This article will discuss a simple method to use moving averages to increase a portfolio's return, while reducing risk, and drastically reducing drawdown. When building a portfolio there is a lot to say about diversifying the funds in the portfolio. According to the International Monetary Fund, the United States is poised for a 2.6% growth in both 2016 and 2017, while the emerging market as a whole is poised for a 4.3% and 4.7% growth in 2016 and 2017 respectively. However, over the past 5 years some emerging market funds have lost close to 50% of their value, while the S&P 500 has increased close to 50%. The bond market has grown in value over the past 10 years, but many investors are expressing concern over the bond markets performance as rates begin to rise. These types of considerations are the rational behind diversifying a portfolio, one fund may not be doing well but other funds may be doing great, one fund may not have much growth over the next 10 years, but another fund has huge growth potential. So by stacking different kinds of fund, AKA diversification, we can gain exposure to a wide range of different markets growth potentials.
On the other side of the diversification coin, to gain exposure to different markets upswing, we also have to be involved in their respective down turns. So the question comes to mind, "Can we avoid the down turns in a market, but be involved in the upswing of each market?". Meb Faber introduced this topic to a wider audience in 2006, with his paper "A Quantitative Approach To Tactical Asset Allocation". In this paper he tested applying a moving average to funds within a portfolio over the past 30 years to see if this simple method could be valuable in avoiding drawdowns while gaining exposure to a diversified portfolios upturns. In our article we want to backtest a simple, and practical method for building a 2 fund portfolio that is 60% stock, 40% bond, and then seeing if the moving average can benefit this simple and popular portfolio.
The 60/40 portfolio we will test from January 1992 to February 2016, the portfolio will consisting of US Stocks (VFINX - S&P 500 Index Mutual Fund) 40%, and longer term treasury bonds (VUSTX - Long-Term Treasury Mutual Fund) 40%. When any of these funds falls below the 200 day simple moving average, we will de-invest in the respective fund, and instead invest in a short term bond, cash like fund.
Buy and Hold Portfolio Results
For our results we will use yearly rebalancing, in blue, as it is the highest return and most normal rebalancing interval. Note, without rebalancing (yellow line), the performance is much worse, and the drawdown much higher.
Results: 9.01% CAGR, 0.89 Sharpe, 10.24% daily volatility, 29.68% maximum daily drawdown.
Adding Market Timing to Each Fund
For the below results we applied a 200 day simple moving average to each of the funds, if it falls below that fund we instead invest in a cash fund to try to avoid that market funds drawdown.
For these results we are using the cash filter fund: VSGBX (Short-Term Federal Mutual Fund).
Results: 9.29% CAGR, 1.17 Sharpe, 7.89% daily volatility, 9.98% maximum daily drawdown.
One thing you will notice is that the cash filter is a very important part of the portfolio, see pie chart below, the cash filter accounts for approximately 1/3 of the total investment duration. Since the cash filter is very important part of the portfolio, lets try an intermediate term bond fund, instead of short term one, to see if that can better our results with a higher yielding fund.
Adding Market Timing to Each Fund + Intermediate Term Bond Cash Filter
For the below results we applied a 200 day simple moving average to each of the funds, if it falls below that fund we instead invest in a cash fund that will be an intermediate term bond fund instead of the short term bond fund used in the last backtest.
For these results we are using the cash filter fund: VFITX (Interm-Term Treasury Mutual Fund).
Results: 10.23% CAGR, 1.2 Sharpe, 8.45% daily volatility, 10.17% maximum daily drawdown.
|Buy and Hold||Market Timing||Market Timing with Intermediate Term Bonds|
|Annual Return %||9.01%||9.29%||10.23%|
Length of Drawdown
When we zoom in on the drawdown chart we can see the maximum duration of drawdowns is very different for the buy and hold portfolio as compared to the market timing portfolio. The portfolio timing results in a much shorter drawdown duration, vastly improving the length of time a the strategy equity is in the red.
Buy and Hold Results
The maximum drawdown duration was from September 1, 2000 to Jan 12, 2004 - more than 3 years.
The maximum drawdown duration was from January 31, 1994 to April 13, 1995 - a little over 1 year. The drawdown period from 2000 to 2004 that the Buy and Hold portfolio suffered from was much shorter than this, and each of those years was positive for the portfolio timing tool.
Comments on Results
The worst year for the moving average timing strategy was -5.6%, compared to -17.7% on the buy and hold strategy. The best year for the moving average timing system was 33.7% in 1995, while the best year for the buy and hold portfolio was 34.6% in 1995. Our moving average was able to drastically limit our exposure to down markets and drawdown, but still gave us almost complete exposure to the great returns of 1995. This pattern of gaining exposure to a majority of uptrends, and avoiding downtrends has been a pattern. In the bear market of 2000-2002, when the buy and hold portfolio lost money for every one of those 3 years, the moving average timing strategy gained money every one of those 3 years. Also we experienced a drawdown duration that was drastically shorter with the portfolio timing strategy, sitting in drawdown for 1 year is a lot better than the 3 years of drawdown the buy and hold portfolio suffered from.
One thing we can note about the performance of the portfolio is that it is only as good as the funds you are investing in. If both long term bonds and stocks were on a downward spiral, the best this strategy could do is invest in the cash fund and get its low returns. So the importance of diversification is not eliminated by using this strategy in any way, if there is no upside to be had on what you are investing in, there won't be much of an upside with or without this strategy.
|Pros of Portfolio Moving Average||Cons of Portfolio Moving Average|
|Avoid Each fund's Drawdown||Miss out on some of each fund or market's recovery to the upside|
|Get out of the market and into cash during market down trends.||The cash filter fund is invested ~1/3 of the time, thus it is a very important choice for the portfolio.|
|It is possible to see gains in down years like 2000-2002.||If funds are not diversified, the strategy has little chance of seeing gains in different market regimes.|
|The duration of the drawdown is much lower than the buy and hold portfolio|
The popular 60/40 Stocks/Bond portfolio performs well over the past 24 years, but adding a simple moving average to this portfolio has increased returns, reduced the duration of drawdowns, and substantially reduced portfolio drawdown. Adding in an intermediate term bond fund as the cash fund accomplished even more, it increased annual returns more than 10% over the buy and hold portfolio, while having close to 1/3 of the daily drawdown numbers. Avoiding drawdown and still being involved in market upswings was the goal of this strategy, and it worked well in this instance. There were a few concerns, namely being involved in the cash filter fund for too much duration, not being diverse enough to capitalize on gains across different markets, and the potential of missing out on some of the market upsides. However, these concerns did not prevent us from accomplishing the goals of reducing drawdown and risk along with increasing return in this particular example.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.