Using Venn to Analyze Risk Premia Strategies

By Alex Botte, CFA, CAIA on April 4, 2019

Risk Premia strategies performed poorly in 2018, as represented by the SG Multi Alternative Risk Premia Index. In this post, Venn uses the Two Sigma Factor Lens to analyze how the index’s factor exposures and residual may have contributed to its poor performance.

What are “Risk Premia”?

Factor-based investing is far from a new phenomenon. In fact, the first well-known factor, the equity market itself, was introduced in the 1960s with the Capital Asset Pricing Model.1 The idea behind the model is that drivers of stock returns include systematic risk – exposure to overall economic growth and corporate profitability – in addition to idiosyncratic risk that is particular to a specific company. The former is priced and has a corresponding return premium, while the latter introduces extra risk that may or may not be compensated.

The systematic risk of the equity market can be mimicked by a market-capitalization weighted index. Therefore, index investing was the original form of factor investing, as investors who simply wanted exposure to systematic equity risk could purchase the global equity market through a capitalization-weighted index-tracking fund. That risk has become commonly known as equity beta. Researchers have identified other beta exposures in markets since the 1960s, including exposure to interest rates risk, failure-to-pay (or credit) risk, and commodity risk.2

Recently, academic research of risk factors has expanded beyond traditional asset class oriented betas. For example, in the early 1990s, renowned academics Eugene Fama and Kenneth French suggested that there may be other sources of systematic risk within an asset class. They introduced two factors, Value and Size, to show that stocks that are cheaply priced relative to fundamentals and that have smaller market capitalizations tend to have higher returns and higher correlations to one another than stocks that are expensive and have larger market caps.3

Ever since this breakthrough from Fama and French, researchers have “discovered” an ever-increasing number of new factors that seek to explain more risk and return within asset classes.4 This has resulted in investment products that tilt towards factors like Value and Size to outperform the overall market (e.g., smart beta funds). Another implementation of this phenomenon that has gained traction are “risk premia” funds that invest in the market-neutral versions of the factors (i.e., buying securities that exhibit characteristics of a particular factor and selling those that do not) to generate a positive absolute return.

Risk premia funds, while perhaps appearing similar on the surface, require several factor selection, construction, and design-related choices that may result in large tracking errors across managers. Questions like “which factors to include in the fund” and “how do I define and capture a well-known factor like Value” may result in meaningful differences across managers that manifest in performance dispersion.5

Venn provides allocators with tools to help analyze risk premia manager performance and better understand their manager’s factor choices.

Case Study: SG Multi Alternative Risk Premia Index

In order to demonstrate how investors can use Venn’s factor-based analytics to help understand risk premia performance, we’ll analyze the publicly available, daily priced SG Multi Alternative Risk Premia Index.6 In particular, we’ll explore the index’s negative performance of -4.7% in 2018.

The index is composed of a variety of risk premia investment managers that invest in multiple asset classes (including equities, fixed income, currencies, and commodities) using certain risk factors (most commonly Value, Momentum, and Carry).7

Given the composition of the index, we would expect it to have low correlations to core macro factors like Equity and Interest Rates (as risk premia managers typically construct long-short portfolios, thereby minimizing exposure to the principal drivers of asset class returns) and have positive exposure to equity style factors, particularly Value and Momentum. Displayed below are factor contributions to risk and factor exposures for this index, as determined by the Two Sigma Factor Lens on Venn.

Exhibit 1 | SG Multi Alternative Risk Premia Index Factor Contributions to Risk
Source: Venn analysis. February 2019. Time period: January 1, 2016 – February 22, 2019, using daily data.

Exhibit 2 | SG Multi Alternative Risk Premia Index Factor Exposures
Source: Venn analysis. February 2019. Time period: January 1, 2016 – February 22, 2019, using daily data.

Since the index’s inception in 2016,8 the majority of risk that can be explained by Venn is attributable to positive exposures to the following five factors: Interest Rates and Equity (core macro factors), and Momentum, Low Risk, and Value (market-neutral equity style factors).

Another notable exposure is Equity Short Volatility, which seeks to capture strategies that have negative exposure to moves in equity market volatility. Strategies that seek to harvest the “volatility risk premium” typically involve writing put options on the equity market (and possibly other markets) that bet on prices remaining steady or increasing. The expectation is to receive an extra return for bearing the risk of large losses if realized volatility increases suddenly before the option expires.9 The index’s positive exposure to this factor may imply that managers in the index are implementing this strategy.

Finally, over 50% of the index’s risk is unexplained by the Two Sigma Factor Lens. The next section will examine why this might be the case.

Now that we know the index’s factor exposures according to Venn, let’s analyze how these factors contributed to the index’s return in the 2018 calendar year.

Exhibit 3 | SG Multi Alternative Risk Premia Index Factor Contributions to Return
Source: Venn analysis. March 2019. Time period: January 1, 2018 – December 31, 2018, using daily data.

Three out of the five main factors the index has been exposed to over the long term detracted from the index’s performance in 2018. Namely, the Equity factor, which was down over 8% in 2018,10 detracted over 1% from the index’s return. Many investors may use alternative risk premia funds as a way to generate returns with low correlations to equity and macro betas, making this observation notable in terms of the diversification benefits realized.

Two other factors were notable detractors in 2018. First, a negative exposure to Local Equity (which implies that the index is overweight or long international stocks and underweight or short U.S. stocks) detracted another 1.5%. The Local Equity factor outperformed in 2018,11 as the U.S. outperformed the rest of the world on a risk-adjusted basis. Second, the aforementioned positive exposure to Equity Short Volatility also contributed to the index’s negative performance, as there were several events in 2018 in which volatility meaningfully spiked, especially in early February 2018.12

Finally, it is clear that the residual, which, as we saw previously, represents more than half of the index’s risk, was the most meaningful detractor from the index over this period. Let’s discuss what factors may have caused this large negative return.

Understanding the Residual

Exhibit 1 indicates that 50% of the SG Multi Alternative Risk Premia Index’s risk is unexplained by the Two Sigma Factor Lens. That residual has been contributing negatively to returns, as displayed in Exhibit 3. What could be causing this residual, or unexplained, risk?

The first cause of the residual risk could be factors not included in the Two Sigma Factor Lens. While the Two Sigma Factor Lens covers several types of market betas, like Equity and Interest Rates, as well as risk premia factors within the equity universe, like Value and Momentum, it does not include risk premia within macro asset classes. The SG Multi Alternative Risk Premia Index includes managers that invest in risk premia across many asset classes, including fixed income, currencies, and commodities. Factors that attempt to explain risk across these types of asset classes, like Trend Following, Currency Value, or Fixed Income Carry, are not included in the Two Sigma Factor Lens, so any returns coming from those factors would fall in the residual category.

Trend Following is an example of a risk premia factor that generally performed poorly in 2018. According to the SG Trend Index, which tracks the 10 largest trend-following CTAs,13 the category was down over 8% in 2018.14 To the extent that risk premia managers pursue Trend Following, any poor performance attributable to that strategy would fall within the residual category on Venn’s factor analysis.

The second cause of residual risk could be differences between the managers’ factor construction and the construction of the Two Sigma Factor Lens. Decisions like how a manager defines a risk premia factor within an asset class can have a meaningful impact.

For example, we know that the index exhibited a statistically significant positive exposure to the Momentum equity style factor, as displayed in Exhibit 2. The Momentum factor in the Two Sigma Factor Lens is defined by a single metric: how did the stock’s price perform over the trailing year? If it outperformed, the stock has positive momentum exposure, and vice versa for stocks that underperformed. However, there are many other ways to define Momentum. For example, a manager may evaluate stock price performance over different periods (e.g., trailing three or six months), or they may consider more fundamental measures of momentum, such as earnings trends over time. While these measures may all be positively correlated, they can exhibit tracking error to one another, especially over short periods of analysis.

An example of metric divergence in 2018 is Low Risk. The Two Sigma Factor Lens defines the Low Risk factor in two ways: the stock’s beta and its residual return volatility. The first metric posted gains of 5.8% in 2018, while the latter delivered -1.5%.15 Therefore, a manager using solely beta to define their Low Risk factor would have outperformed another manager that used solely residual return volatility, all else equal.

Finally, it’s important to note that the managers in the SG Multi Alternative Risk Premia Index must report returns net of fees and transaction costs.16 The factors in the Two Sigma Factor Lens do not account for fees. Further, the factors are not actually traded. As such, their returns are not directly reduced for transaction, financing, or any similar costs.

However, the equity style factors do consider transaction and financing costs in their construction implicitly by making various adjustments, including tilting the factor portfolios toward more heavily traded stocks, slowing the implicit trading of factor portfolios by smoothing the incorporation of new stock data, and applying certain constraints on their investment universes, like eliminating certain thinly traded and illiquid equities. These methods are not necessarily reflective of any actual or potential costs and will affect the residual return estimates for managers when analyzed through the Two Sigma Factor Lens.

In summary, it may be helpful to consider the risk premia, asset classes, and factor construction choices the manager makes and how those differ from the risk factors used for analysis when analyzing risk premia manager performance through a risk factor lens framework.

What does the 2018 performance say about risk premia strategies?

As mentioned previously, the SG Multi Alternative Risk Premia Index was down -4.7% in 2018, representing a disappointing year for the risk premia category. Does this mean that allocators should dismiss the category? While the index’s poor performance may have shed a negative light on the risk premia strategy last year, it’s important to reflect on why investors found this strategy attractive in the first place.

First, it has delivered positive returns over its full history. Exhibit 4 shows the 7.9% cumulative return of the index since its inception in 2016 through the end of January 2019.

Exhibit 4: SG Multi Alternative Risk Premia Index Cumulative Returns
Source: Venn analysis. March 2019. Time period: January 1, 2016 – January 31, 2019, using daily data.

Additionally, risk premia managers seek exposure to factors that have strong empirical evidence and fundamental justifications for long-term return premia. For example, as displayed in Exhibit 5 below, the equity style factors in the Two Sigma Factor Lens have all have delivered positive returns17 going back over 16 years.

Exhibit 5: Equity Style Factor Cumulative Returns
Source: Venn analysis. March 2019. Time period: December 26, 2002 – February 27, 2019, using daily data.

Further, most equity style factors have a fundamental reason for why investors should be rewarded for holding exposure to that risk factor over time. For example, small-cap stocks tend to be more exposed to downturns in the equity market. Recessions could affect the viability of a small-cap company more than that of a larger one, as worsening economic conditions are associated with a systematically larger decline in sales and investment for smaller firms than for larger firms.18 Therefore, those who invest in smaller-capitalization stocks should be compensated for bearing that extra risk.

Further, risk premia strategies have been additive to a traditional 60/40 portfolio.19 The category was intended to provide diversification to traditional market exposures and therefore had the opportunity to increase the portfolio’s risk-adjusted return. As demonstrated in Exhibit 6, adding a 10% risk premia allocation to a pure 60/40 portfolio20 on Venn produced a better risk-adjusted return, lower volatility, and lower maximum drawdown over the short period since the SG Multi Alternative Risk Premia Index was incepted.

Exhibit 6: Pro Forma 60/40 Portfolio Analysis with and without Risk Premia
Source: Venn analysis. March 2019. Time period: January 4, 2016 – February 22, 2019, using daily data.

Conclusion

The SG Multi Alternative Risk Premia Index, a proxy for the risk premia category, experienced poor performance last year that was partly explainable by exposures to negative-returning factors, like Equity. The index also has a large residual component that delivered negative returns.

The large residual component of the index may be driven by factors that are not included in the Two Sigma Factor Lens and/or differences in how the underlying managers construct their factors.

Managers in the risk premia category must make several decisions when designing their funds, including which risk premia factors to include, in which asset classes, and how they’ll construct their factors. These choices can impact their factor exposures (for example, not constructing factors to be market-neutral could lead to unintended macro factor exposures) and cause differences in residual returns and risk.

Venn, by analyzing factor exposures and contributions to risk and return using the Two Sigma Factor Lens, can help allocators understand if a risk premia strategy is additive to their existing portfolio, which risk premia factors they want exposure to, and how their managers meet those criteria.

Visit Venn today to start evaluating risk premia strategies.

References

1 Sharpe, William F. (September 1964). “Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk”. The Journal of Finance.

2 For example: Joehnk, Michael D. and Frank F. Reilly (December 1976). “The Association Between Market-Determined Risk Measures for Bonds and Bond Ratings”. The Journal of Finance.

3 Fama, Eugene F. and Kenneth R. French (June 1992). “The Cross-Section of Expected Stock Returns”. The Journal of Finance.

4 Feng, Guanhao, Stefano Giglio, and Dacheng Xiu (January 2, 2019). “Taming the Factor Zoo: A Test of New Factors”. Fama-Miller Center for Research in Finance.

5 For more information, read Two Sigma’s white paper Risk Factors Are Not Generic.

6 Source: Société Générale Prime Services Indices.

7 Source: Société Générale Prime Services Indices.

8 Source: Société Générale Prime Services Indices.

9 Fallon, William, James Park, and Danny Yu (2015). “Asset Allocation Implications of the Global Volatility Premium,” Financial Analysts Journal.

10 Source: Venn Factor Insights. April 2019. Using daily data.

11 Source: Venn Factor Insights. February 2019. Using daily data.

12 Stronger growth and improving confidence around tax reform in the U.S. led to concerns of a more hawkish Federal Reserve.

13 Commodity Trading Advisors.

14 Source: Société Générale Trend Index.

15 Source: Venn Factor Insights. April 2019. Using daily data.

16 Source: SG Multi Alternative Risk Premia Index Construction Methodology.

17 Gross of fees and transaction costs, as mentioned previously.

18 Crouzet, Nicolas and Neil R. Mehrotra (2017). “Small and Large Firms over the Business Cycle,” Research Division Federal Reserve Bank of Minneapolis.

19 A 60/40 portfolio of stocks and bonds is used to present an illustrative example of a portfolio that has a mix of common asset class exposures.

20 The 60/40 portfolio is composed of 60% to the MSCI ACWI Price Index and 40% to the Bloomberg Barclays Global Aggregate Index.

This article is not an endorsement by Two Sigma Investor Solutions, LP or any of its affiliates (collectively, “Two Sigma”) of the topics discussed.  The views expressed above reflect those of the authors and are not necessarily the views of Two Sigma. This article (i) is only for informational and educational purposes, (ii) is not intended to provide, and should not be relied upon, for investment, accounting, legal or tax advice, and (iii) is not a recommendation as to any portfolio, allocation, strategy or investment.  This article is not an offer to sell or the solicitation of an offer to buy any securities or other instruments. This article is current as of the date of issuance (or any earlier date as referenced herein) and is subject to change without notice. The analytics or other services available on Venn change frequently and the content of this article should be expected to become outdated and less accurate over time.  Two Sigma has no obligation to update the article nor does Two Sigma make any express or implied warranties or representations as to its completeness or accuracy. This material uses some trademarks owned by entities other than Two Sigma purely for identification and comment as fair nominative use. That use does not imply any association with or endorsement of the other company by Two Sigma, or vice versa. See the end of the document for other important disclaimers and disclosures. Click here for other important disclaimers and disclosures.