首页 > 最新文献

arXiv - ECON - Econometrics最新文献

英文 中文
BayesSRW: Bayesian Sampling and Re-weighting approach for variance reduction BayesSRW:减少方差的贝叶斯取样和再加权方法
Pub Date : 2024-08-28 DOI: arxiv-2408.15454
Carol Liu
In this paper, we address the challenge of sampling in scenarios wherelimited resources prevent exhaustive measurement across all subjects. Weconsider a setting where samples are drawn from multiple groups, each followinga distribution with unknown mean and variance parameters. We introduce a novelsampling strategy, motivated simply by Cauchy-Schwarz inequality, whichminimizes the variance of the population mean estimator by allocating samplesproportionally to both the group size and the standard deviation. This approachimproves the efficiency of sampling by focusing resources on groups withgreater variability, thereby enhancing the precision of the overall estimate.Additionally, we extend our method to a two-stage sampling procedure in a Bayesapproach, named BayesSRW, where a preliminary stage is used to estimate thevariance, which then informs the optimal allocation of the remaining samplingbudget. Through simulation examples, we demonstrate the effectiveness of ourapproach in reducing estimation uncertainty and providing more reliableinsights in applications ranging from user experience surveys tohigh-dimensional peptide array studies.
在本文中,我们探讨了在资源有限、无法对所有研究对象进行详尽测量的情况下进行抽样的难题。我们考虑了从多个组中抽取样本的情况,每个组都遵循一个具有未知均值和方差参数的分布。我们引入了一种新颖的抽样策略,这种策略的动机很简单,就是考希-施瓦茨不等式,它通过按组别大小和标准差成比例地分配样本,最大限度地减小了群体均值估计值的方差。此外,我们还将方法扩展到贝叶斯方法中的两阶段抽样程序,即贝叶斯 SRW,其中初步阶段用于估计方差,然后为剩余抽样预算的最优分配提供信息。通过模拟示例,我们展示了我们的方法在减少估计不确定性和提供更可靠洞察力方面的有效性,应用范围从用户体验调查到高维肽阵列研究。
{"title":"BayesSRW: Bayesian Sampling and Re-weighting approach for variance reduction","authors":"Carol Liu","doi":"arxiv-2408.15454","DOIUrl":"https://doi.org/arxiv-2408.15454","url":null,"abstract":"In this paper, we address the challenge of sampling in scenarios where\u0000limited resources prevent exhaustive measurement across all subjects. We\u0000consider a setting where samples are drawn from multiple groups, each following\u0000a distribution with unknown mean and variance parameters. We introduce a novel\u0000sampling strategy, motivated simply by Cauchy-Schwarz inequality, which\u0000minimizes the variance of the population mean estimator by allocating samples\u0000proportionally to both the group size and the standard deviation. This approach\u0000improves the efficiency of sampling by focusing resources on groups with\u0000greater variability, thereby enhancing the precision of the overall estimate.\u0000Additionally, we extend our method to a two-stage sampling procedure in a Bayes\u0000approach, named BayesSRW, where a preliminary stage is used to estimate the\u0000variance, which then informs the optimal allocation of the remaining sampling\u0000budget. Through simulation examples, we demonstrate the effectiveness of our\u0000approach in reducing estimation uncertainty and providing more reliable\u0000insights in applications ranging from user experience surveys to\u0000high-dimensional peptide array studies.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"67 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
The effects of data preprocessing on probability of default model fairness 数据预处理对违约概率模型公平性的影响
Pub Date : 2024-08-28 DOI: arxiv-2408.15452
Di Wu
In the context of financial credit risk evaluation, the fairness of machinelearning models has become a critical concern, especially given the potentialfor biased predictions that disproportionately affect certain demographicgroups. This study investigates the impact of data preprocessing, with aspecific focus on Truncated Singular Value Decomposition (SVD), on the fairnessand performance of probability of default models. Using a comprehensive datasetsourced from Kaggle, various preprocessing techniques, including SVD, wereapplied to assess their effect on model accuracy, discriminatory power, andfairness.
在金融信用风险评估方面,机器学习模型的公平性已成为一个重要问题,特别是考虑到有可能出现偏差的预测,对某些人口群体造成不成比例的影响。本研究探讨了数据预处理对违约概率模型的公平性和性能的影响,重点是截断奇异值分解(SVD)。利用从 Kaggle 获取的综合数据集,应用了包括 SVD 在内的各种预处理技术,以评估它们对模型准确性、判别力和公平性的影响。
{"title":"The effects of data preprocessing on probability of default model fairness","authors":"Di Wu","doi":"arxiv-2408.15452","DOIUrl":"https://doi.org/arxiv-2408.15452","url":null,"abstract":"In the context of financial credit risk evaluation, the fairness of machine\u0000learning models has become a critical concern, especially given the potential\u0000for biased predictions that disproportionately affect certain demographic\u0000groups. This study investigates the impact of data preprocessing, with a\u0000specific focus on Truncated Singular Value Decomposition (SVD), on the fairness\u0000and performance of probability of default models. Using a comprehensive dataset\u0000sourced from Kaggle, various preprocessing techniques, including SVD, were\u0000applied to assess their effect on model accuracy, discriminatory power, and\u0000fairness.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Endogenous Treatment Models with Social Interactions: An Application to the Impact of Exercise on Self-Esteem 具有社会互动的内生治疗模型:应用于运动对自尊的影响
Pub Date : 2024-08-26 DOI: arxiv-2408.13971
Zhongjian Lin, Francis Vella
We address the estimation of endogenous treatment models with socialinteractions in both the treatment and outcome equations. We model theinteractions between individuals in an internally consistent manner via a gametheoretic approach based on discrete Bayesian games. This introduces asubstantial computational burden in estimation which we address through asequential version of the nested fixed point algorithm. We also provide somerelevant treatment effects, and procedures for their estimation, which capturethe impact on both the individual and the total sample. Our empiricalapplication examines the impact of an individual's exercise frequency on herlevel of self-esteem. We find that an individual's exercise frequency isinfluenced by her expectation of her friends'. We also find that anindividual's level of self-esteem is affected by her level of exercise and, atrelatively lower levels of self-esteem, by the expectation of her friends'self-esteem.
我们探讨了在治疗方程和结果方程中都存在社会互动的内生治疗模型的估计问题。我们通过基于离散贝叶斯博弈的博弈论方法,以内部一致的方式对个体间的相互作用进行建模。这给估计带来了巨大的计算负担,我们通过嵌套定点算法的后续版本解决了这一问题。我们还提供了一些相关的处理效应及其估算程序,以捕捉对个体和总体样本的影响。我们的实证应用研究了个人锻炼频率对其自尊水平的影响。我们发现,一个人的运动频率会受到她对朋友的期望值的影响。我们还发现,个体的自尊水平受其运动频率的影响,而在自尊水平相对较低的情况下,则受其对朋友自尊的期望的影响。
{"title":"Endogenous Treatment Models with Social Interactions: An Application to the Impact of Exercise on Self-Esteem","authors":"Zhongjian Lin, Francis Vella","doi":"arxiv-2408.13971","DOIUrl":"https://doi.org/arxiv-2408.13971","url":null,"abstract":"We address the estimation of endogenous treatment models with social\u0000interactions in both the treatment and outcome equations. We model the\u0000interactions between individuals in an internally consistent manner via a game\u0000theoretic approach based on discrete Bayesian games. This introduces a\u0000substantial computational burden in estimation which we address through a\u0000sequential version of the nested fixed point algorithm. We also provide some\u0000relevant treatment effects, and procedures for their estimation, which capture\u0000the impact on both the individual and the total sample. Our empirical\u0000application examines the impact of an individual's exercise frequency on her\u0000level of self-esteem. We find that an individual's exercise frequency is\u0000influenced by her expectation of her friends'. We also find that an\u0000individual's level of self-esteem is affected by her level of exercise and, at\u0000relatively lower levels of self-esteem, by the expectation of her friends'\u0000self-esteem.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Modeling the Dynamics of Growth in Master-Planned Communities 总体规划社区的增长动态建模
Pub Date : 2024-08-26 DOI: arxiv-2408.14214
Christopher K. Allsup, Irene S. Gabashvili
This paper describes how a time-varying Markov model was used to forecasthousing development at a master-planned community during a transition from highto low growth. Our approach draws on detailed historical data to model thedynamics of the market participants, producing results that are entirelydata-driven and free of bias. While traditional time series forecasting methodsoften struggle to account for nonlinear regime changes in growth, our approachsuccessfully captures the onset of buildout as well as external economicshocks, such as the 1990 and 2008-2011 recessions and the 2021 post-pandemicboom. This research serves as a valuable tool for urban planners, homeownerassociations, and property stakeholders aiming to navigate the complexities ofgrowth at master-planned communities during periods of both system stabilityand instability.
本文介绍了如何利用时变马尔可夫模型来预测一个总体规划社区从高速增长向低速增长过渡期间的住房开发情况。我们的方法利用了详细的历史数据来模拟市场参与者的动态变化,得出的结果完全由数据驱动,没有任何偏差。传统的时间序列预测方法往往难以解释增长中的非线性制度变化,而我们的方法成功地捕捉到了增长的开始以及外部经济冲击,如 1990 年和 2008-2011 年的经济衰退以及 2021 年大流行后的经济繁荣。这项研究为城市规划者、房主协会和房地产利益相关者提供了一个宝贵的工具,帮助他们在系统稳定和不稳定时期驾驭总体规划社区增长的复杂性。
{"title":"Modeling the Dynamics of Growth in Master-Planned Communities","authors":"Christopher K. Allsup, Irene S. Gabashvili","doi":"arxiv-2408.14214","DOIUrl":"https://doi.org/arxiv-2408.14214","url":null,"abstract":"This paper describes how a time-varying Markov model was used to forecast\u0000housing development at a master-planned community during a transition from high\u0000to low growth. Our approach draws on detailed historical data to model the\u0000dynamics of the market participants, producing results that are entirely\u0000data-driven and free of bias. While traditional time series forecasting methods\u0000often struggle to account for nonlinear regime changes in growth, our approach\u0000successfully captures the onset of buildout as well as external economic\u0000shocks, such as the 1990 and 2008-2011 recessions and the 2021 post-pandemic\u0000boom. This research serves as a valuable tool for urban planners, homeowner\u0000associations, and property stakeholders aiming to navigate the complexities of\u0000growth at master-planned communities during periods of both system stability\u0000and instability.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184153","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Double/Debiased CoCoLASSO of Treatment Effects with Mismeasured High-Dimensional Control Variables 误测高维控制变量治疗效果的双重/偏差 CoCoLASSO
Pub Date : 2024-08-26 DOI: arxiv-2408.14671
Geonwoo Kim, Suyong Song
We develop an estimator for treatment effects in high-dimensional settingswith additive measurement error, a prevalent challenge in modern econometrics.We introduce the Double/Debiased Convex Conditioned LASSO (Double/DebiasedCoCoLASSO), which extends the double/debiased machine learning framework toaccommodate mismeasured covariates. Our principal contributions are threefold.(1) We construct a Neyman-orthogonal score function that remains valid undermeasurement error, incorporating a bias correction term to account forerror-induced correlations. (2) We propose a method of moments estimator forthe measurement error variance, enabling implementation without prior knowledgeof the error covariance structure. (3) We establish the $sqrt{N}$-consistencyand asymptotic normality of our estimator under general conditions, allowingfor both the number of covariates and the magnitude of measurement error toincrease with the sample size. Our theoretical results demonstrate theestimator's efficiency within the class of regularized high-dimensionalestimators accounting for measurement error. Monte Carlo simulationscorroborate our asymptotic theory and illustrate the estimator's robustperformance across various levels of measurement error. Notably, ourcovariance-oblivious approach nearly matches the efficiency of methods thatassume known error variance.
我们提出了双/去偏凸条件 LASSO(Double/DebiasedCoCoLASSO),它扩展了双/去偏机器学习框架,以适应测量误差协变量。我们的主要贡献有三点:(1) 我们构建了一个在测量误差条件下仍然有效的奈曼正交得分函数,并加入了一个偏差修正项,以考虑误差引起的相关性。(2) 我们提出了测量误差方差的矩估计方法,无需事先了解误差协方差结构即可实施。(3) 在一般条件下,允许协变量的数量和测量误差的大小随着样本量的增加而增加,我们建立了估计器的($sqrt{N}$)一致性和渐近正态性。我们的理论结果证明了该估计器在考虑测量误差的正则化高维估计器类别中的效率。蒙特卡罗模拟证实了我们的渐近理论,并说明了估计器在各种测量误差水平下的稳健表现。值得注意的是,我们的不考虑方差的方法几乎可以与假定已知误差方差的方法的效率相媲美。
{"title":"Double/Debiased CoCoLASSO of Treatment Effects with Mismeasured High-Dimensional Control Variables","authors":"Geonwoo Kim, Suyong Song","doi":"arxiv-2408.14671","DOIUrl":"https://doi.org/arxiv-2408.14671","url":null,"abstract":"We develop an estimator for treatment effects in high-dimensional settings\u0000with additive measurement error, a prevalent challenge in modern econometrics.\u0000We introduce the Double/Debiased Convex Conditioned LASSO (Double/Debiased\u0000CoCoLASSO), which extends the double/debiased machine learning framework to\u0000accommodate mismeasured covariates. Our principal contributions are threefold.\u0000(1) We construct a Neyman-orthogonal score function that remains valid under\u0000measurement error, incorporating a bias correction term to account for\u0000error-induced correlations. (2) We propose a method of moments estimator for\u0000the measurement error variance, enabling implementation without prior knowledge\u0000of the error covariance structure. (3) We establish the $sqrt{N}$-consistency\u0000and asymptotic normality of our estimator under general conditions, allowing\u0000for both the number of covariates and the magnitude of measurement error to\u0000increase with the sample size. Our theoretical results demonstrate the\u0000estimator's efficiency within the class of regularized high-dimensional\u0000estimators accounting for measurement error. Monte Carlo simulations\u0000corroborate our asymptotic theory and illustrate the estimator's robust\u0000performance across various levels of measurement error. Notably, our\u0000covariance-oblivious approach nearly matches the efficiency of methods that\u0000assume known error variance.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"67 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Inference on Consensus Ranking of Distributions 一致同意的分布排名推论
Pub Date : 2024-08-25 DOI: arxiv-2408.13949
David M. Kaplan
Instead of testing for unanimous agreement, I propose learning how broad of aconsensus favors one distribution over another (of earnings, productivity,asset returns, test scores, etc.). Specifically, given a sample from each oftwo distributions, I propose statistical inference methods to learn about theset of utility functions for which the first distribution has higher expectedutility than the second distribution. With high probability, an "inner"confidence set is contained within this true set, while an "outer" confidenceset contains the true set. Such confidence sets can be formed by inverting aproposed multiple testing procedure that controls the familywise error rate.Theoretical justification comes from empirical process results, given that verylarge classes of utility functions are generally Donsker (subject to finitemoments). The theory additionally justifies a uniform (over utility functions)confidence band of expected utility differences, as well as tests with autility-based "restricted stochastic dominance" as either the null oralternative hypothesis. Simulated and empirical examples illustrate themethodology.
我建议,与其测试是否存在一致意见,不如了解有多大程度的一致意见倾向于一种分布而非另一种分布(收入、生产率、资产回报、考试分数等)。具体来说,给定两种分布的样本,我提出了统计推断方法,以了解第一种分布的预期效用高于第二种分布的效用函数集。很有可能,"内部 "置信集包含在这个真实集合中,而 "外部 "置信集包含真实集合。这种置信集可以通过反转所提出的多重检验程序来形成,该程序可以控制全族误差率。理论依据来自于经验过程的结果,因为很大一类效用函数一般都是唐斯克函数(受有限矩影响)。此外,该理论还证明了预期效用差异的统一(效用函数)置信区间,以及基于自变量的 "受限随机支配 "作为零假设或口述替代假设的检验是合理的。模拟和实证例子说明了这一方法。
{"title":"Inference on Consensus Ranking of Distributions","authors":"David M. Kaplan","doi":"arxiv-2408.13949","DOIUrl":"https://doi.org/arxiv-2408.13949","url":null,"abstract":"Instead of testing for unanimous agreement, I propose learning how broad of a\u0000consensus favors one distribution over another (of earnings, productivity,\u0000asset returns, test scores, etc.). Specifically, given a sample from each of\u0000two distributions, I propose statistical inference methods to learn about the\u0000set of utility functions for which the first distribution has higher expected\u0000utility than the second distribution. With high probability, an \"inner\"\u0000confidence set is contained within this true set, while an \"outer\" confidence\u0000set contains the true set. Such confidence sets can be formed by inverting a\u0000proposed multiple testing procedure that controls the familywise error rate.\u0000Theoretical justification comes from empirical process results, given that very\u0000large classes of utility functions are generally Donsker (subject to finite\u0000moments). The theory additionally justifies a uniform (over utility functions)\u0000confidence band of expected utility differences, as well as tests with a\u0000utility-based \"restricted stochastic dominance\" as either the null or\u0000alternative hypothesis. Simulated and empirical examples illustrate the\u0000methodology.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"5 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184169","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Cross-sectional Dependence in Idiosyncratic Volatility 非同步波动性的横截面依赖性
Pub Date : 2024-08-24 DOI: arxiv-2408.13437
Ilze Kalnina, Kokouvi Tewou
This paper introduces an econometric framework for analyzing cross-sectionaldependence in the idiosyncratic volatilities of assets using high frequencydata. We first consider the estimation of standard measures of dependence inthe idiosyncratic volatilities such as covariances and correlations. Naiveestimators of these measures are biased due to the use of the error-ladenestimates of idiosyncratic volatilities. We provide bias-corrected estimatorsand the relevant asymptotic theory. Next, we introduce an idiosyncraticvolatility factor model, in which we decompose the variation in idiosyncraticvolatilities into two parts: the variation related to the systematic factorssuch as the market volatility, and the residual variation. Again, naiveestimators of the decomposition are biased, and we provide bias-correctedestimators. We also provide the asymptotic theory that allows us to testwhether the residual (non-systematic) components of the idiosyncraticvolatilities exhibit cross-sectional dependence. We apply our methodology tothe S&P 100 index constituents, and document strong cross-sectional dependencein their idiosyncratic volatilities. We consider two different sets ofidiosyncratic volatility factors, and find that neither can fully account forthe cross-sectional dependence in idiosyncratic volatilities. For each model,we map out the network of dependencies in residual (non-systematic)idiosyncratic volatilities across all stocks.
本文介绍了一种计量经济学框架,用于利用高频数据分析资产特异波动率的跨期依赖性。我们首先考虑了对特异波动率依赖性的标准度量的估计,如协方差和相关性。由于使用了带有误差的特异波动率估计值,这些指标的天真估计值是有偏差的。我们提供了偏差校正估计值和相关的渐近理论。接下来,我们引入一个特质波动率因子模型,将特质波动率的变化分解为两部分:与市场波动率等系统性因子相关的变化和残差变化。同样,分解的天真估计值是有偏差的,我们提供了偏差校正估计值。我们还提供了渐近理论,使我们能够检验特异性波动率的残差(非系统性)成分是否表现出横截面依赖性。我们将我们的方法应用于标准普尔 100 指数成分股,并记录了其特异性波动率中强烈的横截面依赖性。我们考虑了两组不同的特异波动率因子,发现这两组因子都不能完全解释特异波动率的横截面依赖性。对于每种模型,我们都绘制出了所有股票的残差(非系统性)特异波动率的依赖网络。
{"title":"Cross-sectional Dependence in Idiosyncratic Volatility","authors":"Ilze Kalnina, Kokouvi Tewou","doi":"arxiv-2408.13437","DOIUrl":"https://doi.org/arxiv-2408.13437","url":null,"abstract":"This paper introduces an econometric framework for analyzing cross-sectional\u0000dependence in the idiosyncratic volatilities of assets using high frequency\u0000data. We first consider the estimation of standard measures of dependence in\u0000the idiosyncratic volatilities such as covariances and correlations. Naive\u0000estimators of these measures are biased due to the use of the error-laden\u0000estimates of idiosyncratic volatilities. We provide bias-corrected estimators\u0000and the relevant asymptotic theory. Next, we introduce an idiosyncratic\u0000volatility factor model, in which we decompose the variation in idiosyncratic\u0000volatilities into two parts: the variation related to the systematic factors\u0000such as the market volatility, and the residual variation. Again, naive\u0000estimators of the decomposition are biased, and we provide bias-corrected\u0000estimators. We also provide the asymptotic theory that allows us to test\u0000whether the residual (non-systematic) components of the idiosyncratic\u0000volatilities exhibit cross-sectional dependence. We apply our methodology to\u0000the S&P 100 index constituents, and document strong cross-sectional dependence\u0000in their idiosyncratic volatilities. We consider two different sets of\u0000idiosyncratic volatility factors, and find that neither can fully account for\u0000the cross-sectional dependence in idiosyncratic volatilities. For each model,\u0000we map out the network of dependencies in residual (non-systematic)\u0000idiosyncratic volatilities across all stocks.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"8 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184173","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Machine Learning and the Yield Curve: Tree-Based Macroeconomic Regime Switching 机器学习与收益率曲线:基于树的宏观经济制度转换
Pub Date : 2024-08-23 DOI: arxiv-2408.12863
Siyu Bie, Francis X. Diebold, Jingyu He, Junye Li
We explore tree-based macroeconomic regime-switching in the context of thedynamic Nelson-Siegel (DNS) yield-curve model. In particular, we customize thetree-growing algorithm to partition macroeconomic variables based on the DNSmodel's marginal likelihood, thereby identifying regime-shifting patterns inthe yield curve. Compared to traditional Markov-switching models, our modeloffers clear economic interpretation via macroeconomic linkages and ensurescomputational simplicity. In an empirical application to U.S. Treasury bondyields, we find (1) important yield curve regime switching, and (2) evidencethat macroeconomic variables have predictive power for the yield curve when theshort rate is high, but not in other regimes, thereby refining the notion ofyield curve ``macro-spanning".
我们以动态 Nelson-Siegel(DNS)收益率曲线模型为背景,探讨了基于树的宏观经济制度转换。特别是,我们定制了树状生长算法,以根据 DNS 模型的边际似然率划分宏观经济变量,从而识别收益率曲线的制度转换模式。与传统的马尔可夫转换模型相比,我们的模型通过宏观经济联系提供了清晰的经济解释,并确保了计算的简便性。在对美国国债收益率的实证应用中,我们发现:(1)收益率曲线存在重要的制度转换;(2)有证据表明,当空头利率较高时,宏观经济变量对收益率曲线具有预测能力,但在其他制度下则没有,从而完善了收益率曲线 "宏观跨度 "的概念。
{"title":"Machine Learning and the Yield Curve: Tree-Based Macroeconomic Regime Switching","authors":"Siyu Bie, Francis X. Diebold, Jingyu He, Junye Li","doi":"arxiv-2408.12863","DOIUrl":"https://doi.org/arxiv-2408.12863","url":null,"abstract":"We explore tree-based macroeconomic regime-switching in the context of the\u0000dynamic Nelson-Siegel (DNS) yield-curve model. In particular, we customize the\u0000tree-growing algorithm to partition macroeconomic variables based on the DNS\u0000model's marginal likelihood, thereby identifying regime-shifting patterns in\u0000the yield curve. Compared to traditional Markov-switching models, our model\u0000offers clear economic interpretation via macroeconomic linkages and ensures\u0000computational simplicity. In an empirical application to U.S. Treasury bond\u0000yields, we find (1) important yield curve regime switching, and (2) evidence\u0000that macroeconomic variables have predictive power for the yield curve when the\u0000short rate is high, but not in other regimes, thereby refining the notion of\u0000yield curve ``macro-spanning\".","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"12 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184171","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Difference-in-differences with as few as two cross-sectional units -- A new perspective to the democracy-growth debate 只有两个横截面单位的差异 -- 民主与增长辩论的新视角
Pub Date : 2024-08-23 DOI: arxiv-2408.13047
Gilles Koumou, Emmanuel Selorm Tsyawo
Pooled panel analyses tend to mask heterogeneity in unit-specific treatmenteffects. For example, existing studies on the impact of democracy on economicgrowth do not reach a consensus as empirical findings are substantiallyheterogeneous in the country composition of the panel. In contrast to pooledpanel analyses, this paper proposes a Difference-in-Differences (DiD) estimatorthat exploits the temporal dimension in the data and estimates unit-specificaverage treatment effects on the treated (ATT) with as few as twocross-sectional units. Under weak identification and temporal dependenceconditions, the DiD estimator is asymptotically normal. The estimator isfurther complemented with a test of identification granted at least twocandidate control units. Empirical results using the DiD estimator suggestBenin's economy would have been 6.3% smaller on average over the 1993-2018period had she not democratised.
集合面板分析往往会掩盖特定单位治疗效果的异质性。例如,关于民主对经济增长影响的现有研究并没有达成共识,因为实证研究结果在面板的国家构成方面存在很大的异质性。与集合面板分析不同,本文提出了一种差分(DiD)估计方法,利用数据中的时间维度,估计特定单位对被治疗者的平均治疗效果(ATT),只需两个跨部门单位。在弱识别和时间依赖性条件下,DiD 估计器是渐近正态的。此外,该估计器还得到了至少两个候选控制单元的识别检验的补充。使用 DiD 估计器的实证结果表明,如果贝宁没有实现民主化,1993-2018 年期间的经济规模平均会缩小 6.3%。
{"title":"Difference-in-differences with as few as two cross-sectional units -- A new perspective to the democracy-growth debate","authors":"Gilles Koumou, Emmanuel Selorm Tsyawo","doi":"arxiv-2408.13047","DOIUrl":"https://doi.org/arxiv-2408.13047","url":null,"abstract":"Pooled panel analyses tend to mask heterogeneity in unit-specific treatment\u0000effects. For example, existing studies on the impact of democracy on economic\u0000growth do not reach a consensus as empirical findings are substantially\u0000heterogeneous in the country composition of the panel. In contrast to pooled\u0000panel analyses, this paper proposes a Difference-in-Differences (DiD) estimator\u0000that exploits the temporal dimension in the data and estimates unit-specific\u0000average treatment effects on the treated (ATT) with as few as two\u0000cross-sectional units. Under weak identification and temporal dependence\u0000conditions, the DiD estimator is asymptotically normal. The estimator is\u0000further complemented with a test of identification granted at least two\u0000candidate control units. Empirical results using the DiD estimator suggest\u0000Benin's economy would have been 6.3% smaller on average over the 1993-2018\u0000period had she not democratised.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Integrating an agent-based behavioral model in microtransit forecasting and revenue management 在微观交通预测和收入管理中整合基于代理的行为模型
Pub Date : 2024-08-22 DOI: arxiv-2408.12577
Xiyuan Ren, Joseph Y. J. Chow, Venktesh Pandey, Linfei Yuan
As an IT-enabled multi-passenger mobility service, microtransit has thepotential to improve accessibility, reduce congestion, and enhance flexibilityin transportation options. However, due to its heterogeneous impacts ondifferent communities and population segments, there is a need for better toolsin microtransit forecast and revenue management, especially when actual usagedata are limited. We propose a novel framework based on an agent-based mixedlogit model estimated with microtransit usage data and synthetic trip data. Theframework involves estimating a lower-branch mode choice model with synthetictrip data, combining lower-branch parameters with microtransit data to estimatean upper-branch ride pass subscription model, and applying the nested model toevaluate microtransit pricing and subsidy policies. The framework enablesfurther decision-support analysis to consider diverse travel patterns andheterogeneous tastes of the total population. We test the framework in a casestudy with synthetic trip data from Replica Inc. and microtransit data fromArlington Via. The lower-branch model result in a rho-square value of 0.603 onweekdays and 0.576 on weekends. Predictions made by the upper-branch modelclosely match the marginal subscription data. In a ride pass pricing policyscenario, we show that a discount in weekly pass (from $25 to $18.9) andmonthly pass (from $80 to $71.5) would surprisingly increase total revenue by$102/day. In an event- or place-based subsidy policy scenario, we show that a100% fare discount would reduce 80 car trips during peak hours at AT&T Stadium,requiring a subsidy of $32,068/year.
作为一种由信息技术支持的多乘客流动服务,微型公交具有改善交通可达性、减少拥堵和提高交通选择灵活性的潜力。然而,由于其对不同社区和人群的影响各不相同,因此需要更好的工具来进行微型公交预测和收入管理,尤其是在实际使用数据有限的情况下。我们提出了一个基于代理混合逻辑模型的新框架,该模型利用微型公交使用数据和合成行程数据进行估算。该框架包括利用合成行程数据估算下分支模式选择模型,将下分支参数与微型公交数据相结合以估算上分支乘车证订购模型,并应用嵌套模型评估微型公交定价和补贴政策。通过该框架可以进行进一步的决策支持分析,以考虑总人口的不同出行模式和异质性品味。我们利用 Replica 公司提供的合成出行数据和阿灵顿 Via 公司提供的微型公交数据进行了案例研究,对该框架进行了测试。下分支模型得出的工作日 rho-square 值为 0.603,周末为 0.576。上分支模型的预测结果与边际订购数据基本吻合。在乘车证定价政策情景下,我们发现周票(从 25 美元降至 18.9 美元)和月票(从 80 美元降至 71.5 美元)的折扣会使总收入出人意料地增加 102 美元/天。在基于活动或地点的补贴政策方案中,我们显示,100% 的票价折扣将减少 AT&T 体育馆高峰时段的 80 次汽车出行,每年需要补贴 32,068 美元。
{"title":"Integrating an agent-based behavioral model in microtransit forecasting and revenue management","authors":"Xiyuan Ren, Joseph Y. J. Chow, Venktesh Pandey, Linfei Yuan","doi":"arxiv-2408.12577","DOIUrl":"https://doi.org/arxiv-2408.12577","url":null,"abstract":"As an IT-enabled multi-passenger mobility service, microtransit has the\u0000potential to improve accessibility, reduce congestion, and enhance flexibility\u0000in transportation options. However, due to its heterogeneous impacts on\u0000different communities and population segments, there is a need for better tools\u0000in microtransit forecast and revenue management, especially when actual usage\u0000data are limited. We propose a novel framework based on an agent-based mixed\u0000logit model estimated with microtransit usage data and synthetic trip data. The\u0000framework involves estimating a lower-branch mode choice model with synthetic\u0000trip data, combining lower-branch parameters with microtransit data to estimate\u0000an upper-branch ride pass subscription model, and applying the nested model to\u0000evaluate microtransit pricing and subsidy policies. The framework enables\u0000further decision-support analysis to consider diverse travel patterns and\u0000heterogeneous tastes of the total population. We test the framework in a case\u0000study with synthetic trip data from Replica Inc. and microtransit data from\u0000Arlington Via. The lower-branch model result in a rho-square value of 0.603 on\u0000weekdays and 0.576 on weekends. Predictions made by the upper-branch model\u0000closely match the marginal subscription data. In a ride pass pricing policy\u0000scenario, we show that a discount in weekly pass (from $25 to $18.9) and\u0000monthly pass (from $80 to $71.5) would surprisingly increase total revenue by\u0000$102/day. In an event- or place-based subsidy policy scenario, we show that a\u0000100% fare discount would reduce 80 car trips during peak hours at AT&T Stadium,\u0000requiring a subsidy of $32,068/year.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"22 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142184172","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
期刊
arXiv - ECON - Econometrics
全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1