Imagine a trading system that allows one to capture a majority of the stock market’s uptrend while limiting drawdowns to roughly one-third of portfolio value versus fifty percent or more for the market. Imagine that to execute this system one needs to spend only a few minutes each month performing some basic calculations to determine the trading signal, decide which risky or risk-free asset to hold, and execute the required trades the next day. Lastly, imagine a system that is not time-consuming to monitor and run, requires minimal trading, and employs liquid trading vehicles.
Well, there’s no need to imagine because that system exists today, and it’s called Dual Momentum. In this article I’ll explain how Dual Momentum works, point out its key advantages, and why you should trust the process and consider replacing your current buy and hold or 60/40 strategy with Dual Momentum.
What is Dual Momentum?
The standard dual momentum model uses relative momentum to select the best performing model assets and incorporates absolute momentum as a filter to invest in cash if the excess return of the selected asset over cash is negative. The following flowchart depicts the standard implementation of the Dual Momentum system, using the S&P 500 and the All Country World (excluding U.S.) Index as the "risky" assets, and US T-Bill as the "risk-free" asset. The corresponding ETF ticker symbols are shown in parenthesis.
Source: Author’s chart.
Test Results Which Influenced Our Implementation
The following chart and table correspond to the standard implementation (defined above) using $10,000 as the starting balance. For both practicality and realism reasons, trades are made at the closing price of the first business day following month end:
The following chart and table correspond to the implementation that we’ve been using since 2015:
All returns represent back tested results and are hypothetical (not based on actual trades) and assume the reinvestment of all dividend and any capital gains distributions that were made. All returns are gross of fees except for underlying ETF expense ratios which are adjusted daily in their pricing. All trades were assumed to have been made at the closing price of the next day (more realistic although not perfect). Comparison is made against a 60/40 stock-bond portfolio as represented by the Vanguard Balanced Index Fund Institutional class fund (VBIAX). We chose the 60/40 portfolio for comparison because it more closely adheres to the pattern that Dual Momentum is in the market only about 50% to 80% of time.
Firstly, our research indicates that stock markets outside of the U.S. are not more compelling as an investment. Specifically, we don’t see strong evidence that a collection of western developed markets can generally perform better than the U.S. on average over long periods of time. And when they do it’s because the U.S. stock market was performing extraordinarily well.
Secondly, and similarly, we’ve found emerging and developing country stock markets strongly outperform predominantly when the U.S. stock market was performing extraordinarily well.
Our research has also found that foreign markets have tended to be more cyclical than the U.S. market. In other words, the U.S. stock market can be prone to long and secular uptrends (punctuated by mostly cyclical bear markets), but which are not exhibited to the same degree by developed western markets, and certainly not by developing or emerging markets.
Why We like Dual Momentum
Full Confession: We are predominantly trend following traders.
That is, we go long things that are rising in price, and go short things that are falling in price. We don’t try to second-guess why as we believe the market, as reflected by an asset's price, is the final arbiter. Furthermore, we don’t overcomplicate things by using moving averages, oscillators and other forms of technical analysis which have been shown not to work over the long term.
Dual Momentum fits in nicely with our way of thinking. Why? It starts by comparing the asset price today versus the asst price from 12 months ago. If the price difference in percentage terms is positive, then it’s assumed the asset is in an uptrend. Similarly, when an asset’s price is lower than it was 12 months ago then it’s assumed the asset is trending downward. It’s as simple as that.
The “momentum” aspect of Dual Momentum allows for the fact that assets may not be trending, relatively speaking, to the same degree. Thus, it picks the asset, risky or risk-free, exhibiting the strongest trend. That is, the asset exhibiting the most oomph!
We also like that the trading signals are generated monthly and that it’s low maintenance and low cost to run. This means we don’t have to waste time monitoring the markets on a daily basis. Once we’ve chosen the risky and risk-free assets it only takes a few minutes per month to generate the trading signal and execute the trades. And because we’ve implemented it using liquid ETFs, trading commissions, bid/ask spreads, and fund expenses are kept low.
Why we prefer QQQ over SPY for Dual Momentum
While we are not against an investor implementing SPY as his or her risky asset, we much prefer using the Invesco QQQ ETF (QQQ) instead. The main reason is that QQQ represents large non-financial companies in industries that represent growth and the new economy. By not using SPY we avoid the negative influence of so-called “value stocks” that are included in the S&P 500 index, along with banks and financials which are interest rate sensitive.
The low level of interest rates, the risk of curve inversion, and the continuing tug-of-war taking place between current monetary policy and modern monetary policy will take years to play out. While that happens, we believe that it creates an environment in which banks, financials and value stocks will continue to underperform. We want to avoid this situation.
Moving on, the following chart demonstrates the long-term relative outperformance of QQQ over SPY:
Lastly, for the risk-free asset we've chosen to hold the Vanguard Total Bond Market ETF (BND) instead of the SPDR Bloomberg Barclays 1-3 Month T-Bill ETF (BIL). We still generate our trading signals using BIL. However, we instead take a position in BND when the system calls for switching out of stocks (risky asset) and into cash/bonds (risk-free asset).
We do this because our testing has shown that returns can be increased by 50-70 basis points per annum simply switching to a low-cost intermediate term bond fund whose yield is significantly higher than BIL's.
Drawdowns and Whipsaws Appear Acceptable
During this current ten-year bull market in stocks we’ve not yet seen a significant price drop of comparable severity to that of late 2007/early 2009. To make the point visually, see the following chart where Dual Momentum lost less than half of what the S&P 500 (VFIAX) lost.
Despite the Dual Momentum (Timing Portfolio) making marginally less over this time period than the Market Portfolio (Vanguard 500 Index Admiral), its drawdown was less than half the Market’s!
Studies put out from other other Dual Momentum practitioners (i.e. Alpha Architect, Newfound Research) typically show drawdowns in the 18% to 33% range for S&P 500 implementations of the system. Relatively low drawdowns is a key feature of the Dual Momentum strategy and one that we believe goes underappreciated when equity markets are doing well. The significantly lower expected drawdown is what allows us to sleep well at night, knowing our worst case should not be as bad as the market’s.
It’s also worth pointing out that Dual Momentum avoided the Q4’18 market whipsaws (both under the standard implementation and our own), having only exited the risky asset in January this year before getting back into it at the start of February, and remaining in the risky asset as of this writing.
Dual Momentum generally allows an investor to capture a majority of the uptrend in the major equity indices (sometimes even outperforming) while significantly reducing drawdowns when the market tanks. Given its monthly trading frequency, this system also appears to be less prone to whipsaws which can reduce returns and increase costs.
It's worth noting that should we ever see a Japan-like bear market situation in the US, or a repeat of what followed the Great Depression or Great Recession, Dual Momentum would allow an investor to side-step a multi-year loss situation. If and when trend and momentum were to finally turn positive for the risky asset (and exceed that of the risk-free asset), an investor would once again be participating in the risky asset's (i.e. QQQ, SPY) uptrend as dictated by the system.
While past performance is no guarantee of future results, we believe that ‘buy and hold’ and 60/40 portfolio investors should seriously consider adopting the Dual Momentum strategy as a means to capture most of what those strategies have to offer (mainly capturing the secular uptrend in US stocks), while avoiding most random whipsaws and steep drawdowns during prolonged bear markets. All of this comes with the added benefit of letting one sleep better at night.
Disclosure: I am/we are long QQQ.