- September 12, 2020
• When Will the Covid-19 Pandemic Peak? (with Oliver Linton)
Journal of Econometrics, Volume 220, Issue 1, September 2020, Pages 130-157
– We carry out some analysis of the daily data on the number of new cases and the number of new deaths by (191) countries as reported to the European Centre for Disease Prevention and Control (ECDC). Our benchmark models are a quadratic model, a quartic model and a gamma model of time trends, which are applied to the log of new cases and deaths for each country. We use our model to predict when the peak of the epidemic will arise in terms of new cases or new deaths in each country and the peak level. We also predict how long the number of new daily cases in each country will fall by an order of magnitude. Finally, we forecast the total number of cases and deaths for each country. We also consider two models that link the joint evolution of new cases and new deaths.
• Dynamic Peer Groups of Arbitrage Characteristics (with Shuyi Ge and Oliver Linton)
Forthcoming, Journal of Business & Economic Statistics
Keywords: Semiparametric; Characteristics-based; Asset pricing; Power-enhanced test;
– We propose an asset pricing factor model constructed with semi-parametric characteristics-based mispricing and factor loading functions. We approximate the unknown functions by B-splines sieve where the number of B-splines coefficients is diverging. We estimate this model and test the existence of the mispricing function by a power-enhanced hypothesis test. The enhanced test solves the low power problem caused by diverging B-spline coefficients, with the strengthened power approaches to one asymptotically. We also investigate the structure of mispricing components through Hierarchical K-means Clusterings. We apply our methodology to CRSP (Center for Research in Security Prices) and Compustat data for the US stock market with one-year rolling windows during 1967-2017. This empirical study shows the presence of mispricing functions in certain time blocks. We also find that distinct clusters of the same characteristics lead to similar arbitrage returns, forming a “peer group” of arbitrage characteristics.
• News-Implied Linkages and Local Dependency in the Equity Market (with Shuyi Ge and Oliver Linton)
Forthcoming, Journal of Econometrics
Keywords: Spatial asset pricing model, weak and strong cross-sectional dependence, local dependency, networks, textual analysis, big data, large heterogeneous panel
– This paper studies a heterogeneous coefficient spatial factor model that separately addresses both common factor risks (strong cross-sectional dependence) and local dependency (weak cross-sectional dependence) in the equity returns. From the asset pricing perspective, we derive the theoretical implications of no asymptotic arbitrage for the heterogeneous spatial factor model. In empirical work, it is challenging to measure granular firm-to-firm connectivity for a high-dimensional panel of equity returns. We use extensive business news to construct firms’ links via which local shocks transmit, and we use those news-implied linkages as a proxy for the connectivity among firms. Empirically, we document a considerable degree of local dependency among S&P 500 stocks, and the spatial component does a great job in capturing the remaining correlations in the de-factored returns. We find that adding spatial interactions to factor models reduces mispricing and mean-squared errors. We also show that our news-implied linkages provide a comprehensive and integrated proxy for firm-to-firm connectivity, and it out-performs other existing networks in the literature.
Augment Large Covariance Matrix Estimation with Auxiliary Information (with Shuyi Ge, Oliver Linton, and Weiguang Liu)
–To estimate a large covariance matrix is a challenging job. Suppose we have identified a network G among cross-sections from auxiliary information such as network data. We propose a linear projection method to incorporate such information in the estimation of the large covariance matrix to improve efficiency. The simulation shows improvement in both Frobenius Norm and Matrix 1-norm over the linear shrinkage method, sample covariance matrix, and thresholding estimator.
Topic: A Semiparametric Characteristics-based Factor Model
A Dynamic Semiparametric Characteristics-based Model for Optimal Portfolio Selection (with Gregory Connor and Oliver Linton)
Keywords: Portfolio Management; Single index; GMM;
– This paper develops a two-step semiparametric methodology for portfolio weight selection for characteristics-based factor-tilt and factor-timing investment strategies. We build upon the expected utility maximization framework of Brandt (1999) and Aït-sahalia and Brandt (2001). We assume that assets’ returns obey a characteristics-based factor model with time-varying factor risk premia as in Li and Linton (2020). We prove under our return-generating assumptions that in a market with a large number of assets, an approximately optimal portfolio can be established using a two-step procedure. The first step finds optimal factor-mimicking sub-portfolios using a quadratic objective function over linear combinations of characteristics-based factor loadings. The second step dynamically combines these factor-mimicking sub-portfolios based on a time-varying signal, using the investor’s expected utility as the objective function. We develop and implement a two-stage semiparametric estimator. We apply it to CRSP (Center for Research in Security Prices) and FRED (Federal Reserve Economic Data) data and find excellent in-sample and out-sample performance that are consistent with investors’ risk aversion levels.
Specification-Lasso and An Application in Financial Markets (with Chaohua Dong)
Keywords: Interactive; Lasso; Variable selection; Model specification;
– This paper proposes the method of Specification-Lasso in a flexible semi-parametric regression model that allows for the interactive effects between different covariates. Specification-Lasso extends Lasso and Adaptive Group Lasso to achieve both relevant variable selection and model specification. Specification-Lasso also gives preliminary estimates that facilitate the estimation of the regression model. Monte Carlo simulations show that the Specification-Lasso can accurately specify partially linear additive models with interactive effects. Finally, the proposed methods are applied in an empirical study, which examines the topic proposed by Freyberger, Neuhierl and Weber (2017), arguing that firms’ sizes may have interactive effects with other security-specific characteristics, which can explain the stocks’ excess returns together.
Working in Progress
Generalized EGARCH Model: Factor-EGARCH (with Shuyi Ge and Weiguang Liu)
–One interesting stylized fact of the financial market is although firms’ idiosyncratic returns share very little common variation, their idiosyncratic volatilities tend to move together. While the volatility clustering in time has been widely explored, the volatility clustering in cross-sections has been less touched until recently. In this paper, we propose a Generalized EGARCH model with multiplicative common volatility factor, which aims to capture volatility clustering in both time and cross-sectional dimensions.