Investing in a Multi-Asset Multi-Factor World
Alexandar Cherkezov (Invesco), et al.
August 31, 2017
In this article, we advance the use of factor investing across multiple asset classes. It turns out that style factors well established in the equity domain – such as value, momentum or quality – do extend to other asset classes as well. Even more so, multi-asset multi-factors significantly expand the investment opportunity set relative to a traditional multi-asset universe. Seeking to exploit this potential, we put forward an innovative diversified risk parity strategy that is designed to strive for maximum diversification in the multi-asset multi-factor world. To illustrate the strategy’s merits, we investigate its stylized facts vis-à-vis more standard allocation approaches.
Building Multi-Factor Portfolios:
Challenges to Consider and How to Overcome Them
Tony Guida and Benjamin Raffin (RPMI Railpen Investments)
September 27, 2017
Over the last 18 months many research papers have been released about different portfolio construction for multi-factor portfolio. However, it is not clear which methodology can provide superior returns superior between top-down (portfolio blending, mixed) and bottom-up (signal blending, integrated). This lively research corpus and debate leave investors with no clear guidance on what to do. The purpose of this paper is twofold: first to give a framework to tackle the main challenges and steps that one has to consider when building a multi-factor portfolio, second to give empirical evidence regarding the more efficient methodology to build a multi-factor portfolio through practitioner’s lenses. In order to do so, we empirically compare both approaches and find that bottom-up portfolios show more compelling characteristics in terms of factor exposure, risk-adjusted returns, diversification and liquidity.
Gregg S. Fisher (Gerstein Fisher) and Michael B. McDonald IV (Fairfield U.)
September 30, 2017
We examine the time series asset pricing factor returns and their use in a portfolio that varies over time based on an investor’s remaining human capital. Using of data for a common set of four different risk factors for the period 1980 to 2013, we show that risk premiums to different factors are not constant over time, and that investors may improve their risk return trade-off by weighting or “tilting” their portfolios as they age. Our results suggest that those investors targeting higher returns should tilt towards the size and value factors, while investors favoring lower levels of risk, should tilt towards quality. Our results raise questions about the current industry approach to asset allocation and the driving forces behind the magnitude of risk premiums over time.
Value Timing: Risk and Return Across Asset Classes
Fahiz M. Baba Yara (New University of Lisbon), et al.
October 16, 2017
Returns to value strategies in individual equities, commodities, currencies, global government bonds and stock indexes are predictable by the value spread. The value spread captures the strength of the value signal in the long relative to the short portfolio of a value strategy. We show that both common and asset-class-specific components of the value spread contribute to this predictability. Whereas return variation due to common value is largely driven by standard proxies of risk, such as the default spread and real uncertainty, the return variation due to specific value presents a challenge for asset pricing models. Evidence from a range of value timing and rotation strategies shows that investors can benefit from information in the value spread in real-time.
Style Concentration in Ownership and Expected Stock Returns
Gikas A. Hardouvelis and Georgios I. Karalas ((University of Piraeus)
We examine the relation of expected stock returns with fund style concentration in stock ownership over the period 1997-2015. Concentration is measured by the Herfindahl index H of the shares of different investment styles in the ownership of stocks and represents a measure of investor inattention in stocks. Decile portfolios on H reveal a strong positive association of H with future returns, with the long-short portfolio on H having significant alphas after passing through the five-factor Fama-French (2015) model… The results are consistent with the model of Merton (1987), which claims a stock’s excess risk premium over the CAPM premium, is the product of investor participation (which is proxied by H in our framework), idiosyncratic volatility and size. These results also shed light on the small firm effect.
‘Know When to Hodl ‘Em, Know When to Fodl ‘Em’:
An Investigation of Factor Based Investing in the Cryptocurrency Space
Stefan Hubrich (T.Rowe Price)
October 28, 2017
It has been known since at least the groundbreaking work of Fama and French (1992) that there are specific attributes, so called factors, that can help predict the returns of individual assets above the return of the broader market. Since these predictive characteristics arise out of sample (with currently observable factor values predicting future returns), investors can earn excess returns with portfolios that are constructed to align with the factors. First introduced in the cross section of returns and focusing on individual equity securities, the efficacy of such factors has since been demonstrated at the asset class level as well, and found to work not only in the cross section but also longitudinally (for individual assets, through time). Factors like value, momentum, and carry have been found to work so broadly across different asset classes, security universes, countries, and time periods, that Asness et al. simply titled their influential 2013 Journal of Finance paper “Value and Momentum Everywhere”. Our paper provides a first application of momentum, value, and carry based factor investing to the cryptocurrencies. We show that these same factors are effective in this relatively new and unexplored asset class, permitting the construction of portfolios that can earn excess returns over the cryptocurrency “market” as a whole.
Another Look at Currency Risk in International Stock Returns
George A. Karolyi (Cornell U.) and Ying Wu (Stevens Institute of Tech.)
October 21, 2017
This paper offers new evidence on how currency risk is priced in the cross-section of international stock returns. Our experiment examines this long-standing question for a wide variety of test asset portfolios comprised of monthly returns for over 37,000 stocks from 46 countries over a two-decade period. We obtain some positive evidence of the pricing of currency risk, but the implied premia on the currency risk factors are economically small. The inferences are fragile – they depend critically on the benchmark factor models used, the sub-period evaluated, and even the composition of the test asset portfolios assessed. Overall, we judge that currency risk factors do not help to capture much of the time-series or cross-sectional variation in stock returns for global and regional portfolios.