Testing Conditional Factor Models

Abstract

Using nonparametric techniques, we develop a methodology for estimating conditional alphas and betas and long-run alphas and betas, which are the averages of conditional alphas and betas, respectively, across time. The tests can be performed for a single asset or jointly across portfolios. The traditional Gibbons, Ross, and Shanken (1989) test arises as a special case of no time variation in the alphas and factor loadings and homoskedasticity. As applications of the methodology, we estimate conditional CAPM and multifactor models on book-to-market and momentum decile portfolios. We reject the null that long-run alphas are equal to zero even though there is substantial variation in the conditional factor loadings of these portfolios.

Growth options, macroeconomic conditions, and the cross section of credit risk

“…Empirically, the yearly value premium is between 6.29% and 12.55% when comparing the top and bottom book-to-market deciles (Fama and French, 2002;Patton and Timmermann, 2010;Ang and Kristensen, 2011 The value premium is higher in Panel B than in Panel A because the equity risk premium is an increasing and convex function of the leverage ratio. Hence, whenever the economy switches to recession in the dynamic simulation, the equity risk premium of value firms with an initially larger leverage increases, on average, more than the one of growth firms with an initially lower leverage.…”

Time-Varying Coefficient Estimation in SURE Models: Application to Portfolio Management

“…Recently, research in time-varying coefficients in nonparametric asset pricing models has been popular lately, see Ang and Kristensen (2012), Ferreira et al (2011), Esteban et al (2015) and Cai et al (2015) for further details. Time-varying coefficients account for the variations in the asset/portfolio sensitivity to the risk factors, along time.…”

“…Methodologically, we propose a kernel smoothing estimator for the above models, within the family of Orbe et al (2003Orbe et al ( , 2005; Cai (2007) ;Ferreira et al (2011); Ang and Kristensen (2012); Phillips et al (2017) among others. This approach enables us to relax the usual assumption of stationarity and consider locally stationary variables as defined in Dahlhaus (1997) and Dahlhaus (2000).…”

Nonparametric test for a constant beta between Itô semi-martingales based on high-frequency data

“…First, there is an enormous amount of literature on parameter shifts and breaks (Kejriwal et al, 2013, and references therein), but the results are all based on a long span ergodic-type theory rather than fixed length in-fill conducted here. Second, Ang and Kristensen (2012) propose a test for constant beta based on a Hausman type statistic that compares a nonparametric kernel-based estimate of betas at fixed time points and a long-run estimate of beta. Ang and Kristensen (2012) do not consider formally the role of the discretization error in their analysis.…”

“…Second, Ang and Kristensen (2012) propose a test for constant beta based on a Hausman type statistic that compares a nonparametric kernel-based estimate of betas at fixed time points and a long-run estimate of beta. Ang and Kristensen (2012) do not consider formally the role of the discretization error in their analysis. By contrast, we rely here solely on a fixed span and the associated in-fill or high-frequency asymptotics, and we are interested in checking whether beta is constant on the whole time interval, not only at fixed points in time.…”