This study shows that since 1927, investors would have earned a statistically significant excess return of nearly two percent per month by investing in the U.S. equity market from November through April in presidential pre-election years. On the other hand, Treasury bond returns performed inversely to the equity returns, i.e., they have been higher in summer (May to October) months and in other-than-pre-election-years (especially in midterm election years). Our equity results suggest that the previously documented Halloween and pre-election year effects are intertwined. The combined Halloween–pre-election year effect shows up consistently in sub-periods; in an extended sample period since 1871; and in international stock markets. It appears to be separate from a January anomaly; it is independent of the political party in the White House; and it doesn’t appear to be a compensation for higher risk. In contrast, small (value) stocks outperform large (growth) stocks in the November–to–April period in years other than presidential pre-election years. We show that the winter–pre-election year premiums align with the Baker et al. (2016) measure of economic policy uncertainty, and we propose that models of political uncertainty potentially explain both the equity and bond results.
Alpha Signals, Smart Beta and Factor Model Alignment
Terry Marsh and Paul Pfleiderer, Forthcoming: Journal of Portfolio Management, May 2016
The authors consider the case for augmenting a user risk model in light of a manager’s alphas to be used in portfolio construction. They consider both “smart beta” models and a case where alpha signals are partly factor-driven but incorrectly perceived to be stock-specific. In the smart beta case, the authors argue that mechanically augmenting the risk model risks distorting an otherwise-correct factor structure. For the case of ex ante alpha signals and potentially missing factors, errors of omission in factors are shown to generate larger expected losses in portfolio efficiency than do errors of commission when “phantom factors” are unintentionally included. In the case where the alpha signals simply contain noise, then mechanically augmenting the risk model with a custom penalty to offset that noise can improve portfolio efficiency. But in this case of alpha noise, the penalty has nothing to do with missing factors per se and the “customization” is not with respect to the risk model but to the manager’s alphas.
Flight to Quality and Asset Allocation in a Financial Crisis
Terry Marsh and Paul Pfleiderer, Financial Analysts Journal, July-August 2013
With respect to the recent financial crisis, the authors argue that the appropriate adjustments to portfolio allocations in response to the market dislocation are determined by equilibrium considerations (supply must equal demand) and depend on individual investors’ characteristics relative to societal averages. Using a simple model that captures the magnitude of the recent crisis, the authors show that the optimal tactical adjustments for most portfolios require a turnover of less than 10%.
Terry Marsh and Paul Pfleiderer, Review of Pacific Basin Financial Markets and Policies, Vol. 15, No. 2 (2012) 1250008
Posted with permission by:World Scientific Publishing Company
Post-mortems of the financial crisis typically mention “black swans" as the rare events that were the Achilles heel of financial models, manifesting themselves as “25 standard deviation events occurring several days in a row." Here, we briefly discuss the implications of “black swan" events in asset pricing and risk management. We then show that the “black swans" problem virtually disappears for S&P Index returns when surprises are measured relative to the standard deviation of the conditional S&P distribution. In our illustration, we use the one-day-lagged VIX as an easy-to-understand measure of that conditional S&P standard deviation.
Keywords: Black swans; fat tails; unknown unknowns; conditional S&P returns; VIX; financial crisis, model failure.
Terry Marsh and Paul Pfleiderer, Preface: The Recent Trend of Hedge Fund Strategies, Ed. by Yasuaki Watanabe, Nova Publisher, 2010.
In this Preface, we offer some analysis of the 2008-2009 financial crisis and its implications for financial industry reform and research. We primarily focus on issues relating to transparency and the measurement of risk and how these are affected by management incentives that are often misaligned with the incentives of those who are exposed in various ways to the risk being measured. In the aftermath of the crisis many have called for increased transparency; we suggest that while transparency is no doubt a desirable goal in many ways, enhancing it could prove to be quite difficult.
The Relation between Fixed Income and Equity Return Factors
Jaime Lee, Terry Marsh, Robert Maxim, and Paul Pfleiderer, Journal of Investment Management, 4(4), January 2006, 52-72.
The paper provides an analysis of the relation between equity and fixed income returns over time. As measured by realized correlation, this relation has changed substantially over the last decade, from positive to negative through the market collapse and is currently around zero. We find "jumps" in the co-movements of equity and bond returns at a daily frequency; these jumps can at times be attributed to "flight to liquidity" phenomena in the markets, and at other times, to apparent surprise announcements in expected inflation or related macro conditions. We find no evidence of short-run persistence in the jumps in daily co-movement of bond and equity returns, but there does seem to be a "regime-like" longer-run persistence in them, perhaps associated with Federal Reserve "management over the last decade.
Surprise volume and heteroskedasticity in equity market returns
Niklas Wagner and Terry Marsh, Quantitative Finance, Vol. 5, No. 2 April 2005, 153-168.
Heteroskedasticity in returns may be explainable by trading volume. We use different volume variables, including surprise volume - i.e. unexpected above-average trading activity - which is derived from uncorrelated volume innovations. Assuming weakly exogenous volume, we extend the Lamoureux and Lastrapes (1990) model by an asymmetric GARCH in-mean specification following Golsten et al. (1993). Model estimation for the US as well as six large equity markets shows that surprise volume provides superior model fit and helps to explain volatility persistence as well as excess kurtosis. Surpirse volume reveals a significant positive market risk premium, asymmetry and a surprise volume effect in conditional variance. The findings suggest that e.g. a surprise volume shock (breakdown) - i.e. large (small) contemporaneous and small (large) lagged surprise volume - relates to increased (decreased) conditional market variance and return.
Measuring tail thickness under GARCH and an application to extreme exchange rate changes
Niklas Wagner and Terry Marsh, Journal of Empirical Finance, 12 (2005), 165-185.
Accurate modeling of extreme price changes is vital to financial risk management. We examine the small sample properties of adaptive tail index estimators under the class of student-t marginal distribution functions including generalized autoregressive conditional heteroskedastic (GARCH) models and propose a model-based bias-corrected estimation approach. Our simulation results indicate that bias relates to the underlying model and may be positively as well as negatively signed. The empirical study of daily exchange rate changes reveals substantial differences in measured tail thickness due to small sample bias. Thus, high quantile estimation may lead to a substantial underestimation of tail risk.
David Tien, Paul Pfleiderer, Robert Maxim, and Terry Marsh, Linear Factor Models in Finance, Chapter 13: Ed. by John Knight and Stephen Satchell, Elsevier Finance, Amsterdam.
This study addresses the problem of accurately forecasting and attributing risk in equity portfolios. It develops a hybrid methodology that takes advantage of the superior forecasting power of implicit factor models while also attributing portfolio risk to economic factors and firm-specific characteristics. It then compares the relative accuracy of risk attribution using this hybrid approach versus an explicit cross-sectional factor model. It presents simulation results suggesting, given realistic parameter values, that the estimation efficiency gained by using the hybrid approach yields substantial improvements over explicit models.
The Role of Country and Industry Effects in Explaining Stock Returns
Terry Marsh and Paul Pfleiderer, Working Paper, 1997
The relative role of industry and country factors in explaining global returns on individual stocks is studied. In contrast to past studies, e.g. Heston and Rouwenhorst (1994) and Roll (1992), the sensitivities of stocks' returns to these factors are allowed to differ across stocks. We find that the industry factor explains 20% - 30% of the variation in stock returns that can be accounted for by country and industry, and about 7% of the variation explained when a global factor is also included. This relative explanatory power for the industry factor is much higher than the 1%-or-less estimate reported by Heston and Rouwenhorst because, we argue, they focus on index returns. Strikingly, however, we find that the broad nine-segment "industrial classification" in the Dow Jones Global Index (DJGI®) explains as much return variation as the finer 68-industry classification, and that the 38-industry classification in the Morgan Stanley Capital International (MSCI®) World Index explains slightly more. Finally, we find that measuring global stock returns in a common currency decreases the return variation attributable to industry factors by about 15%.