What is the Proper Benchmark for Momentum or Trend-Following Strategies?

Most academics and practitioners tend to compare momentum or trend-following strategies to a buy and hold investment strategy. They do so by comparing the results of the strategies to a benchmark that is a proxy for buy and hold.  From this type of analysis many ‘experts’ justify why the strategies are either superior to buy and hold, or on the contrary side why they aren’t worth pursuing. Both sides of the argument are wrong—they are wrong because they are using the wrong benchmark altogether.

The classic passive method for portfolio construction is to hold a portion of the portfolio in risky assets like stocks and offset their risk by holding bonds and cash. In contrast, tactical strategies by their very nature will generally spend time in bonds or cash only when conditions are unfavorable for risky assets. To make matters even more complex, different tactical strategies can spend more or less time in cash.

In order to properly evaluate a tactical manager, we need to look at longer periods of time and compare to a properly designed benchmark to determine if they are skilled or just lucky.  Why you may ask?  Suppose that during a bear market (a relatively short period of time), a tactical manager had 70% of his portfolio in cash which saved his investors from experiencing significant losses – was the manager skilled or just lucky? In comparison, if the same manager over a 10-year period had an average allocation of 65% in cash then this would prove that he was in fact lucky purely because he favored higher cash allocations.  This would become painfully obvious to his investors as they drastically underperformed during times when risky assets did well. This is one of the many important things to consider in performance analysis and attribution. What we can learn from this simple example is that if we want to compare apples to apples it is important to compare the average allocation in the underlying investment versus the allocation to cash.

The same principle can be applied to momentum or trend following strategies. After all, the premise of these strategies is to primarily avoid bear market or recession risk. That means that they will invariably spend long periods in cash when the market is going down. There are two classic strategies discussed in the investment literature:

• TREND FOLLOWING- 200-day Simple Moving Average strategy (200-SMA): this is essentially a “trend-following” strategy that buys when the price of the S&P500 (SPY) is greater than the moving average (or trend) and sells when the price is below the moving average. Traditionally this is traded using daily data.

• MOMENTUM- 12-month or 252-day Absolute Momentum or Time Series Momentum strategy (252-TSMOM): this is essentially a “momentum” strategy that buys when the 1-year or 252-day return of the S&P500 (SPY) is greater than the return on cash or short-term treasurys (SHY). Traditionally this employs monthly rebalancing.

These strategies (like any approach) cannot possibly perfectly identify when the market is going down or up. They will end up spending a fair amount of time in cash when you would have been better off being invested.

The actual allocation to cash will vary as a function of the time period that you investigate. If the time period is only a bull market, then they will tend to have lower cash allocations. If the time period contains primarily bear markets, then a momentum strategy will tend to have fairly sizable cash allocations. That is why it is important to assess these strategies over a full market cycle that includes both bear and bull markets to make any honest comparison. Using a 2 or 3-year period when the S&P500 was only going up to perform benchmark analysis is not going to tell you whether the strategy is useful or not.

The proper comparison/benchmark for these strategies is to look at a passively rebalanced asset mix that is invested equal to the average exposure of the momentum or trend-following strategy.

Let’s look at the average allocation of a typical momentum strategy (252-TSMOM):

Notice that the momentum strategy spends nearly ¼ or 25% of the time holding cash. Given that this 20+ year period covers a range of both bull and bear market periods, this is a good sample to make a proper benchmark comparison. What about the trend following strategy? Let’s look at the average allocation of a typical trend-following strategy (200-SMA):

Again we see similar statistics which is not surprising. The trend-following strategy also spends nearly ¼ or 25% of the time holding cash.

To construct our proper benchmarks, we will simply compare a fixed weight portfolio that has the same average cash levels (or risk exposure) as each strategy and rebalance on the same time frequency. For the momentum strategy this means that the benchmark will use a fixed weight portfolio of 76%/24% mix in S&P500 and cash and rebalance monthly. For the trend-following strategy this means that the benchmark will use a fixed weight portfolio of 74%/26% and rebalance daily. Now let’s look at the final comparison:

Here are some of the conclusions we can draw from this analysis:

• Momentum and trend-following outperform: Both momentum and trend-following outperformed their proper benchmarks. This shows that these strategies have the potential to outperform a comparable passive asset allocation.

• Momentum and trend-following don’t reduce conventional measures of risk: While the volatility of momentum and trend-following is lower than buy and hold their volatility is not significantly different from their benchmarks

• Momentum and trend-following significantly reduce downside risk: Both momentum and trend-following had much lower downside risk than their benchmarks when looking at maximum drawdowns (Max DD)

• Better risk-adjusted returns: Both momentum and trend-following had better risk-adjusted returns (Sharpe ratio) than their benchmarks

• Momentum and trend-following exhibit asymmetry: they have low comparable downside risk, but good upside capture. In contrast both the benchmarks and buy and hold have symmetric risk.

• Lag in Bull Markets: Analysis shows that momentum and trend-following have lower upside capture than their benchmarks, and significantly lower upside capture versus buy and hold—this means you can expect to lag during bull markets.

• Thrive in Bear Markets: Analysis shows that momentum and trend-following have much lower downside risk than their benchmarks and especially vs buy and hold. The gap is much wider for downside risk improvement versus upside capture. This means that you can expect better performance during bear markets and that most of the strategy outperformance will arrive during bear markets.

Conclusion

The only way to properly benchmark momentum or trend-following portfolios is to use a comparable passive or fixed weight portfolio that has the same average cash (or risky asset) exposure. Results show that while momentum and trend-following strategies outperform their benchmarks, they do not -contrary to popular wisdom- reduce volatility. However, most importantly they do significantly reduce the most important investor measure which is downside risk or maximum drawdowns. Careful analysis shows that these strategies exhibit an asymmetric returns profile:  historically they have actually underperformed their benchmarks in bull markets but more than make up for this by significantly outperforming in bear markets.  A true understanding of the nature of these strategies is necessary as an investor or as an advisor in order to know what you can expect going forward. Don’t compare these strategies to passive asset allocation during bull markets—your goal as a momentum or trend-following investor is just to capture a portion of the upside. Where you will likely shine is when the market falls into a recession and you are safely tucked away in cash during turbulent waters.