Commodities | Credit | Equities | Rates
The holy grail of financial market investing is to find investment or trading rules that consistently make money over time. And while no one has found a rule that always avoids losses, some rules do better than others.
The simplest profitable strategy is to go long in asset market like equities or bonds. This type of return is known as beta. After this, the most widely known, profitable strategies relate to factors or factor premia. These are trading rules that have been shown to be profitable over long stretches of time and have developed academic support. Four common ones that can applied to several different markets are value, momentum, carry, and defensive.
But what are the benefits of each? And do any have serious pitfalls? Here, we take a look at an academic paper recently published by researchers from the quantitative hedge fund AQR to find some answers.[1]…
This article is only available to Macro Hive subscribers. Sign-up to receive world-class macro analysis with a daily curated newsletter, podcast, original content from award-winning researchers, cross market strategy, equity insights, trade ideas, crypto flow frameworks, academic paper summaries, explanation and analysis of market-moving events, community investor chat room, and more.
The holy grail of financial market investing is to find investment or trading rules that consistently make money over time. And while no one has found a rule that always avoids losses, some rules do better than others.
The simplest profitable strategy is to go long in asset market like equities or bonds. This type of return is known as beta. After this, the most widely known, profitable strategies relate to factors or factor premia. These are trading rules that have been shown to be profitable over long stretches of time and have developed academic support. Four common ones that can applied to several different markets are value, momentum, carry, and defensive.
But what are the benefits of each? And do any have serious pitfalls? Here, we take a look at an academic paper recently published by researchers from the quantitative hedge fund AQR to find some answers.[1]
Defining Factor Trading Rules
First, the researchers constructed an impressive dataset that covers equity, commodity, bond, and currency markets that in some cases stretch back to 1920. They then created trading rules based on the four factors to determine their profitability. All the rules are based on ranking markets by factor and then buying the more attractive ones and selling the less attractive ones. The factors are defined for each market as described in Figure 1.
Figure 1: Factor Rules for Different Asset Classes
So, for example, the value trading rule for equity indices would be to rank the countries CAPE ratios from cheapest (lowest CAPE) to most expensive (highest CAPE). Then, buy the higher ranked (cheap) countries and sell the lower ranked (expensive countries).
Trading Results
Using these rules across all the markets delivers impressive results. The best returns are for individual stocks, where the value factor delivered a Sharpe ratio (excess return over volatility) of 1.1 for international and 0.74 for US stocks. The weakest results are for fixed income and currencies where the momentum factor delivered a paltry 0.12 and 0.18, respectively. The defensive factor for fixed income was even weaker at 0.05 (see Figure 2).
Figure 2: Sharpe Ratios of Following Factor Rules in Each Asset Class
Trading Rules Are Not Arbitraged Away
The paper then provides a supporting case for these trading rules or factors. They also perform some useful tests to determine whether these factors were data-mined or could be arbitraged away now that they are well known, finding that the answer is no for a significant number.
The Factor Return Cycle
The variability of the returns of these factors or trading rules is an issue because although the average returns over long stretches of time are impressive, there are periods of times when the returns are poor and even negative. For example, the value factor for equity indices returns has seen 90th decile annual returns of over 20% but 10th decline annual returns have been around -15% (see Figure 3).
Figure 3: Distribution of Market and Factor Returns by Asset Class
The figure plots the annualized mean return, 10th, and 90th return percentiles of the market portfolio and the factor portfolios pertaining to value, momentum, carry and defensive for each asset class. Also reported are the distribution of returns for multifactor portfolios for each asset class and the multifactor, multi-asset class portfolio.
Diversification Works
Perhaps the best path to reduce this variability is to exploit the low correlations of the factor returns within and across asset classes. This gives scope for diversification gains. Within an asset class, we can combine value, momentum, carry, and defensive factors. This pushes up the Sharpe ratios (and reduces drawdowns) for each asset class above the best individual factor for that class. For individual US stocks, the Sharpe ratio moves up 1.17; for individual, international stocks it moves up to 1.47; and for equity indices it moves up to 0.61. Meanwhile, for fixed income it moves up to 0.54; for currencies it moves up to 0.65; and for commodities it moves up to 0.65.
A single factor such as carry can also be combined across all asset classes, which raises the Sharpe ratio higher the any individual market. For example, the best market for carry (fixed income), delivers a Sharpe ratio of 0.66. However, if we combine carry in all the markets, we arrive at a Sharpe ratio of 0.84. Finally, combining all the factors with all the markets, the Sharpe ratio jumps to 1.59 with a maximum drawdown of 4%. The variability of returns also improves dramatically – the 90th decline returns are 6.3% and the 10th decile returns are 0.3%.
Other Factor Timing Approaches
The paper spends a significant amount of time attempting to improve these returns by finding additional rules that could time the ups-and-downs of the factor returns. Despite trying a range of indicators and tests, the results were unimpressive.
One of the best-known techniques, value spread timing, computes the valuations of the long positions of a given factor or trading rule and compares them to the valuations of the short positions. If the gap is large, then the overall positions are scaled in proportion to cheapness. A simple version of this strategy appears to improve returns of individual US stocks, but it fails to work well across other stocks and asset classes, which questions the robustness of the results. A more complex version, where one regresses factor returns on lagged value spreads to determine the recent relationship between the two, does show more promise across all markets.
One can also look at the spread within each factor. If the size of carry is very large, then one increases the size of the carry position; or if markets have moved a lot, then one increases the size of the momentum position. This is known as the factor spread. Other techniques include using factor momentum, five-year reversals, the volatility of the factor returns, business cycle measures, and general market timing indicators such as CAPE and VIX. None of these fares particularly well, except the business cycle measure.
The Real World Bottom Line
If we apply transaction costs and more realistic out-of-sample testing procedures, none of these factor timing techniques appear to be that profitable. In the end, it appears that diversifying across factors and asset classes is the most robust method to reduce the variability of returns.
The takeaway: Factor trading rules, such as value, carry, momentum and defensive, consistently deliver positive returns across multiple markets. To minimise the fluctuations of these trading returns, it is best to combine factors and asset classes. Most indicators used to time the returns fail to add much beyond this approach.
[1] Ilmanen, Antti S. and Israel, Ronen and Moskowitz, Tobias J. and Thapar, Ashwin K and Wang, Franklin, “Do Factor Premia Vary Over Time? A Century of Evidence” (June 7, 2019). Available at SSRN: https://ssrn.com/abstract=3400998 or http://dx.doi.org/10.2139/ssrn.3400998.
Bilal Hafeez is the CEO and Editor of Macro Hive. He spent over twenty years doing research at big banks – JPMorgan, Deutsche Bank, and Nomura, where he had various “Global Head” roles and did FX, rates and cross-markets research.
(The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs.)