس خطاهای استاندارد TWCR.خوشبختانه ، مطالعات شبیه سازی با این وجود نشان می دهد که گاوسی نسبت به استثنایی رایج است.این مقاله پشتیبانی نظری را برای این مشاهده دلگرم کننده ارائه می دهد. به طور خاص ، ما یک قضیه محدودیت مرکزی جدید برای خوشه های دو طرفه استخراج می کنیم آرایه های مثلثی که استفاده از TWCR را در زیر بسیار خفیف و شرایط قابل تفسیربنابراین ، ما امیدواریم که این مقاله یک توجیه نظری برای مشروعیت اکثر ، اگر نه همه ، از هزاران مقاله تجربی که از خطاهای استاندارد TWCR استفاده کرده اند.ما راهنمایی در عمل ارائه دهید تا زمانی که یک محقق می تواند TWCR را به کار گیرد خطاهای استاندارد
29,703
This paper is concerned with estimation and inference on average treatment effects in randomized controlled trials when researchers observe potentially many covariates. By employing Neyman's (1923) finite population perspective, we propose a bias-corrected regression adjustment estimator using cross-fitting, and show that the proposed estimator has favorable properties over existing alternatives. For inference, we derive the first and second order terms in the stochastic component of the regression adjustment estimators, study higher order properties of the existing inference methods, and propose a bias-corrected version of the HC3 standard error. The proposed methods readily extend to stratified experiments with large strata. Simulation studies show our cross-fitted estimator, combined with the bias-corrected HC3, delivers precise point estimates and robust size controls over a wide range of DGPs. To illustrate, the proposed methods are applied to real dataset on randomized experiments of incentives and services for college achievement following Angrist, Lang, and Oreopoulos (2009).
این مقاله مربوط به تخمین و استنباط در درمان متوسط ​​است اثرات در کارآزمایی های کنترل شده تصادفی هنگامی که محققان به طور بالقوه مشاهده می کنند بسیاری از متغیرهای متغیر.ما با استفاده از دیدگاه جمعیت محدود Neyman (1923) ، ما یک برآوردگر تنظیم رگرسیون اصلاح شده با تعصب را با استفاده از متناسب با متناسب پیشنهاد دهید ، و نشان دهید که برآوردگر پیشنهادی نسبت به موجود دارای خواص مطلوب است جایگزین، گزینه ها.برای استنباط ، ما شرایط مرتبه اول و دوم را در مؤلفه تصادفی برآوردگرهای تنظیم رگرسیون ، مطالعه بالاتر خواص سفارش روشهای استنباط موجود ، و پیشنهاد a نسخه تعصب اصلاح شده از خطای استاندارد HC3.روشهای پیشنهادی به راحتی تا آزمایشات طبقه بندی شده با اقشار بزرگ گسترش یابد.مطالعات شبیه سازی ما را نشان می دهد برآوردگر متقاطع ، همراه با HC3 اصلاح شده با تعصب ، دقیق را ارائه می دهد برآورد نقطه و کنترل اندازه قوی در طیف گسترده ای از DGP ها.به نشان می دهد ، روشهای پیشنهادی در مجموعه داده های واقعی به صورت تصادفی اعمال می شود آزمایش های مشوق ها و خدمات برای پیشرفت کالج در زیر Angrist ، Lang و Oreopoulos (2009).
29,704
In this paper we construct an inferential procedure for Granger causality in high-dimensional non-stationary vector autoregressive (VAR) models. Our method does not require knowledge of the order of integration of the time series under consideration. We augment the VAR with at least as many lags as the suspected maximum order of integration, an approach which has been proven to be robust against the presence of unit roots in low dimensions. We prove that we can restrict the augmentation to only the variables of interest for the testing, thereby making the approach suitable for high dimensions. We combine this lag augmentation with a post-double-selection procedure in which a set of initial penalized regressions is performed to select the relevant variables for both the Granger causing and caused variables. We then establish uniform asymptotic normality of a second-stage regression involving only the selected variables. Finite sample simulations show good performance, an application to investigate the (predictive) causes and effects of economic uncertainty illustrates the need to allow for unknown orders of integration.
در این مقاله ما یک روش استنباطی برای علیت گرنجر در ایجاد می کنیم مدلهای وکتور غیر ثابت (VAR) (VAR) با ابعاد بالا.روش ما نیازی به دانش در مورد ترتیب ادغام سری های زمانی در زیر ندارد توجه.ما VAR را با حداقل تاخیر به عنوان مظنون افزایش می دهیم حداکثر ترتیب ادغام ، رویکردی که ثابت شده است قوی است در برابر حضور ریشه های واحد در ابعاد کم.ما ثابت می کنیم که می توانیم تقویت را فقط به متغیرهای مورد علاقه برای آزمایش محدود کنید ، بدین ترتیب این رویکرد برای ابعاد بالا مناسب است.ما این تاخیر را با هم ترکیب می کنیم افزودنی با یک روش انتخاب پس از دو برابر که در آن مجموعه ای از اولیه است رگرسیون مجازات شده برای انتخاب متغیرهای مربوطه برای هر دو انجام می شود Granger باعث ایجاد متغیرها می شود.ما سپس بدون علامت یکنواخت ایجاد می کنیم نرمال بودن رگرسیون مرحله دوم که فقط متغیرهای انتخاب شده را شامل می شود. شبیه سازی نمونه محدود عملکرد خوب را نشان می دهد ، برنامه ای برای بررسی (پیش بینی کننده) علل و اثرات عدم قطعیت اقتصادی نشان می دهد نیاز به اجازه دادن به سفارشات ناشناخته ادغام.
29,705
When agents' information is imperfect and dispersed, existing measures of macroeconomic uncertainty based on the forecast error variance have two distinct drivers: the variance of the economic shock and the variance of the information dispersion. The former driver increases uncertainty and reduces agents' disagreement (agreed uncertainty). The latter increases both uncertainty and disagreement (disagreed uncertainty). We use these implications to identify empirically the effects of agreed and disagreed uncertainty shocks, based on a novel measure of consumer disagreement derived from survey expectations. Disagreed uncertainty has no discernible economic effects and is benign for economic activity, but agreed uncertainty exerts significant depressing effects on a broad spectrum of macroeconomic indicators.
هنگامی که اطلاعات نمایندگان ناقص و پراکنده است ، اقدامات موجود از عدم قطعیت کلان براساس واریانس خطای پیش بینی دو محرک های مجزا: واریانس شوک اقتصادی و واریانس پراکندگی اطلاعاتراننده سابق عدم اطمینان را افزایش داده و کاهش می دهد اختلاف نظر نمایندگان (عدم اطمینان توافق شده).دومی هر دو را افزایش می دهد عدم اطمینان و اختلاف نظر (عدم اطمینان مخالف).ما از این پیامدها استفاده می کنیم برای شناسایی تجربی اثرات شوک های توافق شده و مخالفت ، عدم اطمینان ، بر اساس یک معیار جدید از اختلاف مصرف کننده ناشی از نظرسنجی انتظاراتعدم اطمینان مخالف هیچ اثرات اقتصادی قابل تشخیص ندارد و هست خوش خیم برای فعالیت اقتصادی ، اما عدم اطمینان موافق قابل توجه است اثرات افسردگی بر روی طیف گسترده ای از شاخص های کلان اقتصادی.
29,706
This document presents an overview of the bayesmixedlogit and bayesmixedlogitwtp Stata packages. It mirrors closely the helpfile obtainable in Stata (i.e., through help bayesmixedlogit or help bayesmixedlogitwtp). Further background for the packages can be found in Baker(2014).
این سند مروری بر BayesmixedLogit و بسته های Stata BayesmixedLogitWtp.آن را از نزدیک به دست آوردن راهنما می کند در Stata (به عنوان مثال ، از طریق Help BayesmixedLogit یا کمک به BayesmixedLogitWtp). پیش زمینه بیشتر بسته ها را می توان در بیکر (2014) یافت.
29,707
This Appendix (dated: July 2021) includes supplementary derivations related to the main limit results of the econometric framework for structural break testing in predictive regression models based on the OLS-Wald and IVX-Wald test statistics, developed by Katsouris C (2021). In particular, we derive the asymptotic distributions of the test statistics when the predictive regression model includes either mildly integrated or persistent regressors. Moreover, we consider the case in which a model intercept is included in the model vis-a-vis the case that the predictive regression model has no model intercept. In a subsequent version of this study we reexamine these particular aspects in more depth with respect to the demeaned versions of the variables of the predictive regression.
این پیوست (مورخ: ژوئیه 2021) شامل مشتقات تکمیلی مربوط به آن است به نتایج اصلی چارچوب اقتصاد سنجی برای شکستن ساختاری آزمایش در مدل های رگرسیون پیش بینی بر اساس تست OLS-WALD و IVX-WALD آمار ، توسط کاتسوریس C (2021) تهیه شده است.به طور خاص ، ما استخراج می کنیم توزیع بدون علامت آمار آزمون هنگام رگرسیون پیش بینی مدل شامل رگرسیونرهای خفیف یکپارچه یا مداوم است.علاوه بر این ، ما موردی را در نظر بگیرید که در آن یک مدل رهگیری در مدل Vis-a-vis گنجانده شده است این مورد که مدل رگرسیون پیش بینی کننده هیچ رهگیری مدل ندارد.در یک نسخه بعدی این مطالعه ما این جنبه های خاص را در موارد دیگر بررسی می کنیم عمق با توجه به نسخه های متنوع متغیرهای پیش بینی پسرفت.
29,708
This paper is concerned with detecting the presence of out of sample predictability in linear predictive regressions with a potentially large set of candidate predictors. We propose a procedure based on out of sample MSE comparisons that is implemented in a pairwise manner using one predictor at a time and resulting in an aggregate test statistic that is standard normally distributed under the global null hypothesis of no linear predictability. Predictors can be highly persistent, purely stationary or a combination of both. Upon rejection of the null hypothesis we subsequently introduce a predictor screening procedure designed to identify the most active predictors. An empirical application to key predictors of US economic activity illustrates the usefulness of our methods and highlights the important forward looking role played by the series of manufacturing new orders.
این مقاله به تشخیص وجود خارج از نمونه مربوط می شود پیش بینی پذیری در رگرسیون پیش بینی خطی با مجموعه بالقوه بزرگی از پیش بینی کننده های نامزد.ما روشی را بر اساس نمونه MSE پیشنهاد می کنیم مقایسه هایی که با استفاده از یک پیش بینی کننده در a به صورت جفت اجرا می شوند زمان و در نتیجه یک آمار تست کل که به طور عادی استاندارد است تحت فرضیه NULL جهانی از پیش بینی خطی توزیع شده است. پیش بینی کنندگان می توانند بسیار پایدار ، صرفاً ثابت یا ترکیبی از باشند هر دو.پس از رد فرضیه تهی ، متعاقباً ما را معرفی می کنیم روش غربالگری پیش بینی شده برای شناسایی فعال ترین پیش بینی کننده ها طراحی شده است. یک برنامه تجربی برای پیش بینی کننده های کلیدی فعالیت اقتصادی ایالات متحده نشان می دهد سودمندی روشهای ما و نقش مهم رو به جلو را برجسته می کند بازی شده توسط سری ساخت سفارشات جدید.
29,709
We extend the theory from Fan and Li (2001) on penalized likelihood-based estimation and model-selection to statistical and econometric models which allow for non-negativity constraints on some or all of the parameters, as well as time-series dependence. It differs from classic non-penalized likelihood estimation, where limiting distributions of likelihood-based estimators and test-statistics are non-standard, and depend on the unknown number of parameters on the boundary of the parameter space. Specifically, we establish that the joint model selection and estimation, results in standard asymptotic Gaussian distributed estimators. The results are applied to the rich class of autoregressive conditional heteroskedastic (ARCH) models for the modelling of time-varying volatility. We find from simulations that the penalized estimation and model-selection works surprisingly well even for a large number of parameters. A simple empirical illustration for stock-market returns data confirms the ability of the penalized estimation to select ARCH models which fit nicely the autocorrelation function, as well as confirms the stylized fact of long-memory in financial time series data.
ما این تئوری را از فن و لی (2001) در مورد احتمال مجازات گسترش می دهیم تخمین و انتخاب مدل به مدلهای آماری و اقتصاد سنجی که محدودیت های غیر منفی را در برخی یا همه پارامترها نیز فراهم کنید به عنوان وابستگی سری زمانی.این با احتمال کلاسیک غیر حرکتی متفاوت است تخمین ، جایی که محدود کننده توزیع برآوردگرهای مبتنی بر احتمال و آمار تست غیر استاندارد است و به تعداد ناشناخته بستگی دارد پارامترهای در مرز فضای پارامتر.به طور خاص ، ما ایجاد می کنیم که انتخاب و تخمین مدل مشترک ، منجر به بدون علامت استاندارد می شود برآوردگرهای توزیع شده گاوسی.نتایج برای کلاس غنی از مدلهای هتروسیکاستی مشروط خودجوش (ARCH) برای مدل سازی نوسانات متغیر زمان.ما از شبیه سازی ها می فهمیم که تخمین مجازات شده و انتخاب مدل حتی برای تعداد زیادی از مولفه های.یک تصویر ساده تجربی برای بازار سهام بازده داده ها توانایی تخمین مجازات شده برای انتخاب مدل های قوس را تأیید می کند عملکرد همبستگی را به خوبی متناسب کنید ، و همچنین واقعیت تلطیف شده را تأیید می کند از حافظه طولانی در داده های سری زمانی مالی.
29,710
We develop asymptotic approximation results that can be applied to sequential estimation and inference problems, adaptive randomized controlled trials, and other statistical decision problems that involve multiple decision nodes with structured and possibly endogenous information sets. Our results extend the classic asymptotic representation theorem used extensively in efficiency bound theory and local power analysis. In adaptive settings where the decision at one stage can affect the observation of variables in later stages, we show that a limiting data environment characterizes all limit distributions attainable through a joint choice of an adaptive design rule and statistics applied to the adaptively generated data, under local alternatives. We illustrate how the theory can be applied to study the choice of adaptive rules and end-of-sample statistical inference in batched (groupwise) sequential adaptive experiments.
ما نتایج تقریب مجانبی را ایجاد می کنیم که می تواند برای پی در پی استفاده شود مشکلات تخمین و استنباط ، کارآزمایی های کنترل شده تصادفی سازگار و سایر مشکلات تصمیم گیری آماری که شامل چندین گره تصمیم گیری است مجموعه اطلاعات ساختاری و احتمالاً درون زا.نتایج ما گسترش می یابد قضیه نمایشی بدون علامت کلاسیک که به طور گسترده در کارآیی مورد استفاده قرار می گیرد تئوری و تحلیل قدرت محلی.در تنظیمات تطبیقی ​​که در آن تصمیم در یک مرحله می تواند در مراحل بعدی بر مشاهده متغیرها تأثیر بگذارد ، ما نشان می دهیم که a محدود کردن محیط داده ها کلیه توزیع های محدود قابل دستیابی را مشخص می کند از طریق یک انتخاب مشترک از یک قانون طراحی تطبیقی ​​و آمار اعمال شده برای داده های سازگار به صورت سازگار ، تحت گزینه های محلی.ما نشان می دهیم که چگونه تئوری می تواند برای مطالعه انتخاب قوانین تطبیقی ​​و پایان نمونه استفاده شود استنباط آماری در آزمایشات تطبیقی ​​متوالی batched (groupwise).
29,711
In settings with few treated units, Difference-in-Differences (DID) estimators are not consistent, and are not generally asymptotically normal. This poses relevant challenges for inference. While there are inference methods that are valid in these settings, some of these alternatives are not readily available when there is variation in treatment timing and heterogeneous treatment effects; or for deriving uniform confidence bands for event-study plots. We present alternatives in settings with few treated units that are valid with variation in treatment timing and/or that allow for uniform confidence bands.
در تنظیمات با چند واحد تحت درمان ، تفاوت در تفاوت (DID) برآوردگرها سازگار نیستند و به طور کلی از نظر غیر عادی طبیعی نیستند. این چالش های مربوط به استنباط را ایجاد می کند.در حالی که روش های استنتاج وجود دارد که در این تنظیمات معتبر هستند ، برخی از این گزینه ها به راحتی نیستند در صورت وجود تغییر در زمان بندی و ناهمگن در دسترس است اثرات درمانی ؛یا برای به دست آوردن گروههای اعتماد به نفس یکنواخت برای مطالعه رویداد توطئه هاما گزینه های دیگری را در تنظیمات با چند واحد تحت درمان ارائه می دهیم که هستند معتبر با تغییر در زمان بندی درمانی و/یا این امکان را فراهم می آورد گروههای اعتماد به نفس.
29,850
We study mixed-effects methods for estimating equations containing person and firm effects. In economics such models are usually estimated using fixed-effects methods. Recent enhancements to those fixed-effects methods include corrections to the bias in estimating the covariance matrix of the person and firm effects, which we also consider.
ما روشهای اثرات مختلط را برای برآورد معادلات حاوی شخص مطالعه می کنیم اثرات محکم.در اقتصاد چنین مدلهایی معمولاً با استفاده از آن تخمین زده می شود روشهای اثر ثابت.پیشرفت های اخیر به آن روش های اثر ثابت شامل اصلاحات در تعصب در برآورد ماتریس کواریانس اثرات شخص و محکم ، که ما نیز در نظر می گیریم.
29,712
In this paper we test for Granger causality in high-dimensional vector autoregressive models (VARs) to disentangle and interpret the complex causal chains linking radiative forcings and global temperatures. By allowing for high dimensionality in the model we can enrich the information set with all relevant natural and anthropogenic forcing variables to obtain reliable causal relations. These variables have mostly been investigated in an aggregated form or in separate models in the previous literature. Additionally, our framework allows to ignore the order of integration of the variables and to directly estimate the VAR in levels, thus avoiding accumulating biases coming from unit-root and cointegration tests. This is of particular appeal for climate time series which are well known to contain stochastic trends as well as yielding long memory. We are thus able to display the causal networks linking radiative forcings to global temperatures but also to causally connect radiative forcings among themselves, therefore allowing for a careful reconstruction of a timeline of causal effects among forcings. The robustness of our proposed procedure makes it an important tool for policy evaluation in tackling global climate change.
در این مقاله ما برای علیت گرنجر در بردار با ابعاد بالا آزمایش می کنیم مدل های خودجوش (VARS) برای جدا کردن و تفسیر علت پیچیده زنجیرهای پیوند دهنده های تابشی و دمای جهانی.با اجازه بالا ابعادی در مدل می توانیم اطلاعات مجموعه ای را با همه موارد مرتبط غنی کنیم متغیرهای مجبور طبیعی و انسانی برای به دست آوردن علت قابل اعتماد روابطاین متغیرها بیشتر به صورت جمع شده مورد بررسی قرار گرفته اند یا در مدلهای جداگانه در ادبیات قبلی.علاوه بر این ، چارچوب ما اجازه می دهد تا ترتیب ادغام متغیرها و مستقیماً نادیده بگیرند VAR را در سطح تخمین بزنید ، بنابراین از تجمع تعصبات ناشی از آن جلوگیری می شود تست های واحد ریشه ای و ادغام.این جذابیت خاصی برای آب و هوا است سری زمانی که به خوبی شناخته شده است حاوی روند تصادفی و همچنین به دست آوردن حافظه طولانی.بنابراین ما قادر به نمایش شبکه های علی هستیم نیروهای تابشی به دمای جهانی بلکه به اتصال علّی نیز اجبار تابشی در بین خود ، بنابراین امکان دقت را فراهم می آورد بازسازی یک جدول زمانی از اثرات علّی در بین نیروها.استحکام از روش پیشنهادی ما آن را به ابزاری مهم برای ارزیابی سیاست در مقابله با تغییرات جهانی آب و هوا.
29,713
This paper studies inference on the average treatment effect in experiments in which treatment status is determined according to "matched pairs" and it is additionally desired to adjust for observed, baseline covariates to gain further precision. By a "matched pairs" design, we mean that units are sampled i.i.d. from the population of interest, paired according to observed, baseline covariates and finally, within each pair, one unit is selected at random for treatment. Importantly, we presume that not all observed, baseline covariates are used in determining treatment assignment. We study a broad class of estimators based on a "doubly robust" moment condition that permits us to study estimators with both finite-dimensional and high-dimensional forms of covariate adjustment. We find that estimators with finite-dimensional, linear adjustments need not lead to improvements in precision relative to the unadjusted difference-in-means estimator. This phenomenon persists even if the adjustments are interacted with treatment; in fact, doing so leads to no changes in precision. However, gains in precision can be ensured by including fixed effects for each of the pairs. Indeed, we show that this adjustment is the "optimal" finite-dimensional, linear adjustment. We additionally study two estimators with high-dimensional forms of covariate adjustment based on the LASSO. For each such estimator, we show that it leads to improvements in precision relative to the unadjusted difference-in-means estimator and also provide conditions under which it leads to the "optimal" nonparametric, covariate adjustment. A simulation study confirms the practical relevance of our theoretical analysis, and the methods are employed to reanalyze data from an experiment using a "matched pairs" design to study the effect of macroinsurance on microenterprise.
در این مقاله استنباط از میانگین اثر درمانی در آزمایشات بررسی شده است که در آن وضعیت درمانی با توجه به "جفت های همسان" تعیین می شود و این است علاوه بر این ، برای به دست آوردن متغیرهای پایه ، متغیرهای پایه را تنظیم می کند دقت بیشترمنظور ما از طرح "جفت های همسان" است. I.I.D.از جمعیت مورد علاقه ، با توجه به مشاهده ، جفت شده است متغیرهای متغیر و در آخر ، در هر جفت ، یک واحد به طور تصادفی انتخاب می شود رفتار.مهمتر از همه ، ما فرض می کنیم که همه مشاهده نشده ، متغیرهای پایه در تعیین تکلیف درمانی استفاده می شود.ما یک کلاس گسترده از برآوردگرها بر اساس یک لحظه "مضاعف قوی" که به ما امکان مطالعه می دهد برآوردگرها با هر دو شکل محدود و ابعاد بالا متغیرهای متغیر تنظیمما می دانیم که برآوردگرها با تنظیمات خطی و خطی محدود نیازی به پیشرفت در دقت نسبت به افراد غیر قابل تنظیم نیست برآوردگر اختلاف در میانگین.این پدیده حتی اگر تنظیمات ادامه یابد با درمان در تعامل هستند.در حقیقت ، انجام این کار منجر به هیچ تغییری در دقت، درستی.با این حال ، با استفاده از ثابت می توان از دستاوردهای دقیق اطمینان حاصل کرد اثرات برای هر یک از جفت ها.در واقع ، ما نشان می دهیم که این تنظیمات است تنظیم خطی "بهینه" محدود ، تنظیم خطی.ما علاوه بر این دو مطالعه می کنیم برآوردگرهای با اشکال بالایی از تنظیم متغیر متغیر بر اساس کمند.برای هر چنین برآوردگر ، ما نشان می دهیم که منجر به پیشرفت در دقت نسبت به برآوردگر اختلاف غیر قابل تنظیم و همچنین شرایطی را فراهم کنید که تحت آن به غیرپارامتری "بهینه" منجر شود ، تنظیم متغیر.یک مطالعه شبیه سازی ، ارتباط عملی را تأیید می کند تجزیه و تحلیل نظری ما ، و روش ها برای تجزیه و تحلیل داده ها از آزمایشی با استفاده از یک طرح "جفت همسان" برای بررسی تأثیر ماکرو بیمه در ریزگردها.
29,714
This paper presents a new perspective on the identification at infinity for the intercept of the sample selection model as identification at the boundary via a transformation of the selection index. This perspective suggests generalizations of estimation at infinity to kernel regression estimation at the boundary and further to local linear estimation at the boundary. The proposed kernel-type estimators with an estimated transformation are proven to be nonparametric-rate consistent and asymptotically normal under mild regularity conditions. A fully data-driven method of selecting the optimal bandwidths for the estimators is developed. The Monte Carlo simulation shows the desirable finite sample properties of the proposed estimators and bandwidth selection procedures.
در این مقاله چشم انداز جدیدی در مورد شناسایی در Infinity ارائه شده است رهگیری مدل انتخاب نمونه به عنوان شناسایی در مرز از طریق تغییر شاخص انتخاب.این دیدگاه نشان می دهد تعمیم تخمین در بی نهایت برآورد رگرسیون هسته در مرز و تخمین خطی محلی در مرز.در برآوردگرهای نوع هسته پیشنهادی با تحول تخمین زده شده اثبات شده است با سرعت غیر پارامتری و بدون علامت طبیعی تحت خفیف قرار بگیرید شرایط منظمیک روش کاملاً داده محور برای انتخاب بهینه پهنای باند برای برآوردگرها توسعه یافته است.شبیه سازی مونت کارلو را نشان می دهد خواص نمونه محدود مطلوب از برآوردگرهای پیشنهادی و پهنای باند روشهای انتخاب
29,715
We propose an econometric environment for structural break detection in nonstationary quantile predictive regressions. We establish the limit distributions for a class of Wald and fluctuation type statistics based on both the ordinary least squares estimator and the endogenous instrumental regression estimator proposed by Phillips and Magdalinos (2009a, Econometric Inference in the Vicinity of Unity. Working paper, Singapore Management University). Although the asymptotic distribution of these test statistics appears to depend on the chosen estimator, the IVX based tests are shown to be asymptotically nuisance parameter-free regardless of the degree of persistence and consistent under local alternatives. The finite-sample performance of both tests is evaluated via simulation experiments. An empirical application to house pricing index returns demonstrates the practicality of the proposed break tests for regression quantiles of nonstationary time series data.
ما یک محیط اقتصاد سنجی برای تشخیص شکستگی ساختاری در پیشنهاد می کنیم رگرسیون پیش بینی کننده کمی غیر ایستا.ما حد را تعیین می کنیم توزیع برای یک کلاس از آمار والد و نوسانات بر اساس هر دو برآوردگر حداقل مربعات معمولی و رگرسیون ابزاری درون زا برآوردگر پیشنهاد شده توسط فیلیپس و مگدالینوس (2009a ، استنباط اقتصاد سنجی در مجاورت وحدت.مقاله کار ، دانشگاه مدیریت سنگاپور). اگرچه به نظر می رسد توزیع بدون علامت این آمار آزمون بستگی دارد در برآوردگر انتخاب شده ، تست های مبتنی بر IVX به صورت بدون علامت نشان داده شده است بدون توجه به میزان پایداری و سازگار بدون پارامتر تحت گزینه های محلی.عملکرد نمونه محدود هر دو آزمایش است از طریق آزمایش های شبیه سازی ارزیابی می شود.یک برنامه تجربی برای قیمت گذاری خانه بازده شاخص نشان دهنده عملی تست های استراحت پیشنهادی برای مقادیر رگرسیون داده های سری زمانی غیر ایستگاه.
29,716
Machine learning (ML) estimates of conditional average treatment effects (CATE) can guide policy decisions, either by allowing targeting of individuals with beneficial CATE estimates, or as inputs to decision trees that optimise overall outcomes. There is limited information available regarding how well these algorithms perform in real-world policy evaluation scenarios. Using synthetic data, we compare the finite sample performance of different policy learning algorithms, machine learning techniques employed during their learning phases, and methods for presenting estimated policy values. For each algorithm, we assess the resulting treatment allocation by measuring deviation from the ideal ("oracle") policy. Our main finding is that policy trees based on estimated CATEs outperform trees learned from doubly-robust scores. Across settings, Causal Forests and the Normalised Double-Robust Learner perform consistently well, while Bayesian Additive Regression Trees perform poorly. These methods are then applied to a case study targeting optimal allocation of subsidised health insurance, with the goal of reducing infant mortality in Indonesia.
یادگیری ماشین (ML) تخمین اثرات درمانی متوسط ​​مشروط (CATE) می تواند تصمیمات مربوط به سیاست را راهنمایی کند ، یا با اجازه هدف قرار دادن افراد با برآوردهای مفید CATE ، یا به عنوان ورودی برای تصمیم گیری در مورد درختان که بهینه می شوند نتایج کلیاطلاعات محدودی در مورد میزان خوب وجود دارد این الگوریتم ها در سناریوهای ارزیابی سیاست در دنیای واقعی عمل می کنند.استفاده كردن داده های مصنوعی ، ما عملکرد نمونه محدود سیاست های مختلف را مقایسه می کنیم الگوریتم های یادگیری ، تکنیک های یادگیری ماشین که در طول یادگیری خود به کار می روند مراحل و روشهای ارائه مقادیر تخمین زده شده سیاست.برای هر الگوریتم ، ما تخصیص درمان حاصل را با اندازه گیری انحراف از سیاست ایده آل ("اوراکل").یافته اصلی ما این است که درختان سیاست بر اساس تخمین زده شده از درختان بهتر از درختان آموخته شده از نمرات مضاعف.در سراسر تنظیمات ، جنگل های علی و یادگیرنده دو طرفه عادی انجام می شود به طور مداوم خوب ، در حالی که درختان رگرسیون افزودنی بیزی عملکرد ضعیفی دارند. این روشها سپس برای یک مطالعه موردی که اختصاصی بهینه از آن را هدف قرار می دهد ، اعمال می شود بیمه درمانی یارانه ای ، با هدف کاهش مرگ و میر نوزادان در اندونزی
29,717
Designing individualized allocation of treatments so as to maximize the equilibrium welfare of interacting agents has many policy-relevant applications. Focusing on sequential decision games of interacting agents, this paper develops a method to obtain optimal treatment assignment rules that maximize a social welfare criterion by evaluating stationary distributions of outcomes. Stationary distributions in sequential decision games are given by Gibbs distributions, which are difficult to optimize with respect to a treatment allocation due to analytical and computational complexity. We apply a variational approximation to the stationary distribution and optimize the approximated equilibrium welfare with respect to treatment allocation using a greedy optimization algorithm. We characterize the performance of the variational approximation, deriving a performance guarantee for the greedy optimization algorithm via a welfare regret bound. We implement our proposed method in simulation exercises and an empirical application using the Indian microfinance data (Banerjee et al., 2013), and show it delivers significant welfare gains.
طراحی تخصیص فردی از درمان ها به حداکثر رساندن رفاه تعادل عوامل در تعامل دارای بسیاری از سیاست های مربوط به سیاست است برنامه های کاربردی.با تمرکز بر بازی های تصمیم گیری متوالی عوامل تعامل ، این مقاله روشی را برای به دست آوردن قوانین بهینه واگذاری درمان ایجاد می کند که با ارزیابی توزیع های ثابت از معیار رفاه اجتماعی حداکثر عواقب.توزیع ثابت در بازی های تصمیم گیری متوالی توسط توزیع گیبس ، که بهینه سازی آنها با توجه به a دشوار است تخصیص درمان به دلیل پیچیدگی تحلیلی و محاسباتی.ما اعمال می کنیم تقریب متغیر به توزیع ثابت و بهینه سازی رفاه تعادل تقریبی با توجه به تخصیص درمان با استفاده از a الگوریتم بهینه سازی حریص.ما عملکرد عملکرد را توصیف می کنیم تقریب متغیر ، به دست آوردن ضمانت عملکرد برای حریص الگوریتم بهینه سازی از طریق یک رفاه رفاه محدود.ما پیشنهادی خود را پیاده سازی می کنیم روش در تمرینات شبیه سازی و یک برنامه تجربی با استفاده از هندی داده های مالی خرد (بانرجی و همکاران ، 2013) ، و نشان می دهد که قابل توجه است سود رفاه.
29,718
We provide a simple method to estimate the parameters of multivariate stochastic volatility models with latent factor structures. These models are very useful as they alleviate the standard curse of dimensionality, allowing the number of parameters to increase only linearly with the number of the return series. Although theoretically very appealing, these models have only found limited practical application due to huge computational burdens. Our estimation method is simple in implementation as it consists of two steps: first, we estimate the loadings and the unconditional variances by maximum likelihood, and then we use the efficient method of moments to estimate the parameters of the stochastic volatility structure with GARCH as an auxiliary model. In a comprehensive Monte Carlo study we show the good performance of our method to estimate the parameters of interest accurately. The simulation study and an application to real vectors of daily returns of dimensions up to 148 show the method's computation advantage over the existing estimation procedures.
ما یک روش ساده برای برآورد پارامترهای چند متغیره ارائه می دهیم مدل های نوسانات تصادفی با ساختارهای فاکتور نهفته.این مدل ها هستند بسیار مفید است زیرا آنها لعنت استاندارد از ابعاد را کاهش می دهند ، اجازه می دهند تعداد پارامترها برای افزایش خطی فقط با تعداد تعداد سریال را برگردانید.اگرچه از نظر تئوری بسیار جذاب است ، این مدل ها فقط دارند به دلیل بارهای محاسباتی عظیم ، کاربرد عملی محدودی پیدا کرد.ما روش تخمین در اجرای ساده است زیرا از دو مرحله تشکیل شده است: ابتدا بارهای و واریانس های بی قید و شرط را حداکثر تخمین می زنیم احتمال ، و سپس ما از روش کارآمد لحظات برای تخمین استفاده می کنیم پارامترهای ساختار نوسانات تصادفی با GARCH به عنوان یک کمکی مدل.در یک مطالعه جامع مونت کارلو عملکرد خوب ما را نشان می دهیم روش برای برآورد پارامترهای مورد نظر به طور دقیق.مطالعه شبیه سازی و برنامه ای برای بردارهای واقعی بازده روزانه ابعاد تا 148 مزیت محاسبه روش را نسبت به تخمین موجود نشان دهید رویه ها
29,719
This paper presents an inference method for the local average treatment effect (LATE) in the presence of high-dimensional covariates, irrespective of the strength of identification. We propose a novel high-dimensional conditional test statistic with uniformly correct asymptotic size. We provide an easy-to-implement algorithm to infer the high-dimensional LATE by inverting our test statistic and employing the double/debiased machine learning method. Simulations indicate that our test is robust against both weak identification and high dimensionality concerning size control and power performance, outperforming other conventional tests. Applying the proposed method to railroad and population data to study the effect of railroad access on urban population growth, we observe that our methodology yields confidence intervals that are 49% to 92% shorter than conventional results, depending on specifications.
در این مقاله یک روش استنباط برای درمان متوسط ​​محلی ارائه شده است اثر (دیر) در حضور متغیرهای متغیر با ابعاد بالا ، صرف نظر از قدرت شناسایی.ما یک شرطی با ابعاد بالا رمان را پیشنهاد می کنیم آمار آزمون را با اندازه یکنواخت بدون علامت درست کنید.ما ارائه می دهیم الگوریتم آسان برای اجرای برای استنباط اواخر با ابعاد بالا با وارونه کردن ما آمار آزمون و استفاده از روش یادگیری ماشین مضاعف/debiased. شبیه سازی ها نشان می دهد که آزمایش ما در برابر هر دو شناسایی ضعیف قوی است و ابعاد بالا در مورد کنترل اندازه و عملکرد قدرت ، بهتر از سایر آزمایشات معمولی.استفاده از روش پیشنهادی در داده های راه آهن و جمعیت برای بررسی تأثیر دسترسی راه آهن بر شهری رشد جمعیت ، ما مشاهده می کنیم که روش ما فواصل اطمینان را به همراه دارد بسته به نوع 49 ٪ تا 92 ٪ کوتاهتر از نتایج معمولی است مشخصات فنی.
29,720
This paper studies settings where the analyst is interested in identifying and estimating the average causal effect of a binary treatment on an outcome. We consider a setup in which the outcome realization does not get immediately realized after the treatment assignment, a feature that is ubiquitous in empirical settings. The period between the treatment and the realization of the outcome allows other observed actions to occur and affect the outcome. In this context, we study several regression-based estimands routinely used in empirical work to capture the average treatment effect and shed light on interpreting them in terms of ceteris paribus effects, indirect causal effects, and selection terms. We obtain three main and related takeaways. First, the three most popular estimands do not generally satisfy what we call \emph{strong sign preservation}, in the sense that these estimands may be negative even when the treatment positively affects the outcome conditional on any possible combination of other actions. Second, the most popular regression that includes the other actions as controls satisfies strong sign preservation \emph{if and only if} these actions are mutually exclusive binary variables. Finally, we show that a linear regression that fully stratifies the other actions leads to estimands that satisfy strong sign preservation.
در این مقاله تنظیماتی که تحلیلگر علاقه مند به شناسایی است ، مطالعه می کند و تخمین میانگین اثر علیت یک درمان باینری بر روی نتیجه. ما یک مجموعه را در نظر می گیریم که در آن تحقق نتیجه بلافاصله حاصل نمی شود پس از انجام معالجه درمان ، ویژگی ای که در همه جا وجود دارد تنظیمات تجربی.دوره بین درمان و تحقق نتیجه اجازه می دهد تا سایر اقدامات مشاهده شده اتفاق بیفتد و بر نتیجه تأثیر بگذارد.در این زمینه ، ما چندین برآورد مبتنی بر رگرسیون را که بطور معمول در آن استفاده می شود مطالعه می کنیم کار تجربی برای گرفتن میانگین اثر درمانی و روشن کردن نور تفسیر آنها از نظر اثرات paribus ceteris ، اثرات علیت غیرمستقیم ، و شرایط انتخابما سه غذای اصلی و مرتبط را به دست می آوریم.اول ، سه برآورد محبوب به طور کلی آنچه را که ما می نامیم \ itm {قوی نیست حفظ نشانه} ، به این معنا که این برآوردها حتی در مواردی منفی باشند این درمان تأثیر مثبت بر هرگونه مشروط بر هرگونه ممکن است ترکیبی از اقدامات دیگر.دوم ، محبوب ترین رگرسیون که شامل می شود اقدامات دیگر به عنوان کنترل ، حفظ علائم قوی را برآورده می کند. فقط در صورتی که این اقدامات متغیرهای باینری متقابل منحصر به فرد هستند.بالاخره ، ما نشان دهید که یک رگرسیون خطی که به طور کامل اقدامات دیگر را طبقه بندی می کند منجر به برآوردهایی که حفظ علائم قوی را برآورده می کنند.
29,721
Different proxy variables used in fiscal policy SVARs lead to contradicting conclusions regarding the size of fiscal multipliers. In this paper, we show that the conflicting results are due to violations of the exogeneity assumptions, i.e. the commonly used proxies are endogenously related to the structural shocks. We propose a novel approach to include proxy variables into a Bayesian non-Gaussian SVAR, tailored to accommodate potentially endogenous proxy variables. Using our model, we show that increasing government spending is a more effective tool to stimulate the economy than reducing taxes. We construct new exogenous proxies that can be used in the traditional proxy VAR approach resulting in similar estimates compared to our proposed hybrid SVAR model.
متغیرهای مختلف پروکسی مورد استفاده در SVAR های سیاست مالی منجر به متناقض می شوند نتیجه گیری در مورد اندازه ضربهای مالی.در این مقاله ، ما نشان می دهیم اینکه نتایج متناقض به دلیل نقض اگزوژن است فرضیات ، یعنی پروکسی های متداول که به صورت درون زا در ارتباط هستند با شوک های ساختاری.ما یک رویکرد جدید را پیشنهاد می کنیم تا متغیرهای پروکسی را در آن قرار دهیم یک SVAR غیر گاوسی بیزی ، متناسب با قرار دادن بالقوه درون زا متغیرهای پروکسی.با استفاده از مدل خود ، ما نشان می دهیم که افزایش هزینه های دولت ابزاری مؤثرتر برای تحریک اقتصاد از کاهش مالیات است.ما پروکسی های جدید اگزوژن را بسازید که می تواند در پروکسی سنتی استفاده شود رویکرد در نتیجه تخمین های مشابه در مقایسه با SVAR ترکیبی پیشنهادی ما مدل.
29,722
Local Projection is widely used for impulse response estimation, with the Fixed Effect (FE) estimator being the default for panel data. This paper highlights the presence of Nickell bias for all regressors in the FE estimator, even if lagged dependent variables are absent in the regression. This bias is the consequence of the inherent panel predictive specification. We recommend using the split-panel jackknife estimator to eliminate the asymptotic bias and restore the standard statistical inference. Revisiting three macro-finance studies on the linkage between financial crises and economic contraction, we find that the FE estimator substantially underestimates the post-crisis economic losses.
طرح ریزی محلی به طور گسترده ای برای برآورد پاسخ تکانه استفاده می شود. برآوردگر اثر ثابت (FE) پیش فرض برای داده های پانل است.این کاغذ وجود تعصب نیکل را برای همه رگرسیونرها در برآوردگر FE برجسته می کند ، حتی اگر متغیرهای وابسته به عقب در رگرسیون وجود نداشته باشند.این تعصب است نتیجه مشخصات پیش بینی پانل ذاتی.توصیه می کنیم با استفاده از برآوردگر Jackknife تقسیم شده برای از بین بردن تعصب بدون علامت و استنتاج آماری استاندارد را بازیابی کنید.تجدید نظر در سه کلان مالی مطالعات در مورد پیوند بین بحران های مالی و انقباض اقتصادی ، ما دریابید که برآوردگر آهن به طور قابل ملاحظه ای پس از بحران را دست کم می گیرد خسارات اقتصادی
29,723
Experiments are an important tool to measure the impacts of interventions. However, in experimental settings with one-sided noncompliance, extant empirical approaches may not produce the estimands a decision-maker needs to solve their problem. For example, these experimental designs are common in digital advertising settings, but typical methods do not yield effects that inform the intensive margin -- how much should be spent or how many consumers should be reached with a campaign. We propose a solution that combines a novel multi-cell experimental design with modern estimation techniques that enables decision-makers to recover enough information to solve problems with an intensive margin. Our design is straightforward to implement. Using data from advertising experiments at Facebook, we demonstrate our approach outperforms standard techniques in recovering treatment effect parameters. Through a simple advertising reach decision problem, we show that our approach generates better decisions relative to standard techniques.
آزمایشات یک ابزار مهم برای اندازه گیری تأثیر مداخلات است. با این حال ، در تنظیمات آزمایشی با عدم سازگاری یک طرفه ، موجود است رویکردهای تجربی ممکن است تخمین هایی را که تصمیم گیرنده لازم است تولید نکنند مشکل آنها را حل کنید.به عنوان مثال ، این طرح های آزمایشی در تنظیمات تبلیغات دیجیتال ، اما روشهای معمولی اثرات آن را به همراه نمی آورند حاشیه فشرده را آگاه کنید - چقدر باید هزینه شود یا چه تعداد مصرف کننده باید با یک کمپین به دست بیاید.ما راه حلی را پیشنهاد می کنیم که یک رمان را ترکیب می کند طراحی آزمایشی چند سلولی با تکنیک های تخمین مدرن که امکان پذیر است تصمیم گیرندگان برای بازیابی اطلاعات کافی برای حل مشکلات با یک حاشیه فشردهطراحی ما ساده برای پیاده سازی است.استفاده از داده ها آزمایش های تبلیغاتی در فیس بوک ، ما رویکرد خود را بهتر نشان می دهیم تکنیک های استاندارد در بازیابی پارامترهای اثر درمانی.از طریق یک ساده تبلیغات به مشکل تصمیم گیری می رسد ، ما نشان می دهیم که رویکرد ما بهتر تولید می کند تصمیمات نسبت به تکنیک های استاندارد.
29,724
We examine the incremental value of news-based data relative to the FRED-MD economic indicators for quantile predictions (now- and forecasts) of employment, output, inflation and consumer sentiment. Our results suggest that news data contain valuable information not captured by economic indicators, particularly for left-tail forecasts. Methods that capture quantile-specific non-linearities produce superior forecasts relative to methods that feature linear predictive relationships. However, adding news-based data substantially increases the performance of quantile-specific linear models, especially in the left tail. Variable importance analyses reveal that left tail predictions are determined by both economic and textual indicators, with the latter having the most pronounced impact on consumer sentiment.
ما ارزش افزایشی داده های مبتنی بر اخبار را نسبت به فرد MD بررسی می کنیم شاخص های اقتصادی برای پیش بینی های کمی (اکنون و پیش بینی) اشتغال ، خروجی ، تورم و احساسات مصرف کننده.نتایج ما نشان می دهد که داده های خبری حاوی اطلاعات ارزشمندی است که توسط شاخص های اقتصادی ضبط نشده است ، به ویژه برای پیش بینی های دم چپ.روش هایی که خاصیت کمی را ضبط می کنند غیر خطی ها پیش بینی های برتر را نسبت به روشهایی که از آن استفاده می کنند تولید می کنند روابط پیش بینی خطی.با این حال ، اضافه کردن داده های مبتنی بر اخبار قابل ملاحظه ای عملکرد مدلهای خطی خاص کمی را افزایش می دهد ، به خصوص در دم چپتجزیه و تحلیل اهمیت متغیر نشان می دهد که پیش بینی دم چپ هستند تعیین شده توسط شاخص های اقتصادی و متنی ، با دومی بیشترین تأثیر بر احساسات مصرف کننده.
29,725
This article discusses the use of dynamic factor models in macroeconomic forecasting, with a focus on the Factor-Augmented Error Correction Model (FECM). The FECM combines the advantages of cointegration and dynamic factor models, providing a flexible and reliable approach to macroeconomic forecasting, especially for non-stationary variables. We evaluate the forecasting performance of the FECM model on a large dataset of 117 Moroccan economic series with quarterly frequency. Our study shows that FECM outperforms traditional econometric models in terms of forecasting accuracy and robustness. The inclusion of long-term information and common factors in FECM enhances its ability to capture economic dynamics and leads to better forecasting performance than other competing models. Our results suggest that FECM can be a valuable tool for macroeconomic forecasting in Morocco and other similar economies.
این مقاله در مورد استفاده از مدلهای فاکتور پویا در اقتصاد کلان بحث شده است پیش بینی ، با تمرکز بر روی مدل تصحیح خطای فاکتور. (FECM).FECM مزایای ادغام و فاکتور پویا را ترکیب می کند مدل ها ، ارائه یک رویکرد انعطاف پذیر و قابل اعتماد برای اقتصاد کلان پیش بینی ، به ویژه برای متغیرهای غیر ثابت.ما ارزیابی می کنیم پیش بینی عملکرد مدل FECM در یک مجموعه داده بزرگ 117 مراکش سری اقتصادی با فرکانس سه ماهه.مطالعه ما نشان می دهد که FECM بهتر است مدلهای سنتی اقتصاد سنجی از نظر پیش بینی دقت و استحکام. گنجاندن اطلاعات طولانی مدت و عوامل مشترک در FECM آن را تقویت می کند توانایی ضبط پویایی اقتصادی و منجر به پیش بینی بهتر عملکرد نسبت به سایر مدل های رقیب.نتایج ما نشان می دهد که FECM می تواند یک باشد ابزار ارزشمند برای پیش بینی کلان اقتصادی در مراکش و سایر موارد مشابه اقتصاد
29,726
This paper proposes a linear categorical random coefficient model, in which the random coefficients follow parametric categorical distributions. The distributional parameters are identified based on a linear recurrence structure of moments of the random coefficients. A Generalized Method of Moments estimation procedure is proposed also employed by Peter Schmidt and his coauthors to address heterogeneity in time effects in panel data models. Using Monte Carlo simulations, we find that moments of the random coefficients can be estimated reasonably accurately, but large samples are required for estimation of the parameters of the underlying categorical distribution. The utility of the proposed estimator is illustrated by estimating the distribution of returns to education in the U.S. by gender and educational levels. We find that rising heterogeneity between educational groups is mainly due to the increasing returns to education for those with postsecondary education, whereas within group heterogeneity has been rising mostly in the case of individuals with high school or less education.
در این مقاله یک مدل ضریب تصادفی طبقه بندی شده خطی ارائه شده است ، که در آن ضرایب تصادفی از توزیع های طبقه بندی پارامتری پیروی می کنند.در پارامترهای توزیع بر اساس یک ساختار عود خطی مشخص می شوند از لحظات ضرایب تصادفی.یک روش کلی از لحظات روش تخمین نیز توسط پیتر اشمیت و وی به کار گرفته شده است همکاران برای پرداختن به ناهمگونی در اثرات زمانی در مدل های داده پانل.استفاده كردن شبیه سازی مونت کارلو ، می فهمیم که لحظه های ضرایب تصادفی می تواند باشد به طور منطقی تخمین زده می شود ، اما برای برآورد نمونه های بزرگ مورد نیاز است از پارامترهای توزیع طبقه بندی اساسی.ابزار برآوردگر پیشنهادی با برآورد توزیع بازده نشان داده شده است به آموزش در ایالات متحده بر اساس سطح جنسیت و تحصیلات.ما می بینیم که در حال افزایش است ناهمگونی بین گروههای آموزشی عمدتاً ناشی از افزایش است برای کسانی که دارای آموزش بعد از دوره متوسطه هستند ، به آموزش و پرورش باز می گردد ، در حالی که در ناهمگونی گروه بیشتر در مورد افراد دارای زیاد در حال افزایش است مدرسه یا آموزش کمتر.
29,727
This paper revisits the Lagrange multiplier type test for the null hypothesis of no cross-sectional dependence in large panel data models. We propose a unified test procedure and its power enhancement version, which show robustness for a wide class of panel model contexts. Specifically, the two procedures are applicable to both heterogeneous and fixed effects panel data models with the presence of weakly exogenous as well as lagged dependent regressors, allowing for a general form of nonnormal error distribution. With the tools from Random Matrix Theory, the asymptotic validity of the test procedures is established under the simultaneous limit scheme where the number of time periods and the number of cross-sectional units go to infinity proportionally. The derived theories are accompanied by detailed Monte Carlo experiments, which confirm the robustness of the two tests and also suggest the validity of the power enhancement technique.
در این مقاله آزمون نوع ضرب LaGrange برای فرضیه تهی تجدید نظر شده است هیچ وابستگی مقطعی در مدلهای داده پانل بزرگ وجود ندارد.ما پیشنهاد می کنیم روش آزمایش یکپارچه و نسخه تقویت قدرت آن ، که نشان دهنده استحکام است برای یک کلاس گسترده از زمینه های مدل پانل.به طور خاص ، این دو روش است قابل استفاده برای هر دو مدل داده پانل جلوه های ناهمگن و ثابت با وجود رگرسیون های وابسته به اگزوژن و همچنین وابسته به عقب مانده ، اجازه می دهد برای یک شکل کلی از توزیع خطای غیر طبیعی.با ابزارهای تصادفی نظریه ماتریس ، اعتبار بدون علامت روشهای آزمون ایجاد شده است تحت طرح حد همزمان که تعداد دوره های زمانی و تعداد واحدهای مقطعی به طور متناسب به بی نهایت می روند.مشتق شده نظریه ها با آزمایش های دقیق مونت کارلو همراه هستند که تأیید می کنند استحکام دو آزمایش و همچنین اعتبار قدرت را نشان می دهد تکنیک تقویت.
29,728
This paper develops a new specification test for the instrument weakness when the number of instruments $K_n$ is large with a magnitude comparable to the sample size $n$. The test relies on the fact that the difference between the two-stage least squares (2SLS) estimator and the ordinary least squares (OLS) estimator asymptotically disappears when there are many weak instruments, but otherwise converges to a non-zero limit. We establish the limiting distribution of the difference within the above two specifications, and introduce a delete-$d$ Jackknife procedure to consistently estimate the asymptotic variance/covariance of the difference. Monte Carlo experiments demonstrate the good performance of the test procedure for both cases of single and multiple endogenous variables. Additionally, we re-examine the analysis of returns to education data in Angrist and Keueger (1991) using our proposed test. Both the simulation results and empirical analysis indicate the reliability of the test.
در این مقاله یک تست مشخصات جدید برای ضعف ابزار ایجاد می شود تعداد سازهای $ k_n $ با بزرگی قابل مقایسه با اندازه نمونه $ n $.آزمون به این واقعیت متکی است که تفاوت بین حداقل مربعات حداقل مربع (2SL) و حداقل مربعات معمولی (OLS) برآوردگر به صورت مجانبی وقتی سازهای ضعیف زیادی وجود دارد ، اما اما در غیر این صورت به حد غیر صفر همگرا می شود.ما توزیع محدود را تعیین می کنیم از تفاوت در دو مشخصات فوق ، و معرفی حذف- $ d $ jackknife برای برآورد مداوم بدون علامت واریانس/کواریانس تفاوت.آزمایش های مونت کارلو نشان می دهد عملکرد خوب روش آزمایش برای هر دو مورد تک و چندگانه متغیرهای درون زا.علاوه بر این ، ما دوباره تجزیه و تحلیل بازده ها را بررسی می کنیم داده های آموزش در Angrist و Keueger (1991) با استفاده از آزمون پیشنهادی ما.هر دو نتایج شبیه سازی و تجزیه و تحلیل تجربی نشانگر قابلیت اطمینان آزمون است.
29,729
This paper investigates the behavior of Stock and Yogo (2005)'s first-stage F statistic and the Cragg-Donald statistic (Cragg and Donald, 1993) when the number of instruments and the sample size go to infinity in a comparable magnitude. Our theory shows that the first-stage F test is oversized for detecting many weak instruments. We next propose an asymptotically valid correction of the F statistic for testing weakness of instruments. The theory is also used to construct confidence intervals for the strength of instruments. As for the Cragg-Donald statistic, we obtain an asymptotically valid correction in the case of two endogenous variables. Monte Carlo experiments demonstrate the satisfactory performance of the proposed methods in both situations of a single and multiple endogenous variables. The usefulness of the proposed tests is illustrated by an analysis of the returns to education data in Angrist and Keueger (1991).
در این مقاله به بررسی رفتار سهام و یوگو (2005) در مرحله اول f می پردازیم آمار و آمار Cragg-Donald (کراگ و دونالد ، 1993) وقتی تعداد ابزارها و اندازه نمونه در یک مقایسه به بی نهایت می روند اندازه.نظریه ما نشان می دهد که آزمون مرحله اول F برای تشخیص بسیاری از ابزارهای ضعیف.ما در مرحله بعد یک معتبر بدون علامت پیشنهاد می کنیم تصحیح آمار F برای آزمایش ضعف ابزارها.نظریه همچنین برای ایجاد فواصل اطمینان برای استحکام ابزارها استفاده می شود. در مورد آمار Cragg-Donald ، ما یک اصلاح معتبر بدون علامت به دست می آوریم در مورد دو متغیر درون زا.آزمایش های مونت کارلو نشان می دهد عملکرد رضایت بخش روشهای پیشنهادی در هر دو حالت a متغیرهای درون زا تک و چندگانه.سودمندی تست های پیشنهادی با تجزیه و تحلیل بازده داده های آموزش در Angrist و Keueger (1991).
31,608
This study examines the influence of grandchildren's gender on grandparents' voting behavior using independently collected individual-level data. The survey was conducted immediately after the House of Councilors election in Japan. I observed that individuals with a granddaughter were more likely to vote for female candidates by around 10 % than those without. However, having a daughter did not affect the parents' voting behavior. Furthermore, having a son or a grandson did not influence grandparents' voting behavior. This implies that grandparents voted for their granddaughter's future benefit because granddaughters may be too young vote in a male-dominated and aging society.
این مطالعه به بررسی تأثیر جنسیت نوه ها بر پدربزرگ و مادربزرگ و مادربزرگ می پردازد رفتار رأی گیری با استفاده از داده های مستقل جمع آوری شده در سطح فردی.بررسی بلافاصله پس از انتخابات مجلس شورا در ژاپن انجام شد.من مشاهده کرد که افراد مبتلا به نوه احتمالاً به آن رأی می دهند نامزدهای زن حدود 10 ٪ از افراد بدون آن.با این حال ، داشتن یک دختر بر رفتار رای گیری والدین تأثیر نمی گذارد.علاوه بر این ، داشتن یک پسر یا یک نوه در رفتار رای گیری مادربزرگ و مادربزرگ تأثیر نگذاشت.این به معنای این است که پدربزرگ و مادربزرگ به سود آینده نوه خود رای دادند زیرا نوه ها ممکن است در یک جامعه تحت سلطه مردان و پیر شدن بسیار جوان باشند.
29,730
Dynamic logit models are popular tools in economics to measure state dependence. This paper introduces a new method to derive moment restrictions in a large class of such models with strictly exogenous regressors and fixed effects. We exploit the common structure of logit-type transition probabilities and elementary properties of rational fractions, to formulate a systematic procedure that scales naturally with model complexity (e.g the lag order or the number of observed time periods). We detail the construction of moment restrictions in binary response models of arbitrary lag order as well as first-order panel vector autoregressions and dynamic multinomial logit models. Identification of common parameters and average marginal effects is also discussed for the binary response case. Finally, we illustrate our results by studying the dynamics of drug consumption amongst young people inspired by Deza (2015).
مدل های ورود به سیستم پویا ابزارهای محبوب در اقتصاد برای اندازه گیری دولت هستند وابستگی.در این مقاله یک روش جدید برای به دست آوردن محدودیت های لحظه ای در کلاس بزرگی از چنین مدلهایی با رگرسیون های کاملاً بیرونی و ثابت اثراتما از ساختار مشترک احتمالات انتقال از نوع logit سوءاستفاده می کنیم و خصوصیات ابتدایی کسری منطقی ، برای تدوین یک سیستماتیک روشی که به طور طبیعی با پیچیدگی مدل مقیاس می یابد (به عنوان مثال ترتیب تاخیر یا تعداد دوره های زمانی مشاهده شده).ما جزئیات ساخت لحظه را شرح می دهیم محدودیت در مدل های پاسخ باینری از نظم تاخیر خودسرانه و همچنین اتورهای وکتور پانل مرتبه اول و مدلهای ورود به سیستم چندمجمی پویا. شناسایی پارامترهای مشترک و میانگین اثرات حاشیه ای نیز هست مورد بحث برای پرونده پاسخ باینری.سرانجام ، ما نتایج خود را نشان می دهیم مطالعه پویایی مصرف مواد مخدر در بین جوانان با الهام از دزا (2015).
29,731
This paper studies averages of intersection bounds -- the bounds defined by the infimum of a collection of regression functions -- and other similar functionals of these bounds, such as averages of saddle values. Examples of such parameters are Frechet-Hoeffding bounds, Makarov (1981) bounds on distributional effects. The proposed estimator classifies covariate values into the regions corresponding to the identity of the binding regression function and takes the sample average. The paper shows that the proposed moment function is insensitive to first-order classification mistakes, enabling various nonparametric and regularized/machine learning classifiers in the first (classification) step. The result is generalized to cover bounds on the values of linear programming problems and best linear predictor of intersection bounds.
این مقاله به طور متوسط ​​از مرزهای تقاطع - مرزهای تعریف شده توسط اطلاعات مجموعه ای از توابع رگرسیون - و موارد مشابه عملکرد این مرزها ، مانند میانگین مقادیر زین.نمونه هایی از این پارامترها مرزهای Frechet-Hoeffding ، Makarov (1981) در محدود هستند اثرات توزیع.برآوردگر پیشنهادی مقادیر متغیر را به مناطق مربوط به هویت عملکرد رگرسیون اتصال و میانگین نمونه را می گیرد.مقاله نشان می دهد که عملکرد لحظه پیشنهادی نسبت به اشتباهات طبقه بندی مرتبه اول بی حس است ، طبقه بندی کننده های غیرپارامتری و منظم و منظم/ماشین در مرحله اول (طبقه بندی) مرحله.نتیجه برای پوشاندن مرزها بر روی مقادیر تعمیم یافته است مشکلات برنامه نویسی خطی و بهترین پیش بینی کننده خطی تقاطع مرزها
29,732
Monthly and weekly economic indicators are often taken to be the largest common factor estimated from high and low frequency data, either separately or jointly. To incorporate mixed frequency information without directly modeling them, we target a low frequency diffusion index that is already available, and treat high frequency values as missing. We impute these values using multiple factors estimated from the high frequency data. In the empirical examples considered, static matrix completion that does not account for serial correlation in the idiosyncratic errors yields imprecise estimates of the missing values irrespective of how the factors are estimated. Single equation and systems-based dynamic procedures that account for serial correlation yield imputed values that are closer to the observed low frequency ones. This is the case in the counterfactual exercise that imputes the monthly values of consumer sentiment series before 1978 when the data was released only on a quarterly basis. This is also the case for a weekly version of the CFNAI index of economic activity that is imputed using seasonally unadjusted data. The imputed series reveals episodes of increased variability of weekly economic information that are masked by the monthly data, notably around the 2014-15 collapse in oil prices.
شاخص های اقتصادی ماهانه و هفتگی اغلب به عنوان بزرگترین مورد استفاده می شوند فاکتور مشترک تخمین زده شده از داده های فرکانس بالا و پایین ، به طور جداگانه یا به طور مشترکبرای ترکیب اطلاعات فرکانس مختلط بدون مدل سازی مستقیم آنها ، ما یک شاخص انتشار فرکانس پایین را که از قبل موجود است هدف قرار می دهیم ، و مقادیر فرکانس بالا را از دست ندهید.ما این مقادیر را با استفاده از چندین مورد استفاده می کنیم عوامل تخمین زده شده از داده های فرکانس بالا.در مثالهای تجربی در نظر گرفته شده ، تکمیل ماتریس استاتیک که سریال را به خود اختصاص نمی دهد همبستگی در خطاهای ایدیوسنکراتیک تخمین های نادرست از مقادیر از دست رفته صرف نظر از نحوه تخمین عوامل.معادله و رویه های پویا مبتنی بر سیستم که عملکرد همبستگی سریال را به خود اختصاص می دهد مقادیر محرکی که به کالاهای فرکانس پایین مشاهده شده نزدیکتر هستند.این است مورد در تمرین ضد خلاف واقع که مقادیر ماهانه مصرف کننده را تحمیل می کند سری احساسات قبل از سال 1978 هنگامی که داده ها فقط در یک فصلنامه منتشر شد پایهاین همچنین برای نسخه هفتگی شاخص CFNAI از فعالیت اقتصادی که با استفاده از داده های غیر قابل تنظیم فصلی انجام می شود.منتسب سریال قسمت هایی از افزایش تنوع اطلاعات اقتصادی هفتگی را نشان می دهد که توسط داده های ماهانه پوشانده شده است ، به ویژه در مورد فروپاشی 2014-15 در روغن قیمت.
29,733
This paper develops estimation and inference methods for censored quantile regression models with high-dimensional controls. The methods are based on the application of double/debiased machine learning (DML) framework to the censored quantile regression estimator of Buchinsky and Hahn (1998). I provide valid inference for low-dimensional parameters of interest in the presence of high-dimensional nuisance parameters when implementing machine learning estimators. The proposed estimator is shown to be consistent and asymptotically normal. The performance of the estimator with high-dimensional controls is illustrated with numerical simulation and an empirical application that examines the effect of 401(k) eligibility on savings.
این مقاله روشهای تخمین و استنباط برای کمی سانسور شده را تهیه می کند مدل های رگرسیون با کنترل های بعدی.روشها بر اساس استفاده از چارچوب یادگیری ماشین مضاعف/debiased (DML) برای سانسور برآوردگر رگرسیون کمی از بوچینسکی و هان (1998).من معتبر ارائه می دهم استنتاج برای پارامترهای کم بعدی مورد علاقه در حضور پارامترهای مزاحمت با ابعاد بالا هنگام اجرای یادگیری ماشین برآوردگرهابرآوردگر پیشنهادی نشان داده شده است که سازگار و بدون علامت است طبیعی.عملکرد برآوردگر با کنترل های با ابعاد بالا است با شبیه سازی عددی و یک کاربرد تجربی نشان داده شده است که تأثیر 401 (k) واجد شرایط بودن در پس انداز را بررسی می کند.
29,734
Despite increasing popularity in empirical studies, the integration of machine learning generated variables into regression models for statistical inference suffers from the measurement error problem, which can bias estimation and threaten the validity of inferences. In this paper, we develop a novel approach to alleviate associated estimation biases. Our proposed approach, EnsembleIV, creates valid and strong instrumental variables from weak learners in an ensemble model, and uses them to obtain consistent estimates that are robust against the measurement error problem. Our empirical evaluations, using both synthetic and real-world datasets, show that EnsembleIV can effectively reduce estimation biases across several common regression specifications, and can be combined with modern deep learning techniques when dealing with unstructured data.
با وجود افزایش محبوبیت در مطالعات تجربی ، ادغام یادگیری ماشین متغیرهایی را برای مدل های رگرسیون برای آماری ایجاد کرد استنباط از مشکل خطای اندازه گیری رنج می برد ، که می تواند تخمین تعصب داشته باشد و اعتبار استنباط ها را تهدید می کند.در این مقاله ، ما یک رمان ایجاد می کنیم رویکرد برای کاهش تعصبات برآورد مرتبط.رویکرد پیشنهادی ما ، Ensembleiv ، متغیرهای ابزاری معتبر و قوی را از زبان آموزان ضعیف ایجاد می کند در یک مدل گروه ، و از آنها برای به دست آوردن برآوردهای مداوم که هستند استفاده می کند در برابر مشکل خطای اندازه گیری قوی است.ارزیابی های تجربی ما ، با استفاده از هر دو مجموعه داده مصنوعی و واقعی در دنیای واقعی ، نشان می دهد که Ensembleiv می تواند به طور مؤثر باشد تعصب تخمین را در چندین مشخصات رگرسیون مشترک کاهش دهید ، و می توان هنگام برخورد با تکنیک های مدرن یادگیری عمیق ترکیب شد داده های بدون ساختار.
29,735
This paper studies the identification of perceived ex ante returns in the context of binary human capital investment decisions. The environment is characterised by uncertainty about future outcomes, with some uncertainty being resolved over time. In this context, each individual holds a probability distribution over different levels of returns. The paper uses the hypothetical choice methodology to identify nonparametrically the population distribution of several individual-specific distribution parameters, which are crucial for counterfactual policy analyses. The empirical application estimates perceived returns on overstaying for Afghan asylum seekers in Germany and evaluates the effect of assisted voluntary return policies.
در این مقاله به بررسی شناسایی بازده های سابق درک شده در زمینه تصمیمات سرمایه گذاری باینری سرمایه انسانی.محیط است با عدم اطمینان در مورد نتایج آینده ، با برخی از عدم اطمینان مشخص می شود با گذشت زمان حل شد.در این زمینه ، هر فرد احتمال دارد توزیع در سطوح مختلف بازده.مقاله از فرضیه استفاده می کند روش انتخاب برای شناسایی غیرعادی توزیع جمعیت چندین پارامتر توزیع خاص فردی ، که برای آن بسیار مهم هستند تجزیه و تحلیل سیاست های متقابل.برآورد برنامه تجربی درک شده بازگرداندن بیش از حد برای پناهجویان افغانستان در آلمان و ارزیابی تأثیر سیاست های بازگشت داوطلبانه.
29,736
This paper proposes a nonparametric estimator of the counterfactual copula of two outcome variables that would be affected by a policy intervention. The proposed estimator allows policymakers to conduct ex-ante evaluations by comparing the estimated counterfactual and actual copulas as well as their corresponding measures of association. Asymptotic properties of the counterfactual copula estimator are established under regularity conditions. These conditions are also used to validate the nonparametric bootstrap for inference on counterfactual quantities. Simulation results indicate that our estimation and inference procedures perform well in moderately sized samples. Applying the proposed method to studying the effects of college education on intergenerational income mobility under two counterfactual scenarios, we find that while providing some college education to all children is unlikely to promote mobility, offering a college degree to children from less educated families can significantly reduce income persistence across generations.
در این مقاله یک برآوردگر غیر پارامتری از کوپول ضد خلاف واقع شده است دو متغیر نتیجه که تحت تأثیر مداخله سیاست قرار می گیرند.در برآوردگر پیشنهادی به سیاست گذاران اجازه می دهد تا ارزیابی های سابق را انجام دهند مقایسه کپی های ضد واقع و واقعی تخمین زده شده و همچنین آنها اقدامات مربوط به انجمن.خصوصیات بدون علامت برآوردگر کپی ضد خلاف واقع در شرایط منظم ایجاد می شود. این شرایط همچنین برای اعتبار سنجی بوت استرپ غیر پارامتری برای استفاده می شود استنتاج در مقادیر ضد خلاف.نتایج شبیه سازی نشان می دهد که ما روشهای تخمین و استنباط در نمونه های متوسط ​​عملکرد خوبی دارند. استفاده از روش پیشنهادی برای مطالعه اثرات آموزش دانشگاه بر تحرک درآمد بین نسلی تحت دو سناریوی ضد خلاف واقع ، ما می یابیم این در حالی که ارائه برخی از تحصیلات دانشگاهی به همه کودکان بعید است تحرک را ارتقا دهید و دارای مدرک دانشگاهی برای کودکان از تحصیل کمتر است خانواده ها می توانند به طور قابل توجهی پایداری درآمد را در نسل ها کاهش دهند.
29,737
Identification-robust hypothesis tests are commonly based on the continuous updating objective function or its score. When the number of moment conditions grows proportionally with the sample size, the large-dimensional weighting matrix prohibits the use of conventional asymptotic approximations and the behavior of these tests remains unknown. We show that the structure of the weighting matrix opens up an alternative route to asymptotic results when, under the null hypothesis, the distribution of the moment conditions is reflection invariant. In a heteroskedastic linear instrumental variables model, we then establish asymptotic normality of conventional tests statistics under many instrument sequences. A key result is that the additional terms that appear in the variance are negative. Revisiting a study on the elasticity of substitution between immigrant and native workers where the number of instruments is over a quarter of the sample size, the many instrument-robust approximation indeed leads to substantially narrower confidence intervals.
آزمون های فرضیه شناسایی-روبست معمولاً مبتنی بر مداوم است به روزرسانی عملکرد هدف یا نمره آن.وقتی تعداد شرایط لحظه ای به طور متناسب با اندازه نمونه ، وزن بزرگ بعدی ماتریس استفاده از تقریب های بدون علامت معمولی و رفتار این آزمایشات ناشناخته است.ما نشان می دهیم که ساختار ماتریس وزنه برداری یک مسیر جایگزین را برای نتایج بدون علامت باز می کند ، تحت فرضیه تهی ، توزیع شرایط لحظه ای است بازتاب ثابت.در یک مدل متغیرهای ابزاری خطی ناهمگن ، ما سپس عادی بودن بدون علامت آمار آزمایشات متعارف را در زیر ایجاد می کنیم بسیاری از توالی ابزار.نتیجه کلیدی این است که اصطلاحات اضافی که ظاهر شده در واریانس منفی است.تجدید نظر در یک مطالعه در مورد خاصیت ارتجاعی تعویض بین کارگران مهاجر و بومی که تعداد آنها ابزارها بیش از یک چهارم از اندازه نمونه هستند ، بسیاری تقریب در واقع منجر به فواصل اعتماد به نفس باریک تر می شود.
29,738
This paper proposes a novel approach for identifying coefficients in an earnings dynamics model with arbitrarily dependent contemporaneous income shocks. Traditional methods relying on second moments fail to identify these coefficients, emphasizing the need for nongaussianity assumptions that capture information from higher moments. Our results contribute to the literature on earnings dynamics by allowing models of earnings to have, for example, the permanent income shock of a job change to be linked to the contemporaneous transitory income shock of a relocation bonus.
در این مقاله یک رویکرد جدید برای شناسایی ضرایب در یک ارائه شده است مدل دینامیک درآمد با درآمد معاصر وابسته به خودسرانه شوکروشهای سنتی با تکیه بر لحظات دوم در شناسایی اینها ناکام هستند ضرایب ، با تأکید بر نیاز به فرضیات غیرقانونی که ضبط می شود اطلاعات از لحظات بالاتر.نتایج ما به ادبیات کمک می کند دینامیک درآمد با اجازه دادن به مدل های درآمد ، به عنوان مثال ، شوک درآمدی دائمی از تغییر شغلی که به همزمان مرتبط است شوک درآمد گذرا از یک جایزه جابجایی.
29,739
We consider penalized extremum estimation of a high-dimensional, possibly nonlinear model that is sparse in the sense that most of its parameters are zero but some are not. We use the SCAD penalty function, which provides model selection consistent and oracle efficient estimates under suitable conditions. However, asymptotic approximations based on the oracle model can be inaccurate with the sample sizes found in many applications. This paper gives conditions under which the bootstrap, based on estimates obtained through SCAD penalization with thresholding, provides asymptotic refinements of size \(O \left( n^{- 2} \right)\) for the error in the rejection (coverage) probability of a symmetric hypothesis test (confidence interval) and \(O \left( n^{- 1} \right)\) for the error in rejection (coverage) probability of a one-sided or equal tailed test (confidence interval). The results of Monte Carlo experiments show that the bootstrap can provide large reductions in errors in coverage probabilities. The bootstrap is consistent, though it does not necessarily provide asymptotic refinements, even if some parameters are close but not equal to zero. Random-coefficients logit and probit models and nonlinear moment models are examples of models to which the procedure applies.
ما برآورد افراطی مجازات شده از یک ابعاد بالا و احتمالاً در نظر می گیریم مدل غیرخطی که پراکنده است به این معنا که بیشتر پارامترهای آن هستند صفر اما برخی نیستند.ما از عملکرد مجازات SCAD استفاده می کنیم ، که مدل را ارائه می دهد برآوردهای کارآمد و سازگار و کارآمد در شرایط مناسب. با این حال ، تقریبی بدون علامت بر اساس مدل اوراکل می تواند نادرست باشد با اندازه نمونه موجود در بسیاری از برنامه ها.این مقاله شرایط را ارائه می دهد براساس برآوردهای بدست آمده از طریق SCAD ، بوت استرپ است. مجازات با آستانه ، اصلاحات بدون علامت از اندازه را فراهم می کند (o \ سمت چپ (n^{- 2} \ راست) \) برای خطا در احتمال رد (پوشش) آزمون فرضیه متقارن (فاصله اطمینان) و \ (o \ سمت چپ (n^{- 1} \ راست) \) برای خطا در رد (پوشش) احتمال یک یک طرفه یا آزمون دم مساوی (فاصله اطمینان).نتایج آزمایشات مونت کارلو نشان می دهد که بوت استرپ می تواند کاهش زیادی در خطاها در پوشش ایجاد کند احتمالاتبوت استرپ سازگار است ، اگرچه لزوماً نیست تصفیه های بدون علامت را ارائه دهید ، حتی اگر برخی از پارامترها نزدیک باشند اما برابر نیستند به صفرمدلهای پیوسته و مشارکت تصادفی و لحظه غیرخطی مدل ها نمونه هایی از مدل هایی هستند که این روش در آن اعمال می شود.
29,740
We examine asymptotic properties of the OLS estimator when the values of the regressor of interest are assigned randomly and independently of other regressors. We find that the OLS variance formula in this case is often simplified, sometimes substantially. In particular, when the regressor of interest is independent not only of other regressors but also of the error term, the textbook homoskedastic variance formula is valid even if the error term and auxiliary regressors exhibit a general dependence structure. In the context of randomized controlled trials, this conclusion holds in completely randomized experiments with constant treatment effects. When the error term is heteroscedastic with respect to the regressor of interest, the variance formula has to be adjusted not only for heteroscedasticity but also for correlation structure of the error term. However, even in the latter case, some simplifications are possible as only a part of the correlation structure of the error term should be taken into account. In the context of randomized control trials, this implies that the textbook homoscedastic variance formula is typically not valid if treatment effects are heterogenous but heteroscedasticity-robust variance formulas are valid if treatment effects are independent across units, even if the error term exhibits a general dependence structure. In addition, we extend the results to the case when the regressor of interest is assigned randomly at a group level, such as in randomized control trials with treatment assignment determined at a group (e.g., school/village) level.
ما خواص بدون علامت برآوردگر OLS را در هنگام مقادیر مقادیر بررسی می کنیم رگرسیون علاقه به طور تصادفی و مستقل از سایر موارد اختصاص می یابد رگرسیونما می دانیم که فرمول واریانس OLS در این مورد اغلب است ساده ، گاهی به طور قابل ملاحظه ای.به طور خاص ، هنگامی که رگرسیونر علاقه نه تنها از سایر رگرسیونرها بلکه از این خطا مستقل است اصطلاح ، کتاب درسی فرمول واریانس homoskedastic حتی اگر خطا معتبر باشد رگرسیونرهای ترم و کمکی یک ساختار وابستگی عمومی را نشان می دهند.در زمینه کارآزمایی های کنترل شده تصادفی ، این نتیجه گیری به طور کامل ادامه دارد آزمایشات تصادفی با اثرات درمانی مداوم.وقتی اصطلاح خطا است هتروسکوژیک با توجه به رگرسیون مورد علاقه ، فرمول واریانس باید نه تنها برای ناهمگونی بلکه برای همبستگی تنظیم شود ساختار اصطلاح خطا.با این حال ، حتی در مورد دوم ، برخی ساده سازی ها فقط بخشی از ساختار همبستگی ممکن است اصطلاح خطا باید در نظر گرفته شود.در زمینه کنترل تصادفی آزمایشات ، این بدان معنی است که فرمول واریانس همجنسگرا کتاب درسی است به طور معمول اگر اثرات درمانی ناهمگن باشد اما معتبر نیست در صورت اثرات درمانی ، فرمول های واریانس-روبست معتبر هستند مستقل در واحدها ، حتی اگر اصطلاح خطا وابستگی عمومی داشته باشد ساختارعلاوه بر این ، ما نتایج را به پرونده ای می پردازیم که رگرسیون بهره به طور تصادفی در سطح گروهی مانند کنترل تصادفی اختصاص می یابد آزمایشات با تکلیف درمانی که در یک گروه تعیین شده است (به عنوان مثال ، مدرسه/روستا) مرحله.
29,741
Locally Robust (LR)/Orthogonal/Debiased moments have proven useful with machine learning first steps, but their existence has not been investigated for general parameters. In this paper, we provide a necessary and sufficient condition, referred to as Restricted Local Non-surjectivity (RLN), for the existence of such orthogonal moments to conduct robust inference on general parameters of interest in regular semiparametric models. Importantly, RLN does not require either identification of the parameters of interest or the nuisance parameters. However, for orthogonal moments to be informative, the efficient Fisher Information matrix for the parameter must be non-zero (though possibly singular). Thus, orthogonal moments exist and are informative under more general conditions than previously recognized. We demonstrate the utility of our general results by characterizing orthogonal moments in a class of models with Unobserved Heterogeneity (UH). For this class of models our method delivers functional differencing as a special case. Orthogonality for general smooth functionals of the distribution of UH is also characterized. As a second major application, we investigate the existence of orthogonal moments and their relevance for models defined by moment restrictions with possibly different conditioning variables. We find orthogonal moments for the fully saturated two stage least squares, for heterogeneous parameters in treatment effects, for sample selection models, and for popular models of demand for differentiated products. We apply our results to the Oregon Health Experiment to study heterogeneous treatment effects of Medicaid on different health outcomes.
لحظه های محلی قوی (LR)/متعامد/دفع شده با آن مفید بوده اند اولین مراحل یادگیری ماشین ، اما وجود آنها برای آن مورد بررسی قرار نگرفته است پارامترهای عمومی.در این مقاله ، ما لازم و کافی را ارائه می دهیم شرط ، به عنوان غیرقانونی محلی (RLN) محدود ، برای وجود چنین لحظات متعامد برای انجام استنتاج قوی بر روی عمومی پارامترهای مورد علاقه در مدلهای نیمه نیمه منظم.نکته مهم ، RLN این کار را می کند نیازی به شناسایی پارامترهای مورد علاقه یا مزاحمت نیست مولفه های.با این حال ، برای اینکه لحظات متعامد آموزنده باشد ، کارآمد است ماتریس اطلاعات فیشر برای پارامتر باید غیر صفر باشد (هر چند احتمالاً مفرد)بنابراین ، لحظات متعامد وجود دارد و تحت تأثیر بیشتر است شرایط عمومی از آنچه قبلاً شناخته شده بود.ما ابزار را نشان می دهیم نتایج کلی ما با توصیف لحظات متعامد در یک کلاس از مدل ها با ناهمگونی بدون نظارت (UH).برای این کلاس از مدل ها روش ما تفاوت عملکردی را به عنوان یک مورد خاص ارائه می دهد.ارتوگونی برای عمومی عملکردهای صاف توزیع UH نیز مشخص می شود.به عنوان یک ثانیه برنامه اصلی ، ما در مورد وجود لحظات متعامد و آنها بررسی می کنیم ارتباط برای مدل های تعریف شده توسط محدودیت های لحظه ای با احتمالاً متفاوت متغیرهای تهویه.ما لحظات متعامد را برای دو نفر کاملاً اشباع پیدا می کنیم حداقل مربع ها ، برای پارامترهای ناهمگن در اثرات درمانی ، برای مدل های انتخاب نمونه ، و برای مدل های محبوب تقاضا برای تمایز محصولاتما نتایج خود را در آزمایش بهداشت Oregon برای مطالعه اعمال می کنیم اثرات درمانی ناهمگن Medicaid بر پیامدهای مختلف سلامت.
29,742
We discuss estimating conditional treatment effects in regression discontinuity designs with multiple scores. While local linear regressions have been popular in settings where the treatment status is completely described by one running variable, they do not easily generalize to empirical applications involving multiple treatment assignment rules. In practice, the multivariate problem is usually reduced to a univariate one where using local linear regressions is suitable. Instead, we propose a forest-based estimator that can flexibly model multivariate scores, where we build two honest forests in the sense of Wager and Athey (2018) on both sides of the treatment boundary. This estimator is asymptotically normal and sidesteps the pitfalls of running local linear regressions in higher dimensions. In simulations, we find our proposed estimator outperforms local linear regressions in multivariate designs and is competitive against the minimax-optimal estimator of Imbens and Wager (2019). The implementation of this estimator is simple, can readily accommodate any (fixed) number of running variables, and does not require estimating any nuisance parameters of the data generating process.
ما در مورد تخمین اثرات درمانی مشروط در رگرسیون بحث می کنیم طرح های ناپیوستگی با چند امتیاز.در حالی که رگرسیون های خطی محلی دارند در تنظیماتی که وضعیت درمان به طور کامل توصیف شده است محبوب بوده است یک متغیر در حال اجرا ، آنها به راحتی در برنامه های تجربی تعمیم نمی دهند شامل چندین قوانین واگذاری درمانی.در عمل ، چند متغیره معمولاً مشکل به یک متغیره در جایی که از خطی محلی استفاده می شود کاهش می یابد رگرسیون مناسب است.در عوض ، ما یک برآوردگر مبتنی بر جنگل پیشنهاد می کنیم که می تواند با انعطاف پذیری نمرات چند متغیره ، جایی که ما دو جنگل صادقانه در آن می سازیم Sense of Wager and Athey (2018) در هر دو طرف مرز درمانی.این برآوردگر از نظر غیر عادی طبیعی است و خطاهای اجرای محلی را از بین می برد رگرسیون خطی در ابعاد بالاتر.در شبیه سازی ها ، ما پیشنهادی را می یابیم برآوردگر از رگرسیون های خطی محلی در طرح های چند متغیره بهتر عمل می کند و است رقابتی در برابر برآوردگر حداقلی از IMBENS و WAGER (2019). اجرای این برآوردگر ساده است ، می تواند به راحتی هر یک را در خود جای دهد (ثابت) تعداد متغیرهای در حال اجرا ، و نیازی به تخمین ندارد پارامترهای مزاحمت فرآیند تولید داده.
29,743
This paper proposes a novel tool to nonparametrically identify models with a discrete endogenous variable or treatment: semi-instrumental variables (semi-IVs). A semi-IV is a variable that is relevant but only partially excluded from the potential outcomes, i.e., excluded from at least one, but not necessarily all, potential outcome equations. It follows that standard instrumental variables (IVs), which are fully excluded from all the potential outcomes, are a special (extreme) case of semi-IVs. I show that full exclusion is stronger than necessary because the same objects that are usually identified with an IV (Imbens and Angrist, 1994; Heckman and Vytlacil, 2005; Chernozhukov and Hansen, 2005) can be identified with several semi-IVs instead, provided there is (at least) one semi-IV excluded from each potential outcome. For applied work, tackling endogeneity with semi-IVs instead of IVs should be an attractive alternative, since semi-IVs are easier to find: most selection-specific costs or benefits can be valid semi-IVs, for example. The paper also provides a simple semi-IV GMM estimator for models with homogenous treatment effects and uses it to estimate the returns to education.
در این مقاله یک ابزار جدید برای شناسایی غیر پارامتری مدل ها با a ارائه شده است متغیر یا درمان درون زا گسسته: متغیرهای نیمه سازنده (نیمه IVS).نیمه IV متغیری است که مرتبط است اما فقط تا حدی مستثنی از نتایج بالقوه ، یعنی حداقل از یک نفر ، اما نه لزوماً همه ، معادلات نتیجه بالقوه.از آن استاندارد پیروی می کند متغیرهای ابزاری (IVS) ، که کاملاً از تمام پتانسیل ها خارج شده اند نتایج ، یک مورد خاص (شدید) از نیمه IV است.من آن محرومیت کامل را نشان می دهم قوی تر از حد لازم است زیرا همان اشیاء که معمولاً شناسایی می شوند با IV (Imbens and Angrist ، 1994 ؛ Heckman and Vytlacil ، 2005 ؛ Chernozhukov و هانسن ، 2005) به جای آن ، با چندین نیمه IV ، ارائه می شود ، ارائه می شود. (حداقل) یک نیمه IV از هر نتیجه بالقوه مستثنی است.برای کار کاربردی ، مقابله با درون زایی با نیمه IVS به جای IV باید یک باشد جایگزین جذاب ، از آنجا که پیدا کردن نیمه IV ها آسانتر است: به عنوان مثال هزینه ها یا مزایای خاص انتخاب می تواند نیمه IVS معتبر باشد.در مقاله همچنین یک برآوردگر ساده GMM نیمه IV برای مدل هایی با همگن ارائه می دهد اثرات درمانی و از آن برای تخمین بازده به آموزش استفاده می کند.
29,744
This study proposes an estimator that combines statistical identification with economically motivated restrictions on the interactions. The estimator is identified by (mean) independent non-Gaussian shocks and allows for incorporation of uncertain prior economic knowledge through an adaptive ridge penalty. The estimator shrinks towards economically motivated restrictions when the data is consistent with them and stops shrinkage when the data provides evidence against the restriction. The estimator is applied to analyze the interaction between the stock and oil market. The results suggest that what is usually identified as oil-specific demand shocks can actually be attributed to information shocks extracted from the stock market, which explain about 30-40% of the oil price variation.
این مطالعه یک برآوردگر را پیشنهاد می کند که شناسایی آماری است با محدودیت های با انگیزه اقتصادی در تعامل.برآوردگر است مشخص شده توسط (میانگین) شوکهای مستقل غیر گاوسی و اجازه می دهد ترکیب دانش اقتصادی نامشخص قبلی از طریق یک خط الراس تطبیقی مجازاتبرآوردگر وقتی به سمت محدودیت های با انگیزه اقتصادی کوچک می شود داده ها با آنها سازگار است و هنگامی که داده ها ارائه می دهند ، کوچک شدن را متوقف می کنند شواهد در برابر محدودیت.برآوردگر برای تجزیه و تحلیل استفاده می شود تعامل بین سهام و بازار نفت.نتایج نشان می دهد که چیست معمولاً به عنوان شوکهای تقاضای خاص روغن در واقع می توان به آن نسبت داد شوکهای اطلاعاتی استخراج شده از بازار سهام ، که حدود 30-40 ٪ توضیح می دهد از تغییر قیمت نفت
29,745
We introduce a simple tool to control for false discoveries and identify individual signals in scenarios involving many tests, dependent test statistics, and potentially sparse signals. The tool applies the Cauchy combination test recursively on a sequence of expanding subsets of $p$-values and is referred to as the sequential Cauchy combination test. While the original Cauchy combination test aims to make a global statement about a set of null hypotheses by summing transformed $p$-values, our sequential version determines which $p$-values trigger the rejection of the global null. The sequential test achieves strong familywise error rate control, exhibits less conservatism compared to existing controlling procedures when dealing with dependent test statistics, and provides a power boost. As illustrations, we revisit two well-known large-scale multiple testing problems in finance for which the test statistics have either serial dependence or cross-sectional dependence, namely monitoring drift bursts in asset prices and searching for assets with a nonzero alpha. In both applications, the sequential Cauchy combination test proves to be a preferable alternative. It overcomes many of the drawbacks inherent to inequality-based controlling procedures, extreme value approaches, resampling and screening methods, and it improves the power in simulations, leading to distinct empirical outcomes.
ما یک ابزار ساده برای کنترل اکتشافات دروغین و شناسایی معرفی می کنیم سیگنال های فردی در سناریوهایی که شامل بسیاری از تست ها هستند ، آزمون وابسته آمار و سیگنال های بالقوه پراکنده.این ابزار کاوشی را اعمال می کند تست ترکیبی به صورت بازگشتی بر روی دنباله ای از زیر مجموعه های $ p $-$ و به عنوان آزمون ترکیبی متوالی Cauchy گفته می شود.در حالی که آزمون ترکیبی اصلی Cauchy با هدف بیان یک بیانیه جهانی در مورد مجموعه ای از فرضیه های تهی با جمع بندی $ P $-Values ​​، نسخه پی در پی ما تعیین می کند که چه مقدار $ p $ باعث رد تهی جهانی می شود.در آزمایش متوالی به کنترل نرخ خطای خانوادگی قوی دست می یابد ، کمتر نشان می دهد محافظه کاری در مقایسه با روشهای کنترل موجود هنگام برخورد آمار آزمون وابسته ، و تقویت قدرت را فراهم می کند.به عنوان تصاویر ، ما مجدداً دو مشکل آزمایش چندگانه در مقیاس بزرگ را در امور مالی مجدداً مورد بررسی قرار دهید که آمار آزمون دارای وابستگی سریال یا مقطعی است وابستگی ، یعنی نظارت بر انفجار رانش در قیمت دارایی و جستجوی آن دارایی با یک آلفا غیرزرو.در هر دو برنامه ، cauchy متوالی آزمون ترکیبی ثابت می کند که یک جایگزین ارجح است.این غلبه بر بسیاری از اشکالاتی که ذاتی به روشهای کنترل مبتنی بر نابرابری است ، شدید رویکردهای ارزش ، روشهای پر کردن و غربالگری و قدرت را بهبود می بخشد در شبیه سازی ها ، منجر به نتایج تجربی متمایز می شود.
29,746
This paper characterizes point identification results of the local average treatment effect (LATE) using two imperfect instruments. The classical approach (Imbens and Angrist (1994)) establishes the identification of LATE via an instrument that satisfies exclusion, monotonicity, and independence. However, it may be challenging to find a single instrument that satisfies all these assumptions simultaneously. My paper uses two instruments but imposes weaker assumptions on both instruments. The first instrument is allowed to violate the exclusion restriction and the second instrument does not need to satisfy monotonicity. Therefore, the first instrument can affect the outcome via both direct effects and a shift in the treatment status. The direct effects can be identified via exogenous variation in the second instrument and therefore the local average treatment effect is identified. An estimator is proposed, and using Monte Carlo simulations, it is shown to perform more robustly than the instrumental variable estimand.
در این مقاله نتایج شناسایی نقطه میانگین محلی مشخص شده است اثر درمانی (دیر) با استفاده از دو ابزار ناقص.رویکرد کلاسیک (Imbens and Angrist (1994)) شناسایی اواخر از طریق ابزاری که محرومیت ، یکنواختی و استقلال را برآورده می کند.با این حال، ممکن است پیدا کردن یک ابزار واحد که همه اینها را برآورده می کند ، چالش برانگیز باشد فرضیات به طور همزمان.مقاله من از دو ابزار استفاده می کند اما ضعیف تر را تحمیل می کند فرضیات مربوط به هر دو ابزار.اولین ابزار مجاز به نقض آن است محدودیت محرومیت و ابزار دوم نیازی به برآورده کردن ندارد یکنواختیبنابراین ، ابزار اول می تواند از طریق هر دو بر نتیجه تأثیر بگذارد اثرات مستقیم و تغییر وضعیت درمان.اثرات مستقیم می تواند باشد از طریق تنوع اگزوژن در ابزار دوم مشخص شده و بنابراین اثر درمانی متوسط ​​موضعی مشخص شده است.یک برآوردگر پیشنهاد شده است ، و با استفاده از شبیه سازی های مونت کارلو ، نشان داده شده است که قوی تر از آن عمل می کند برآورد متغیر ابزاری.
29,747
This paper proposes a framework to analyze the effects of counterfactual policies on the unconditional quantiles of an outcome variable. For a given counterfactual policy, we obtain identified sets for the effect of both marginal and global changes in the proportion of treated individuals. To conduct a sensitivity analysis, we introduce the quantile breakdown frontier, a curve that (i) indicates whether a sensitivity analysis if possible or not, and (ii) when a sensitivity analysis is possible, quantifies the amount of selection bias consistent with a given conclusion of interest across different quantiles. To illustrate our method, we perform a sensitivity analysis on the effect of unionizing low income workers on the quantiles of the distribution of (log) wages.
در این مقاله چارچوبی برای تجزیه و تحلیل اثرات ضد خلاف واقع شده است سیاست های مربوط به مقادیر بی قید و شرط از یک متغیر نتیجه.برای یک داده شده سیاست ضد خلاف ، ما مجموعه های شناسایی شده ای را برای تأثیر هر دو به دست می آوریم تغییرات حاشیه ای و جهانی در نسبت افراد تحت درمان.به تجزیه و تحلیل حساسیت را انجام دهید ، ما مرز شکست کمی را معرفی می کنیم ، a منحنی که (i) نشان می دهد که آیا در صورت امکان تجزیه و تحلیل حساسیت یا نه ، و (ب) هنگامی که یک تجزیه و تحلیل حساسیت امکان پذیر است ، مقدار آن را تعیین می کند تعصب انتخاب سازگار با نتیجه گیری مشخص از علاقه در سراسر مقادیربرای نشان دادن روش خود ، ما یک تجزیه و تحلیل حساسیت را در مورد انجام می دهیم تأثیر اتحاد کارگران کم درآمد بر تعداد توزیع توزیع (ورود به سیستم) دستمزد.
29,748
We revisit identification based on timing and information set assumptions in structural models, which have been used in the context of production functions, demand equations, and hedonic pricing models (e.g. Olley and Pakes (1996), Blundell and Bond (2000)). First, we demonstrate a general under-identification problem using these assumptions in a simple version of the Blundell-Bond dynamic panel model. In particular, the basic moment conditions can yield multiple discrete solutions: one at the persistence parameter in the main equation and another at the persistence parameter governing the regressor. We then show that the problem can persist in a broader set of models but disappears in models under stronger timing assumptions. We then propose possible solutions in the simple setting by enforcing an assumed sign restriction and conclude by using lessons from our basic identification approach to propose more general practical advice for empirical researchers.
ما شناسایی را بر اساس زمان بندی و فرضیات تنظیم شده در اطلاعات دوباره بررسی می کنیم مدل های ساختاری ، که در زمینه توابع تولید استفاده شده است ، معادلات تقاضا ، و مدل های قیمت گذاری هیدونیک (به عنوان مثال اوللی و پیکس (1996) ، بلوندل و باند (2000)).اول ، ما یک شناسایی عمومی را نشان می دهیم مشکل با استفاده از این فرضیات در یک نسخه ساده از بلوندل بلوندل مدل پانل پویا.به طور خاص ، شرایط اصلی لحظه می تواند به دست آورد چندین راه حل گسسته: یکی در پارامتر پایداری در اصلی معادله و دیگری در پارامتر پایداری حاکم بر رگرسور.ما سپس نشان دهید که مشکل می تواند در مجموعه گسترده تری از مدل ها ادامه یابد اما در مدل ها تحت فرضیات زمان بندی قوی تر ناپدید می شود.سپس پیشنهاد می کنیم راه حل های ممکن در تنظیم ساده با اجرای یک علامت فرضی محدودیت و نتیجه گیری با استفاده از دروس از شناسایی اساسی ما رویکرد برای پیشنهاد مشاوره عمومی تر برای محققان تجربی.
29,749
We study identification and estimation of endogenous linear and nonlinear regression models without excluded instrumental variables, based on the standard mean independence condition and a nonlinear relevance condition. Based on the identification results, we propose two semiparametric estimators as well as a discretization-based estimator that does not require any nonparametric regressions. We establish their asymptotic normality and demonstrate via simulations their robust finite-sample performances with respect to exclusion restrictions violations and endogeneity. Our approach is applied to study the returns to education, and to test the direct effects of college proximity indicators as well as family background variables on the outcome.
ما شناسایی و تخمین خطی و غیرخطی درون زا را مطالعه می کنیم مدل های رگرسیون بدون متغیرهای ابزاری محروم ، بر اساس میانگین شرایط استقلال استاندارد و شرایط ارتباط غیرخطی.مستقر در مورد نتایج شناسایی ، ما دو برآوردگر نیمهرامتری را نیز پیشنهاد می کنیم به عنوان یک برآوردگر مبتنی بر گسسته که نیازی به غیر پارامتری ندارد رگرسیونما عادی بودن بدون علامت آنها را تعیین می کنیم و از طریق نشان می دهیم شبیه سازی عملکردهای نمونه محدود آنها را با توجه به محرومیت شبیه سازی کنید محدودیت های نقض و درون زا.رویکرد ما برای مطالعه به آموزش و پرورش و آزمایش اثرات مستقیم نزدیکی دانشکده باز می گردد شاخص ها و همچنین متغیرهای پس زمینه خانوادگی در نتیجه.
29,750
This paper studies semiparametric identification of substitution and complementarity patterns between two goods using a panel multinomial choice model with bundles. The model allows the two goods to be either substitutes or complements and admits heterogeneous complementarity through observed characteristics. I first provide testable implications for the complementarity relationship between goods. I then characterize the sharp identified set for the model parameters and provide sufficient conditions for point identification. The identification analysis accommodates endogenous covariates through flexible dependence structures between observed characteristics and fixed effects while placing no distributional assumptions on unobserved preference shocks. My method is shown to perform more robustly than the parametric method through Monte Carlo simulations. As an extension, I allow for unobserved heterogeneity in the complementarity, investigate scenarios involving more than two goods, and study a class of nonseparable utility functions.
در این مقاله به بررسی شناسایی نیمه پارامتری جایگزینی و الگوهای مکمل بین دو کالا با استفاده از یک انتخاب چند پانل مدل با بسته ها.این مدل اجازه می دهد تا دو کالای جایگزین یا جایگزین شوند مکمل ناهمگن را از طریق مشاهده شده مکمل و پذیرفته است مشخصات.من برای اولین بار پیامدهای قابل آزمایش را برای مکمل ارائه می دهم رابطه بین کالاها.سپس مجموعه تیز مشخص شده را برای آن مشخص می کنم پارامترهای مدل و شرایط کافی برای نقطه فراهم می کند شناسایی.تجزیه و تحلیل شناسایی در هم متغیرهای درون زا قرار دارد از طریق ساختارهای وابستگی انعطاف پذیر بین خصوصیات مشاهده شده و اثرات ثابت در حالی که هیچ فرضیه توزیع را بر روی آنها قرار نمی دهد شوک اولویتروش من نشان داده شده است که قوی تر از آن عمل می کند روش پارامتری از طریق شبیه سازی مونت کارلو.به عنوان یک پسوند ، من اجازه می دهم ناهمگونی بدون نظارت در مکمل ، سناریوها را بررسی کنید شامل بیش از دو کالا ، و مطالعه یک کلاس از ابزار غیر قابل تفکیک کارکرد.
29,751
Granular instrumental variables (GIV) has experienced sharp growth in empirical macro-finance. The methodology's rise showcases granularity's potential for identification in a wide set of economic environments, like the estimation of spillovers and demand systems. I propose a new estimator--called robust granular instrumental variables (RGIV)--that allows researchers to study unit-level heterogeneity in spillovers within GIV's framework. In contrast to GIV, RGIV also allows for unknown shock variances and does not require skewness of the size distribution of units. I also develop a test of overidentifying restrictions that evaluates RGIV's compatibility with the data, a parameter restriction test that evaluates the appropriateness of the homogeneous spillovers assumption, and extend the framework to allow for observable explanatory variables. Applied to the Euro area, I find strong evidence of country-level heterogeneity in sovereign yield spillovers. In simulations, I show that RGIV produces reliable and informative confidence intervals.
متغیرهای ابزاری گرانول (GIV) رشد شدید را در آن تجربه کرده است کلان کلان تجربی.ویترین روش شناسی به نمایش می رود پتانسیل شناسایی در مجموعه گسترده ای از محیط های اقتصادی ، مانند برآورد اسپیلرها و سیستم های تقاضا.من یک برآوردگر جدید را پیشنهاد می کنم-به نام متغیرهای ابزاری گرانول قوی (RGIV)-که به محققان امکان می دهد مطالعه کنند ناهمگونی سطح واحد در سرریزهای موجود در چارچوب GIV.در مقابل GIV ، RGIV همچنین واریانس شوک ناشناخته را امکان پذیر می کند و نیازی به چرت زدن ندارد توزیع اندازه واحدها.من همچنین آزمایشی از شناسایی بیش از حد ایجاد می کنم محدودیت هایی که سازگاری RGIV با داده ها را ارزیابی می کند ، یک پارامتر تست محدودیت که مناسب بودن همگن را ارزیابی می کند فرض Spillovers ، و چارچوب را گسترش دهید تا قابل مشاهده باشد متغیرهای توضیحی.من در منطقه یورو اعمال می شود ، من شواهد محکمی پیدا می کنم ناهمگونی سطح کشور در سرریزهای عملکرد حاکمیت.در شبیه سازی ها ، من نشان می دهد که RGIV فواصل اعتماد به نفس قابل اعتماد و آموزنده را تولید می کند.
29,752
We introduce a novel framework for individual-level welfare analysis. It builds on a parametric model for continuous demand with a quasilinear utility function, allowing for heterogeneous coefficients and unobserved individual-product-level preference shocks. We obtain bounds on the individual-level consumer welfare loss at any confidence level due to a hypothetical price increase, solving a scalable optimization problem constrained by a new confidence set under an independence restriction. This confidence set is computationally simple, robust to weak instruments, nonlinearity, and partial identification. In addition, it may have applications beyond welfare analysis. Monte Carlo simulations and two empirical applications on gasoline and food demand demonstrate the effectiveness of our method.
ما یک چارچوب جدید برای تجزیه و تحلیل رفاه در سطح فردی معرفی می کنیم.آی تی بر روی یک مدل پارامتری برای تقاضای مداوم با یک ابزار شبه خطی ساخته شده است عملکرد ، اجازه ضرایب ناهمگن و بدون نظارت شوک های اولویت سطح محصول.ما مرزهای خود را بدست می آوریم از دست دادن رفاه مصرف کننده در سطح فردی در هر سطح اطمینان به دلیل افزایش قیمت فرضی ، حل یک مشکل بهینه سازی مقیاس پذیر محدود شده توسط اعتماد به نفس جدیدی که تحت یک محدودیت استقلال قرار گرفته است.این مجموعه اعتماد به نفس از نظر محاسباتی ساده ، قوی برای ابزارهای ضعیف است ، غیرخطی بودن و شناسایی جزئی.علاوه بر این ، ممکن است برنامه هایی داشته باشد فراتر از تحلیل رفاه.شبیه سازی مونت کارلو و دو کاربرد تجربی در مورد بنزین و تقاضای مواد غذایی اثربخشی روش ما را نشان می دهد.
29,753
Many estimators of dynamic discrete choice models with persistent unobserved heterogeneity have desirable statistical properties but are computationally intensive. In this paper we propose a method to quicken estimation for a broad class of dynamic discrete choice problems by exploiting semiparametric index restrictions. Specifically, we propose an estimator for models whose reduced form parameters are injective functions of one or more linear indices (Ahn, Ichimura, Powell and Ruud 2018), a property we term index invertibility. We establish that index invertibility implies a set of equality constraints on the model parameters. Our proposed estimator uses the equality constraints to decrease the dimension of the optimization problem, thereby generating computational gains. Our main result shows that the proposed estimator is asymptotically equivalent to the unconstrained, computationally heavy estimator. In addition, we provide a series of results on the number of independent index restrictions on the model parameters, providing theoretical guidance on the extent of computational gains. Finally, we demonstrate the advantages of our approach via Monte Carlo simulations.
بسیاری ناهمگونی دارای خواص آماری مطلوب اما از نظر محاسباتی است متمرکز.در این مقاله ما روشی را برای سریع تخمین برای یک گسترده پیشنهاد می کنیم کلاس مشکلات انتخاب گسسته پویا با سوء استفاده از شاخص نیمه پارامتری محدودیت های.به طور خاص ، ما برای مدلهایی که کاهش یافته است ، تخمین می زنیم پارامترهای فرم توابع تزریقی یک یا چند شاخص خطی هستند (Ahn ، Ichimura ، Powell and Ruud 2018) ، خاصیتی که ما از این فهرست شاخص استفاده می کنیم.ما تعیین این فهرست عملکرد به معنای مجموعه ای از محدودیت های برابری در پارامترهای مدل.برآوردگر پیشنهادی ما از محدودیت های برابری استفاده می کند ابعاد مشکل بهینه سازی را کاهش دهید ، در نتیجه تولید می شود سود محاسباتی.نتیجه اصلی ما نشان می دهد که برآوردگر پیشنهادی است به صورت مجان برآوردگرعلاوه بر این ، ما مجموعه ای از نتایج را در مورد تعداد ارائه می دهیم محدودیت های شاخص مستقل در پارامترهای مدل ، ارائه نظری راهنمایی در مورد میزان سود محاسباتی.سرانجام ، ما نشان می دهیم مزایای رویکرد ما از طریق شبیه سازی مونت کارلو.
30,218
This paper illustrates two algorithms designed in Forneron & Ng (2020): the resampled Newton-Raphson (rNR) and resampled quasi-Newton (rqN) algorithms which speed-up estimation and bootstrap inference for structural models. An empirical application to BLP shows that computation time decreases from nearly 5 hours with the standard bootstrap to just over 1 hour with rNR, and only 15 minutes using rqN. A first Monte-Carlo exercise illustrates the accuracy of the method for estimation and inference in a probit IV regression. A second exercise additionally illustrates statistical efficiency gains relative to standard estimation for simulation-based estimation using a dynamic panel regression example.
در این مقاله دو الگوریتم طراحی شده در Forneron & ng (2020) نشان داده شده است: مجدداً الگوریتم های نیوتن-رافسون (RNR) و مجدداً مجدداً مجدداً مورد استفاده قرار گرفت که برآورد سرعت و استنباط بوت استرپ برای مدل های ساختاری.در کاربرد تجربی برای BLP نشان می دهد که زمان محاسبه تقریباً کاهش می یابد 5 ساعت با بوت استرپ استاندارد تا بیش از 1 ساعت با RNR و فقط 15 دقیقه با استفاده از RQN.اولین تمرین مونت-کارلو صحت آن را نشان می دهد روش تخمین و استنباط در رگرسیون پروبیت IV.یک لحظه علاوه بر این ، تمرینات بهره وری آماری نسبت به برآورد استاندارد برای برآورد مبتنی بر شبیه سازی با استفاده از یک پانل پویا مثال رگرسیون.
29,754
We present a novel approach to causal measurement for advertising, namely to use exogenous variation in advertising exposure (RCTs) for a subset of ad campaigns to build a model that can predict the causal effect of ad campaigns that were run without RCTs. This approach -- Predictive Incrementality by Experimentation (PIE) -- frames the task of estimating the causal effect of an ad campaign as a prediction problem, with the unit of observation being an RCT itself. In contrast, traditional causal inference approaches with observational data seek to adjust covariate imbalance at the user level. A key insight is to use post-campaign features, such as last-click conversion counts, that do not require an RCT, as features in our predictive model. We find that our PIE model recovers RCT-derived incremental conversions per dollar (ICPD) much better than the program evaluation approaches analyzed in Gordon et al. (forthcoming). The prediction errors from the best PIE model are 48%, 42%, and 62% of the RCT-based average ICPD for upper-, mid-, and lower-funnel conversion outcomes, respectively. In contrast, across the same data, the average prediction error of stratified propensity score matching exceeds 491%, and that of double/debiased machine learning exceeds 2,904%. Using a decision-making framework inspired by industry, we show that PIE leads to different decisions compared to RCTs for only 6% of upper-funnel, 7% of mid-funnel, and 13% of lower-funnel outcomes. We conclude that PIE could enable advertising platforms to scale causal ad measurement by extrapolating from a limited number of RCTs to a large set of non-experimental ad campaigns.
ما یک رویکرد جدید برای اندازه گیری علی برای تبلیغات ارائه می دهیم ، یعنی از تنوع اگزوژن در قرار گرفتن در معرض تبلیغات (RCTs) برای زیر مجموعه AD استفاده کنید کمپین هایی برای ساختن مدلی که می تواند تأثیر علّی تبلیغات تبلیغاتی را پیش بینی کند که بدون RCT اجرا شد.این رویکرد - افزایش پیش بینی کننده توسط آزمایش (PIE) - وظیفه تخمین اثر علی یک کمپین تبلیغاتی به عنوان یک مشکل پیش بینی ، با واحد مشاهده RCT خوددر مقابل ، رویکردهای استنباط علی سنتی با مشاهده داده ها به دنبال تنظیم عدم تعادل متغیر در سطح کاربر هستند.یک بینش کلیدی این است که از ویژگی های پس از انتخابات ، مانند شمارش تبدیل کلیک آخر ، استفاده نکنید. به RCT نیاز دارید ، به عنوان ویژگی های موجود در مدل پیش بینی ما.ما متوجه می شویم که مدل پای ما تبدیل های افزایشی مشتق از RCT را در هر دلار (ICPD) بسیار بهتر از رویکردهای ارزیابی برنامه در گوردون و همکاران مورد تجزیه و تحلیل قرار گرفت.(آینده)در خطاهای پیش بینی از بهترین مدل پای 48 ٪ ، 42 ٪ و 62 ٪ از ICPD متوسط ​​مبتنی بر RCT برای نتایج تبدیل Funnel بالا ، میانی و پایین ، به ترتیب.در مقابل ، در همان داده ها ، خطای پیش بینی متوسط تطبیق نمره گرایش طبقه بندی شده بیش از 491 ٪ و از آن یادگیری ماشین دوتایی/debiased بیش از 2،904 ٪ است.با استفاده از تصمیم گیری چارچوبی با الهام از صنعت ، ما نشان می دهیم که پای منجر به تصمیمات مختلف می شود در مقایسه با RCT ها فقط برای 6 ٪ از Funnel فوقانی ، 7 ٪ از میانه Funnel و 13 ٪ از نتایج پایین تر.نتیجه می گیریم که PIE می تواند سیستم عامل های تبلیغاتی را فعال کند برای مقیاس سنجش علّی AD با برون یابی از تعداد محدودی از RCT ها به مجموعه بزرگی از کمپین های تبلیغاتی غیر تجربی.
29,755
Model mis-specification in multivariate econometric models can strongly influence quantities of interest such as structural parameters, forecast distributions or responses to structural shocks, even more so if higher-order forecasts or responses are considered, due to parameter convolution. We propose a simple method for addressing these specification issues in the context of Bayesian VARs. Our method, called coarsened Bayesian VARs (cBVARs), replaces the exact likelihood with a coarsened likelihood that takes into account that the model might be mis-specified along important but unknown dimensions. Coupled with a conjugate prior, this results in a computationally simple model. As opposed to more flexible specifications, our approach avoids overfitting, is simple to implement and estimation is fast. The resulting cBVAR performs well in simulations for several types of mis-specification. Applied to US data, cBVARs improve point and density forecasts compared to standard BVARs, and lead to milder but more persistent negative effects of uncertainty shocks on output.
مدل سازی نادرست در مدل های اقتصاد سنجی چند متغیره به شدت می تواند بر مقادیر مورد علاقه مانند پارامترهای ساختاری ، پیش بینی تأثیر می گذارد توزیع یا پاسخ به شوک های ساختاری ، حتی بیشتر در صورت مرتب سازی بالاتر پیش بینی ها یا پاسخ ها به دلیل حل و فصل پارامتر در نظر گرفته می شوند.ما پیشنهاد می کنیم یک روش ساده برای پرداختن به این موضوعات مشخصات در متن Vars Bayesian.روش ما ، به نام های Bayesian Vars (CBVARS) نامیده می شود. احتمال دقیق با احتمال درشت که این امر را در نظر می گیرد این مدل ممکن است در ابعاد مهم اما ناشناخته مشخص شود. همراه با یک ترکیب قبلی ، این منجر به یک مدل محاسباتی ساده می شود. برخلاف مشخصات انعطاف پذیر تر ، رویکرد ما از بیش از حد جلوگیری می کند ، اجرای ساده و تخمین سریع است.CBVAR حاصل عملکرد خوبی دارد در شبیه سازی برای چندین نوع مشخصات نادرست.برای داده های ایالات متحده اعمال می شود ، CBVAR پیش بینی های نقطه و چگالی را در مقایسه با BVAR های استاندارد بهبود می بخشد و سرب به اثرات منفی خفیف تر اما مداوم تر شوک های عدم اطمینان بر بازده.
29,756
The accurate prediction of short-term electricity prices is vital for effective trading strategies, power plant scheduling, profit maximisation and efficient system operation. However, uncertainties in supply and demand make such predictions challenging. We propose a hybrid model that combines a techno-economic energy system model with stochastic models to address this challenge. The techno-economic model in our hybrid approach provides a deep understanding of the market. It captures the underlying factors and their impacts on electricity prices, which is impossible with statistical models alone. The statistical models incorporate non-techno-economic aspects, such as the expectations and speculative behaviour of market participants, through the interpretation of prices. The hybrid model generates both conventional point predictions and probabilistic forecasts, providing a comprehensive understanding of the market landscape. Probabilistic forecasts are particularly valuable because they account for market uncertainty, facilitating informed decision-making and risk management. Our model delivers state-of-the-art results, helping market participants to make informed decisions and operate their systems more efficiently.
پیش بینی دقیق قیمت برق کوتاه مدت برای استراتژی های تجاری مؤثر ، برنامه ریزی نیروگاه ، حداکثر سود و عملکرد سیستم کارآمد.با این حال ، عدم قطعیت در عرضه و تقاضا باعث می شود چنین پیش بینی هایی چالش برانگیز است.ما یک مدل ترکیبی را پیشنهاد می کنیم که ترکیبی از a مدل سیستم انرژی فنی و اقتصادی با مدل های تصادفی برای پرداختن به این موضوع چالشمدل فنی و اقتصادی در رویکرد ترکیبی ما عمیق است درک بازار.این عوامل اساسی و آنها را ضبط می کند تأثیر بر قیمت برق ، که با مدل های آماری غیرممکن است تنها.مدلهای آماری جنبه های غیر تکنیکی و اقتصادی را شامل می شوند ، مانند انتظارات و رفتار سوداگرانه شرکت کنندگان در بازار ، از طریق تفسیر قیمت ها.مدل ترکیبی هر دو نقطه معمولی را ایجاد می کند پیش بینی ها و پیش بینی های احتمالی ، ارائه یک جامع درک منظر بازار.پیش بینی های احتمالی به ویژه ارزشمند است زیرا آنها عدم اطمینان در بازار را به خود اختصاص می دهند و آگاهانه را تسهیل می کنند تصمیم گیری و مدیریت ریسک.مدل ما مدرن است نتایج ، کمک به شرکت کنندگان در بازار برای تصمیم گیری و عملکرد آگاهانه سیستم های آنها کارآمدتر است.
29,757
Based on the statistical yearbook data and related patent data of 287 cities in China from 2000 to 2020, this study regards the policy of establishing the national high-tech zones as a quasi-natural experiment. Using this experiment, this study firstly estimated the treatment effect of the policy and checked the robustness of the estimation. Then the study examined the heterogeneity in different geographic demarcation of China and in different city level of China. After that, this study explored the possible influence mechanism of the policy. It shows that the possible mechanism of the policy is financial support, industrial agglomeration of secondary industry and the spillovers. In the end, this study examined the spillovers deeply and showed the distribution of spillover effect.
بر اساس داده های سال آماری و داده های ثبت اختراع مربوط به 287 شهر در چین از سال 2000 تا 2020 ، این مطالعه سیاست ایجاد مناطق ملی پیشرفته به عنوان یک آزمایش شبه طبیعی.با استفاده از این آزمایش ، این مطالعه ابتدا اثر درمانی این سیاست را تخمین زده و بررسی کرد استحکام تخمین.سپس مطالعه ناهمگونی را در مشخص کردن جغرافیایی مختلف چین و در سطح مختلف شهر چین. پس از آن ، این مطالعه به بررسی مکانیسم نفوذ احتمالی این سیاست پرداخت. این نشان می دهد که مکانیسم احتمالی این سیاست پشتیبانی مالی است ، انباشت صنعتی صنعت ثانویه و سرریزها.در پایان، در این مطالعه ، سرریزها عمیقاً مورد بررسی قرار گرفت و توزیع آن را نشان داد اثر سرریز.
29,758
We propose a rate optimal estimator for the linear regression model on network data with interacted (unobservable) individual effects. The estimator achieves a faster rate of convergence $N$ compared to the standard estimators' $\sqrt{N}$ rate and is efficient in cases that we discuss. We observe that the individual effects alter the eigenvalue distribution of the data's matrix representation in significant and distinctive ways. We subsequently offer a correction for the \textit{ordinary least squares}' objective function to attenuate the statistical noise that arises due to the individual effects, and in some cases, completely eliminate it. The new estimator is asymptotically normal and we provide a valid estimator for its asymptotic covariance matrix. While this paper only considers models accounting for first-order interactions between individual effects, our estimation procedure is naturally extendable to higher-order interactions and more general specifications of the error terms.
ما یک برآوردگر مطلوب نرخ را برای مدل رگرسیون خطی در پیشنهاد می کنیم داده های شبکه با اثرات فردی تعامل (غیرقابل کنترل).برآوردگر در مقایسه با برآوردگرهای استاندارد ، به سرعت همگرایی $ n $ دست می یابد $ \ sqrt {n} $ نرخ و در مواردی که در مورد آنها بحث می کنیم کارآمد است.ما مشاهده می کنیم که اثرات فردی توزیع مقادیر ویژه ماتریس داده ها را تغییر می دهد نمایندگی به روشهای مهم و متمایز.ما متعاقباً ارائه می دهیم تصحیح برای \ textit {حداقل مربعات معمولی} ' سر و صدای آماری را که به دلیل اثرات فردی بوجود می آید ، کاهش دهید و در بعضی موارد ، آن را کاملاً از بین ببرید.برآوردگر جدید بدون علامت است طبیعی است و ما یک برآوردگر معتبر برای ماتریس کواریانس بدون علامت آن ارائه می دهیم. در حالی که این مقاله فقط مدل های حسابداری را برای تعامل مرتبه اول در نظر می گیرد بین اثرات فردی ، روش تخمین ما به طور طبیعی قابل گسترش است فعل و انفعالات مرتبه بالاتر و مشخصات کلی تر از اصطلاحات خطا.
29,813
This paper focuses on the task of detecting local episodes involving violation of the standard It\^o semimartingale assumption for financial asset prices in real time that might induce arbitrage opportunities. Our proposed detectors, defined as stopping rules, are applied sequentially to continually incoming high-frequency data. We show that they are asymptotically exponentially distributed in the absence of Ito semimartingale violations. On the other hand, when a violation occurs, we can achieve immediate detection under infill asymptotics. A Monte Carlo study demonstrates that the asymptotic results provide a good approximation to the finite-sample behavior of the sequential detectors. An empirical application to S&P 500 index futures data corroborates the effectiveness of our detectors in swiftly identifying the emergence of an extreme return persistence episode in real time.
این مقاله به وظیفه تشخیص قسمت های محلی مربوط می پردازد نقض استاندارد آن \^o فرض semimartingale برای دارایی مالی قیمت در زمان واقعی که ممکن است فرصت های داوری را القا کند.پیشنهادی ما ردیاب ها ، که به عنوان قوانین توقف تعریف می شوند ، به طور مداوم اعمال می شوند داده های با فرکانس بالا.ما نشان می دهیم که آنها بدون علامت هستند در غیاب تخلفات Semimartingale ITO به صورت تصاعدی توزیع می شود.بر از طرف دیگر ، هنگامی که یک تخلف رخ می دهد ، می توانیم به تشخیص فوری دست یابیم تحت بدون علامت Infill.یک مطالعه مونت کارلو نشان می دهد که بدون علامت نتایج تقریب خوبی به رفتار نمونه محدود از ردیاب های متوالی.یک برنامه تجربی برای داده های آتی شاخص S&P 500 تأیید اثربخشی ردیاب های ما در شناسایی سریع ظهور یک قسمت پایداری بازگشت شدید در زمان واقعی.
29,759
Accurate and reliable prediction of individual travel mode choices is crucial for developing multi-mode urban transportation systems, conducting transportation planning and formulating traffic demand management strategies. Traditional discrete choice models have dominated the modelling methods for decades yet suffer from strict model assumptions and low prediction accuracy. In recent years, machine learning (ML) models, such as neural networks and boosting models, are widely used by researchers for travel mode choice prediction and have yielded promising results. However, despite the superior prediction performance, a large body of ML methods, especially the branch of neural network models, is also limited by overfitting and tedious model structure determination process. To bridge this gap, this study proposes an enhanced multilayer perceptron (MLP; a neural network) with two hidden layers for travel mode choice prediction; this MLP is enhanced by XGBoost (a boosting method) for feature selection and a grid search method for optimal hidden neurone determination of each hidden layer. The proposed method was trained and tested on a real resident travel diary dataset collected in Chengdu, China.
پیش بینی دقیق و قابل اعتماد از انتخاب حالت سفر فردی بسیار مهم است برای توسعه سیستم های حمل و نقل شهری چند حالته ، انجام برنامه ریزی حمل و نقل و تدوین استراتژی های مدیریت تقاضای ترافیک. مدلهای انتخاب سنتی گسسته بر روشهای مدل سازی حاکم شده اند دهه ها هنوز از فرضیات مدل سخت و دقت پیش بینی کم رنج می برند. در سالهای اخیر ، مدل های یادگیری ماشین (ML) مانند شبکه های عصبی و مدل های تقویت کننده ، توسط محققان برای انتخاب حالت سفر به طور گسترده ای استفاده می شود پیش بینی و نتایج امیدوار کننده ای داشته است.با این حال ، با وجود برتر عملکرد پیش بینی ، بدنه بزرگی از روشهای ML ، به ویژه شاخه از مدل های شبکه عصبی نیز با استفاده از مدل بیش از حد و خسته کننده محدود است فرآیند تعیین ساختار.برای برطرف کردن این شکاف ، این مطالعه پیشنهاد می کند Perceptron چند لایه پیشرفته (MLP ؛ یک شبکه عصبی) با دو لایه پنهان برای پیش بینی انتخاب حالت سفر ؛این MLP توسط XGBOOST تقویت می شود (تقویت کننده روش) برای انتخاب ویژگی و یک روش جستجوی شبکه برای بهینه پنهان تعیین نورون هر لایه پنهان.روش پیشنهادی آموزش داده شد و آزمایش شده بر روی یک مجموعه داده خاطرات مسافرتی ساکن واقعی که در چنگدو ، چین جمع آوری شده است.
29,760
This paper focuses on estimating the coefficients and average partial effects of observed regressors in nonlinear panel data models with interactive fixed effects, using the common correlated effects (CCE) framework. The proposed two-step estimation method involves applying principal component analysis to estimate latent factors based on cross-sectional averages of the regressors in the first step, and jointly estimating the coefficients of the regressors and factor loadings in the second step. The asymptotic distributions of the proposed estimators are derived under general conditions, assuming that the number of time-series observations is comparable to the number of cross-sectional observations. To correct for asymptotic biases of the estimators, we introduce both analytical and split-panel jackknife methods, and confirm their good performance in finite samples using Monte Carlo simulations. An empirical application utilizes the proposed method to study the arbitrage behaviour of nonfinancial firms across different security markets.
این مقاله بر برآورد ضرایب و میانگین اثرات جزئی متمرکز شده است رگرسیونرهای مشاهده شده در مدلهای داده پانل غیرخطی با ثابت تعاملی اثرات ، با استفاده از چارچوب اثرات همبسته مشترک (CCE).پیشنهاد روش تخمین دو مرحله شامل استفاده از تجزیه و تحلیل مؤلفه اصلی در تخمین عوامل نهفته بر اساس میانگین مقطعی از رگرسیون در اولین قدم ، و به طور مشترک ضرایب رگرسیون ها و بارگذاری عاملی در مرحله دوم.توزیع بدون علامت از برآوردگرهای پیشنهادی تحت شرایط عمومی با فرض اینکه تعداد مشاهدات سری زمانی با تعداد قابل مقایسه است مشاهدات مقطعی.برای تصحیح تعصبات بدون علامت برآوردگرها ، ما هر دو روش تحلیلی و تقسیم پانل Jackknife را معرفی می کنیم ، و عملکرد خوب آنها را در نمونه های محدود با استفاده از شبیه سازی مونت کارلو تأیید کنید. یک برنامه تجربی از روش پیشنهادی برای مطالعه داوری استفاده می کند رفتار بنگاه های غیر مالی در بازارهای امنیتی مختلف.
29,761
This paper studies the estimation of characteristic-based quantile factor models where the factor loadings are unknown functions of observed individual characteristics while the idiosyncratic error terms are subject to conditional quantile restrictions. We propose a three-stage estimation procedure that is easily implementable in practice and has nice properties. The convergence rates, the limiting distributions of the estimated factors and loading functions, and a consistent selection criterion for the number of factors at each quantile are derived under general conditions. The proposed estimation methodology is shown to work satisfactorily when: (i) the idiosyncratic errors have heavy tails, (ii) the time dimension of the panel dataset is not large, and (iii) the number of factors exceeds the number of characteristics. Finite sample simulations and an empirical application aimed at estimating the loading functions of the daily returns of a large panel of S\&P500 index securities help illustrate these properties.
در این مقاله برآورد فاکتور کمی مبتنی بر مشخصه بررسی شده است مدلهایی که بارگذاری عاملی توابع ناشناخته فرد مشاهده شده است خصوصیات در حالی که اصطلاحات خطای ایدیوسیکاتیک مشروط است محدودیت های کمی.ما یک روش تخمین سه مرحله ای را پیشنهاد می کنیم به راحتی در عمل قابل اجرا است و دارای خواص خوبی است.همگرایی نرخ ، توزیع محدود کننده عوامل تخمین زده شده و بارگیری توابع ، و یک معیار انتخاب مداوم برای تعداد عوامل در هر مقدار در شرایط کلی به دست می آید.برآورد پیشنهادی روش شناسی نشان داده می شود که به طور رضایت بخش کار می کند وقتی: (i) خطاهای ایدیوسیکاتیک دمهای سنگین داشته باشید ، (ب) ابعاد زمان مجموعه داده پانل بزرگ نیست ، و (iii) تعداد عوامل از تعداد مشخصات فراتر می رود.محدود، فانی شبیه سازی نمونه و یک برنامه تجربی با هدف تخمین بارگذاری توابع بازده روزانه یک صفحه بزرگ از اوراق بهادار شاخص S \ & p500 به نشان دادن این خصوصیات کمک کنید.
29,762
This paper studies difference-in-differences (DiD) setups with repeated cross-sectional data and potential compositional changes across time periods. We begin our analysis by deriving the efficient influence function and the semiparametric efficiency bound for the average treatment effect on the treated (ATT). We introduce nonparametric estimators that attain the semiparametric efficiency bound under mild rate conditions on the estimators of the nuisance functions, exhibiting a type of rate doubly-robust (DR) property. Additionally, we document a trade-off related to compositional changes: We derive the asymptotic bias of DR DiD estimators that erroneously exclude compositional changes and the efficiency loss when one fails to correctly rule out compositional changes. We propose a nonparametric Hausman-type test for compositional changes based on these trade-offs. The finite sample performance of the proposed DiD tools is evaluated through Monte Carlo experiments and an empirical application. As a by-product of our analysis, we present a new uniform stochastic expansion of the local polynomial multinomial logit estimator, which may be of independent interest.
این مقاله به بررسی تفاوتهای مختلف (DID) با تکرار می پردازد داده های مقطعی و تغییرات احتمالی ترکیبی در طول دوره های زمانی. ما تجزیه و تحلیل خود را با استخراج عملکرد تأثیر کارآمد و راندمان نیمهرامتری محدود به میانگین اثر درمانی بر روی درمان شده (att).ما برآوردگرهای غیرپارامتری را معرفی می کنیم که به نیمهرامتری می رسند بهره وری که تحت شرایط نرخ خفیف در برآوردگرهای مزاحمت قرار می گیرد توابع ، نشان دادن نوعی از نرخ خاصیت دو برابر-Robust (DR).علاوه بر این ، ما یک تجارت را در رابطه با تغییرات ترکیبی مستند می کنیم: ما نتیجه می گیریم تعصب بدون علامت از DR برآوردگرانی انجام داد که به اشتباه ترکیبی را حذف می کنند تغییرات و از دست دادن کارایی هنگامی که فرد نتواند به درستی رد شود تغییرات ترکیبی.ما یک آزمون غیرپارامتری از نوع را برای تغییرات ترکیبی بر اساس این معاملات.عملکرد نمونه محدود از ابزارهای پیشنهادی DID از طریق آزمایش های مونت کارلو و یک مورد ارزیابی می شود کاربرد تجربیما به عنوان یک محصول جانبی از تحلیل ما ، ما جدید ارائه می دهیم گسترش تصادفی یکنواخت از ورود چند جملهای چند جمله ای محلی برآوردگر ، که ممکن است مورد علاقه مستقل باشد.
29,763
Forecasting financial time series (FTS) is an essential field in finance and economics that anticipates market movements in financial markets. This paper investigates the accuracy of text mining and technical analyses in forecasting financial time series. It focuses on the S&P500 stock market index during the pandemic, which tracks the performance of the largest publicly traded companies in the US. The study compares two methods of forecasting the future price of the S&P500: text mining, which uses NLP techniques to extract meaningful insights from financial news, and technical analysis, which uses historical price and volume data to make predictions. The study examines the advantages and limitations of both methods and analyze their performance in predicting the S&P500. The FinBERT model outperforms other models in terms of S&P500 price prediction, as evidenced by its lower RMSE value, and has the potential to revolutionize financial analysis and prediction using financial news data. Keywords: ARIMA, BERT, FinBERT, Forecasting Financial Time Series, GARCH, LSTM, Technical Analysis, Text Mining JEL classifications: G4, C8
پیش بینی سری زمانی مالی (FTS) یک زمینه اساسی در امور مالی و اقتصاد که پیش بینی حرکت بازار در بازارهای مالی است.این کاغذ بررسی صحت استخراج متن و تجزیه و تحلیل فنی در پیش بینی سری زمانی مالی.در طول شاخص بازار سهام S&P500 متمرکز است پاندمی ، که عملکرد بزرگترین شرکت های معامله شده عمومی را دنبال می کند در امریکا.این مطالعه دو روش پیش بینی قیمت آینده را مقایسه می کند S&P500: Text Mining ، که از تکنیک های NLP برای استخراج معنی دار استفاده می کند بینش از اخبار مالی و تجزیه و تحلیل فنی ، که از تاریخی استفاده می کند داده های قیمت و حجم برای پیش بینی.این مطالعه به بررسی مزایای آن می پردازد و محدودیت های هر دو روش و تجزیه و تحلیل عملکرد آنها در پیش بینی S & P500.مدل فینبرت از نظر قیمت S&P500 از سایر مدل ها بهتر عمل می کند پیش بینی ، همانطور که از ارزش RMSE پایین آن مشهود است ، و این پتانسیل را دارد تحول و پیش بینی مالی با استفاده از داده های خبری مالی. کلمات کلیدی: Arima ، Bert ، Finbert ، پیش بینی سری زمانی مالی ، GARCH ، LSTM ، تجزیه و تحلیل فنی ، طبقه بندی Mining Mining طبقه بندی: G4 ، C8
29,764
In this paper, we study the estimation of the threshold predictive regression model with hybrid stochastic local unit root predictors. We demonstrate the estimation procedure and derive the asymptotic distribution of the least square estimator and the IV based estimator proposed by Magdalinos and Phillips (2009), under the null hypothesis of a diminishing threshold effect. Simulation experiments focus on the finite sample performance of our proposed estimators and the corresponding predictability tests as in Gonzalo and Pitarakis (2012), under the presence of threshold effects with stochastic local unit roots. An empirical application to stock return equity indices, illustrate the usefulness of our framework in uncovering regimes of predictability during certain periods. In particular, we focus on an aspect not previously examined in the predictability literature, that is, the effect of economic policy uncertainty.
در این مقاله ، ما تخمین رگرسیون پیش بینی آستانه را بررسی می کنیم مدل با پیش بینی کننده ریشه واحد محلی تصادفی ترکیبی.ما نشان می دهیم روش تخمین و توزیع بدون علامت کمترین مربع برآوردگر و برآوردگر مبتنی بر IV که توسط مگدالینوس و فیلیپس ارائه شده است (2009) ، تحت فرضیه تهی از یک آستانه کاهش دهنده.شبیه سازی آزمایش ها بر عملکرد نمونه محدود برآوردگرهای پیشنهادی ما تمرکز دارند و تست های پیش بینی مربوطه مانند گونزالو و پیتاراکیس (2012) ، تحت حضور اثرات آستانه با ریشه های واحد محلی تصادفی.در کاربرد تجربی به شاخص های سهام عدالت بازگرداندن سهام ، سودمندی را نشان می دهد از چارچوب ما در کشف رژیم های پیش بینی در طول خاص دوره هابه طور خاص ، ما روی جنبه ای تمرکز می کنیم که قبلاً در آن بررسی نشده است ادبیات پیش بینی ، یعنی تأثیر عدم اطمینان سیاست اقتصادی.
29,765
An input-output table is an important data for analyzing the economic situation of a region. Generally, the input-output table for each region (regional input-output table) in Japan is not always publicly available, so it is necessary to estimate the table. In particular, various methods have been developed for estimating input coefficients, which are an important part of the input-output table. Currently, non-survey methods are often used to estimate input coefficients because they require less data and computation, but these methods have some problems, such as discarding information and requiring additional data for estimation. In this study, the input coefficients are estimated by approximating the generation process with an artificial neural network (ANN) to mitigate the problems of the non-survey methods and to estimate the input coefficients with higher precision. To avoid over-fitting due to the small data used, data augmentation, called mixup, is introduced to increase the data size by generating virtual regions through region composition and scaling. By comparing the estimates of the input coefficients with those of Japan as a whole, it is shown that the accuracy of the method of this research is higher and more stable than that of the conventional non-survey methods. In addition, the estimated input coefficients for the three cities in Japan are generally close to the published values for each city.
جدول ورودی و خروجی یک داده مهم برای تجزیه و تحلیل اقتصادی است وضعیت یک منطقهبه طور کلی ، جدول ورودی و خروجی برای هر منطقه (جدول ورودی و خروجی منطقه ای) در ژاپن همیشه در دسترس عموم نیست ، بنابراین آن برای تخمین جدول لازم است.به طور خاص ، روش های مختلفی انجام شده است برای برآورد ضرایب ورودی ، که بخش مهمی از جدول ورودی-خروجی.در حال حاضر ، از روشهای غیر بررسی اغلب برای برآورد استفاده می شود ضرایب ورودی زیرا به داده ها و محاسبات کمتری نیاز دارند ، اما اینها روشها برخی از مشکلات ، مانند دور انداختن اطلاعات و نیاز به داده های اضافی برای تخمین. در این مطالعه ، ضرایب ورودی با تقریب تخمین زده می شود فرآیند تولید با یک شبکه عصبی مصنوعی (ANN) برای کاهش مشکلات روشهای غیر بررسی و تخمین ضرایب ورودی با دقت بالاتربرای جلوگیری از تناسب بیش از حد به دلیل داده های کوچک مورد استفاده ، داده ها افزودن ، به نام مخلوط ، برای افزایش اندازه داده ها توسط معرفی می شود تولید مناطق مجازی از طریق ترکیب و مقیاس منطقه. با مقایسه برآورد ضرایب ورودی با موارد ژاپن به عنوان a در کل ، نشان داده شده است که صحت روش این تحقیق بیشتر است و پایدارتر از روشهای غیر بررسی معمولی.علاوه بر این، ضرایب ورودی تخمین زده شده برای سه شهر ژاپن به طور کلی است نزدیک به ارزشهای منتشر شده برای هر شهر.
29,766
We consider the problem of extrapolating treatment effects across heterogeneous populations (``sites"/``contexts"). We consider an idealized scenario in which the researcher observes cross-sectional data for a large number of units across several ``experimental" sites in which an intervention has already been implemented to a new ``target" site for which a baseline survey of unit-specific, pre-treatment outcomes and relevant attributes is available. We propose a transfer estimator that exploits cross-sectional variation between individuals and sites to predict treatment outcomes using baseline outcome data for the target location. We consider the problem of obtaining a predictor of conditional average treatment effects at the target site that is MSE optimal within a certain class and subject to data constraints. Our approach is design-based in the sense that the performance of the predictor is evaluated given the specific, finite selection of experimental and target sites. Our approach is nonparametric, and our formal results concern the construction of an optimal basis of predictors as well as convergence rates for the estimated conditional average treatment effect relative to the constrained-optimal population predictor for the target site. We illustrate our approach using a combined data set of five multi-site randomized controlled trials (RCTs) to evaluate the effect of conditional cash transfers on school attendance.
ما مشکل برون یابی اثرات درمانی را در آن در نظر می گیریم جمعیت ناهمگن (`` سایت ها "/` `زمینه").ما ایده آل را در نظر می گیریم سناریویی که محقق در آن داده های مقطعی را برای یک بزرگ مشاهده می کند تعداد واحدها در چندین سایت `` آزمایش "که در آن یک مداخله قبلاً در یک سایت جدید "هدف" اجرا شده است که یک پایه برای آن بررسی نتایج خاص واحد ، قبل از درمان و ویژگی های مربوطه است در دسترس.ما یک برآوردگر انتقال پیشنهاد می کنیم که از مقطعی سوء استفاده می کند تنوع بین افراد و سایتها برای پیش بینی نتایج درمانی با استفاده از داده های نتیجه اولیه برای مکان هدف.ما مشکل را در نظر می گیریم به دست آوردن پیش بینی کننده اثرات درمانی متوسط ​​مشروط در هدف سایتی که MSE بهینه در یک کلاس خاص و در معرض داده ها قرار دارد محدودیت ها.رویکرد ما مبتنی بر طراحی است به این معنا که عملکرد پیش بینی کننده با توجه به انتخاب خاص و محدود تجربی ارزیابی می شود و سایت های هدفرویکرد ما غیرپارامتری است و نتایج رسمی ما نگران است ساخت پایه بهینه پیش بینی کننده ها و همچنین نرخ همگرایی برای تخمین میانگین اثر درمانی مشروط نسبت به پیش بینی کننده جمعیت بهینه محدود برای سایت هدف.ما خود را نشان می دهیم رویکرد با استفاده از یک مجموعه داده ترکیبی از پنج کنترل شده تصادفی چند سایت محاکمات (RCT) برای ارزیابی تأثیر نقل و انتقالات نقدی مشروط بر مدرسه حضور.
29,767
In this paper, we develop a novel large volatility matrix estimation procedure for analyzing global financial markets. Practitioners often use lower-frequency data, such as weekly or monthly returns, to address the issue of different trading hours in the international financial market. However, this approach can lead to inefficiency due to information loss. To mitigate this problem, our proposed method, called Structured Principal Orthogonal complEment Thresholding (Structured-POET), incorporates observation structural information for both global and national factor models. We establish the asymptotic properties of the Structured-POET estimator, and also demonstrate the drawbacks of conventional covariance matrix estimation procedures when using lower-frequency data. Finally, we apply the Structured-POET estimator to an out-of-sample portfolio allocation study using international stock market data.
در این مقاله ، ما یک تخمین ماتریس نوسانات بزرگ رمان ایجاد می کنیم رویه ای برای تجزیه و تحلیل بازارهای مالی جهانی.پزشکان اغلب استفاده می کنند داده های با فرکانس پایین ، مانند بازده هفتگی یا ماهانه ، برای پرداختن به این مسئله ساعت معاملاتی مختلف در بازار مالی بین المللی.با این حال ، این رویکرد می تواند به دلیل از دست دادن اطلاعات منجر به ناکارآمدی شود.برای کاهش این مشکل ، روش پیشنهادی ما ، به نام ساختار اصلی مکمل متعامد آستانه (-شاعر ساختاری) ، شامل اطلاعات ساختاری مشاهده است برای هر دو مدل فاکتور جهانی و ملی.ما بدون علامت ایجاد می کنیم خواص برآوردگر شاعرانه ساختار یافته ، و همچنین اشکالاتی را نشان می دهد از روشهای تخمین ماتریس کواریانس معمولی هنگام استفاده داده های فرکانس پایین.سرانجام ، ما برآوردگر شاعر ساختاری را در یک مطالعه تخصیص نمونه کارها خارج از نمونه با استفاده از داده های بین المللی بورس.