Modeling the cross-section of stock returns using sensible models in a model pool

An increase in the number of asset pricing models intensifies model uncertainties in asset pricing. While a pure “model selection” (singling out a best model) can result in a loss of useful information, a full “model pooling” may increase the risk of including noisy information. We make a trade-off between the two methods and develop a new two-step trimming-then-pooling method to forecast the joint distributions of asset returns using a large pool of asset pricing models. Our method allows investors to focus on certain regions of the distributions. In the first step, we trim the uninformative models from a pool of candidates, and in the second step, we pool the forecasts of the surviving models. We find that our method significantly enhances portfolio performance and predicts downside risk precisely, and the improvements are mainly due to trimming. The pool of sensible models becomes larger when focusing on extreme events, responds rapidly to rising uncertainty, and reflects the magnitude of factor premiums. These findings provide new insights into asset pricing model evaluation.     

On Comparing Asset Pricing Models

Revisiting the framework of (Barillas, Francisco, and Jay Shanken, 2018, Comparing asset pricing models, The Journal of Finance 73, 715–754). BS henceforth, we show that the Bayesian marginal likelihood-based model comparison method in that paper is unsound : the priors on the nuisance parameters across models must satisfy a change of variable property for densities that is violated by the Jeffreys priors used in the BS method. Extensive simulation exercises confirm that the BS method performs unsatisfactorily. We derive a new class of improper priors on the nuisance parameters, starting from a single improper prior, which leads to valid marginal likelihoods and model comparisons. The performance of our marginal likelihoods is significantly better, allowing for reliable Bayesian work on which factors are risk factors in asset pricing models.     

Assessing Specification Errors in Stochastic Discount Factor Models

In this article we develop alternative ways to compare asset pricing models when it is understood that their implied stochastic discount factors do not price all portfolios correctly. Unlike comparisons based on ${\chi }^2$ statistics associated with null hypotheses that models are correct, our measures of model performance do not reward variability of discount factor proxies. One of our measures is designed to exploit fully the implications of arbitrage-free pricing of derivative claims. We demonstrate empirically the usefulness of our methods in assessing some alternative stochastic factor models that have been proposed in asset pricing literature.     

Option pricing under regime switching

Abstract

This paper develops a family of option pricing models when the underlying stock price dynamic is modelled by a regime switching process in which prices remain in one volatility regime for a random amount of time before switching over into a new regime. Our family includes the regime switching models of Hamilton (Hamilton J 1989 Econometrica 57 357–84), in which volatility influences returns. In addition, our models allow for feedback effects from returns to volatilities. Our family also includes GARCH option models as a special limiting case. Our models are more general than GARCH models in that our variance updating schemes do not only depend on levels of volatility and asset innovations, but also allow for a second factor that is orthogonal to asset innovations. The underlying processes in our family capture the asymmetric response of volatility to good and bad news and thus permit negative (or positive) correlation between returns and volatility. We provide the theory for pricing options under such processes, present an analytical solution for the special case where returns provide no feedback to volatility levels, and develop an efficient algorithm for the computation of American option prices for the general case.     

Testing linear factor models on individual stocks using the average F-test

In this paper, we propose the average F-statistic for testing linear asset pricing models. The average pricing error, captured in the statistic, is of more interest than the ex post maximum pricing error of the multivariate F-statistic that is associated with extreme long and short positions and excessively sensitive to small perturbations in the estimates of asset means and covariances. The average F-test can be applied to thousands of individual stocks and thus is free from the information loss or the data-snooping biases from grouping. This test is robust to ellipticity, and more importantly, our simulation and bootstrapping results show that the power of the average F-test continues to increase as the number of stocks increases. Empirical tests using individual stocks from 1967 to 2006 demonstrate that the popular four-factor model (i.e. Fama–French three factors and momentum) is rejected in two sub-periods from 1967 to 1971 and from 1982 to 1986.     

Exchange Rates and the Conversion of Currency‐Specific Risk Premia

How do the risk factors that drive asset prices influence exchange rates? Are the parameters of asset price processes relevant for specifying exchange rate processes? Most international asset pricing models focus on the analysis of asset returns given exchange rate processes. Little work has been done on the analysis of exchange rates dependent on asset returns. This paper uses an international stochastic discount factor (SDF) framework to analyse the interplay between asset prices and exchange rates. So far, this approach has only been implemented in international term structure models. We find that exchange rates serve to convert currency‐specific discount factors and currency‐specific prices of risk – a result linked to the international arbitrage pricing theory (IAPT). Our empirical investigation of exchange rates and stock markets of four countries presents evidence for the conversion of currency‐specific risk premia by exchange rates.     

Risky asset pricing based on safety first fund management

We study a portfolio selection model based on Kataoka's safety-first criterion (KSF model in short). We assume that the market is complete but without risk-free asset, and that the returns are jointly elliptically distributed. With these assumptions, we provide an explicit analytical optimal solution for the KSF model and obtain some geometrical properties of the efficient frontier in the plane of probability risk degree z _α and target return r _α. We further prove a two-fund separation and tangency portfolio theorem in the spirit of the traditional mean-variance analysis. We also establish a risky asset pricing model based on risky funds that is similar to Black's zero-beta capital asset pricing model (CAPM, for short). Moreover, we simplify our risky asset pricing model using a derivative risky fund as a reference for market evaluation.     

An Intertemporal International Asset Pricing Model: Theory and Empirical Evidence

We extend Campbell's (1993) model to develop an intertemporal international asset pricing model (IAPM). We show that the expected international asset return is determined by a weighted average of market risk, market hedging risk, exchange rate risk and exchange rate hedging risk. These weights sum up to one. Our model explicitly separates hedging against changes in the investment opportunity set from hedging against exchange rate changes as well as exchange rate risk from intertemporal hedging risk. A test of the conditional version of our intertemporal IAPM using a multivariate GARCH process supports the asset pricing model. We find that the exchange rate risk is important for pricing international equity returns and it is much more important than intertemporal hedging risk.     

Integrating Factor Models

This paper develops a comprehensive framework to address uncertainty about the correct factor model. Asset pricing inferences draw on a composite model that integrates over competing factor models weighted by posterior probabilities. Evidence shows that unconditional models record near-zero probabilities, while postearnings announcement drift, quality-minus-junk, and intermediary capital are potent factors in conditional asset pricing. Out-of-sample, the integrated model performs well, tilting away from subsequently underperforming factors. Model uncertainty makes equities appear considerably riskier, while model disagreement about expected returns spikes during crash episodes. Disagreement spans all return components involving mispricing, factor loadings, and risk premia.     

Comparing Asset Pricing Models

A Bayesian asset pricing test is derived that is easily computed in closed form from the standard F‐statistic. Given a set of candidate traded factors, we develop a related test procedure that permits the computation of model probabilities for the collection of all possible pricing models that are based on subsets of the given factors. We find that the recent models of Hou, Xue, and Zhang (2015a, 2015b) and Fama and French (2015, 2016) are dominated by a variety of models that include a momentum factor, along with value and profitability factors that are updated monthly.     

Bayesian Solutions for the Factor Zoo: We Just Ran Two Quadrillion Models

We propose a novel framework for analyzing linear asset pricing models: simple, robust, and applicable to high-dimensional problems. For a (potentially misspecified) stand-alone model, it provides reliable price of risk estimates for both tradable and nontradable factors, and detects those weakly identified. For competing factors and (possibly nonnested) models, the method automatically selects the best specification—if a dominant one exists—or provides a Bayesian model averaging–stochastic discount factor (BMA-SDF), if there is no clear winner. We analyze 2.25 quadrillion models generated by a large set of factors and find that the BMA-SDF outperforms existing models in- and out-of-sample.     

Model risk of the implied GARCH-normal model

Model risk causes significant losses in financial derivative pricing and hedging. Investors may undertake relatively risky investments due to insufficient hedging or overpaying implied by flawed models. The GARCH model with normal innovations (GARCH-normal) has been adopted to depict the dynamics of the returns in many applications. The implied GARCH-normal model is the one minimizing the mean square error between the market option values and the GARCH-normal option prices. In this study, we investigate the model risk of the implied GARCH-normal model fitted to conditional leptokurtic returns, an important feature of financial data. The risk-neutral GARCH model with conditional leptokurtic innovations is derived by the extended Girsanov principle. The option prices and hedging positions of the conditional leptokurtic GARCH models are obtained by extending the dynamic semiparametric approach of Huang and Guo [Statist. Sin., 2009, 19, 1037–1054]. In the simulation study we find significant model risk of the implied GARCH-normal model in pricing and hedging barrier and lookback options when the underlying dynamics follow a GARCH-t model.     

Taming the Factor Zoo: A Test of New Factors

We propose a model selection method to systematically evaluate the contribution to asset pricing of any new factor, above and beyond what a high-dimensional set of existing factors explains. Our methodology accounts for model selection mistakes that produce a bias due to omitted variables, unlike standard approaches that assume perfect variable selection. We apply our procedure to a set of factors recently discovered in the literature. While most of these new factors are shown to be redundant relative to the existing factors, a few have statistically significant explanatory power beyond the hundreds of factors proposed in the past.     

Fast Ninomiya–Victoir calibration of the double-mean-reverting model

Abstract

We consider the three-factor double mean reverting (DMR) option pricing model of Gatheral [Consistent Modelling of SPX and VIX Options, 2008], a model which can be successfully calibrated to both VIX options and SPX options simultaneously. One drawback of this model is that calibration may be slow because no closed form solution for European options exists. In this paper, we apply modified versions of the second-order Monte Carlo scheme of Ninomiya and Victoir [Appl. Math. Finance, 2008, 15, 107–121], and compare these to the Euler–Maruyama scheme with full truncation of Lord et al. [Quant. Finance, 2010, 10(2), 177–194], demonstrating on the one hand that fast calibration of the DMR model is practical, and on the other that suitably modified Ninomiya–Victoir schemes are applicable to the simulation of much more complicated time-homogeneous models than may have been thought previously.     

How many factors are important in U.K. stock returns?

I use the sequential approach of Harvey and Liu ([2018]. Lucky factors (Working Paper). Duke University) to build linear factor models in U.K. stock returns among a set of 13 candidate factors using individual stocks and three groups of test portfolios between July 1983 and December 2017. My study finds that the Market factor is the dominant factor in reducing mispricing in individual stocks and test portfolios regardless of the pricing error metric used. The Market factor has a bigger impact when using a value weighting pricing error metric. Whether a second factor is used or not depends upon which metric is used for mispricing and the time period examined. My study finds support for a two-factor model for the whole sample period of the Market factor and the Conservative Minus Aggressive (CMA) factor of Fama and French ([2015]. “A five-factor asset pricing model.” Journal of Financial Economics 116: 1–22) when giving greater weight to the mispricing of larger companies.     

Evaluating asset pricing models using the second Hansen-Jagannathan distance

We develop a specification test and a sequence of model selection procedures for non-nested, overlapping, and nested models based on the second Hansen-Jagannathan distance, which requires a good asset pricing model to not only have small pricing errors but also be arbitrage free. Our methods have reasonably good finite sample performances and are more powerful than existing ones in detecting misspecified models with small pricing errors but are not arbitrage-free and in differentiating models that have similar pricing errors of a given set of test assets. Using the Fama and French size and book-to-market portfolios, we reach dramatically different conclusions on model performances based on our approach and existing methods.     

Deriving the arbitrage pricing theory when the number of factors is unknown

This paper examines the use of proxies (or reference variables) for the true factors in the arbitrage pricing theory (APT). It generalizes other authors' existing work and shows that, when there are more reference variables than the true factors, the APT still holds. The possibility of fewer reference variables than the true factors is also considered, but the APT is not shown to hold, in the same sense, for this case. This work builds on an earlier paper by Ingersoll (Ingersoll J 1984 J. Finance 39 1021-39), and our propositions can be thought of as specializations of his theorems. Similar to Nawalkha (Nawalkha S 1997 J. Financial Economics 46 357-81), our work does not use the mathematics of Hilbert and Banach spaces and, thus, is open to a much wider audience. The practical implication of our results is that model builders should be generous with the number of factors they use, as excessively parsimonious models suffer from inaccuracy.     

Agency and Institutional Investment

In this paper we summarise and extend the agency‐based model of asset pricing of Brennan (1993) to show that the implied agency effects on asset pricing are too small to be empirically detectable: empirical tests confirm this and we show that the positive findings of Gomez and Zapatero (2003) are due to their choice of sample. We also derive new empirical implications for the composition of institutional investment portfolios and empirically confirm the major result, that institutional portfolios will be short the minimum variance portfolio.     

A skeptical appraisal of asset pricing tests

It has become standard practice in the cross-sectional asset pricing literature to evaluate models based on how well they explain average returns on size-B/M portfolios, something many models seem to do remarkably well. In this paper, we review and critique the empirical methods used in the literature. We argue that asset pricing tests are often highly misleading, in the sense that apparently strong explanatory power (high cross-sectional R²s and small pricing errors) can provide quite weak support for a model. We offer a number of suggestions for improving empirical tests and evidence that several proposed models do not work as well as originally advertised.     

Portfolio optimization for student t and skewed t returns

We consider the problem of valuing a European option written on an asset whose dynamics are described by an exponential Lévy-type model. In our framework, both the volatility and jump-intensity are allowed to vary stochastically in time through common driving factors—one fast-varying and one slow-varying. Using Fourier analysis we derive an explicit formula for the approximate price of any European-style derivative whose payoff has a generalized Fourier transform; in particular, this includes European calls and puts. From a theoretical perspective, our results extend the class of multiscale stochastic volatility models of Fouque et al. [Multiscale Stochastic Volatility for Equity, Interest Rate, and Credit Derivatives, 2011] to models of the exponential Lévy type. From a financial perspective, the inclusion of jumps and stochastic volatility allow us to capture the term-structure of implied volatility, as demonstrated in a calibration to S&;P500 options data.