Scalper1 News
Summary Momentum strategies have worked exceedingly well since 2008. It takes some effort to find a diversified portfolio for which momentum strategies fail. Adaptive asset allocation based on portfolio optimization with high volatility target also fails when momentum strategies fail. Adaptive asset allocation based on portfolio optimization with low volatility target performs well even when momentum strategies fail. Momentum strategies are very popular and are readily available at no cost on the internet. In fact, it takes some effort to find a well diversified portfolio of equities and bonds that would have failed. I used the “dual momentum” and the “relative strength” timing models on the portfoliovisualizer.com site and run a sequence of simulation on some ETF portfolios that included stocks, bonds, real estate and commodities. The portfolio I selected for the study is made up of six ETFs and it performed poorly for the momentum strategy with any look back period. As a benchmark we analyze the performance of the portfolio with equal weight targets, rebalanced when the allocation of any asset deviates by more than 20% from the target weight. That portfolio was subjected to 21 rebalancings within the time interval of the study from January 2007 to September 2015. In this article I compare the momentum strategy with the adaptive allocation strategies I described in many previously published articles. We investigate two versions of the strategy: a return maximization with a low volatility target, and another with a high volatility target. The version with low volatility target was subjected to 105 reallocations of the assets, virtually almost every month. The version with high volatility target was subjected to only 52 reallocations because it was allocated, on average, about two months to the same asset. Here is the list of securities used to build the portfolio: SPDR S&P 500 Trust ETF (NYSEARCA: SPY ) iShares U.S. Real Estate ETF (NYSEARCA: IYR ) SPDR Gold Trust ETF (NYSEARCA: GLD ) T he United States Oil ETF, LP (NYSEARCA: USO ) iShares 1-3 Year Treasury Bond ETF (NYSEARCA: SHY ) iShares 20+ Year Treasury Bond ETF (NYSEARCA: TLT ) The data for the study were downloaded from Yahoo Finance on the Historical Prices menu for SPY, IYR, GLD, USO, SHY and TLT. We use the daily price data adjusted for dividend payments. For the adaptive allocation strategy, the portfolio is managed as dictated by the mean-variance optimization algorithm developed on the Modern Portfolio Theory (Markowitz). The allocation is rebalanced monthly at market closing of the first trading day of the month. The optimization algorithm seeks to maximize the return under a constraint on the portfolio risk determined as the standard deviation of daily returns. In table 1 we list the total return, the compound average growth rate (CAGR%), the maximum drawdown (maxDD%), the annual volatility (VOL%), the Sharpe ratio and the Sortino ratio of the portfolios. Table 1. Performance of the portfolios from January 2007 to September 2015. TotRet% CAGR% maxDD% VOL% Sharpe Sortino Equal Weight 36.95 3.70 -35.85 10.46 0.32 0.42 AA LOW volatility 65.03 5.96 -11.05 6.02 0.99 1.32 AA HIGH volatility -4.73 -0.56 -55.18 23.19 -0.02 -0.03 The data in table 1 should be compared to the results of applying the dual momentum strategy as computed with the portfolio visualizer application. The dual momentum strategy investing monthly in the asset with the highest return over the previous 3 months had total return of -10.34%, with CAGR of -1.25%, maximum drawdown of -40.88% and volatility (St Dev) of 20.48%. There were two periods when the momentum strategy suffered huge losses; first in 2011-12 after gold topped, and the second in 2014-15 when oil prices tanked. The AA high volatility results are very similar to the dual momentum results. Most of the difference in drawdown and volatility is due to the fact that I use daily closing data while the portfolio visualizer site uses monthly data. That explains the slightly larger volatility and drawdown of the AA high volatility compared to the dual momentum. The small difference in the total return is due to a different allocation of the two strategies during a few months in 2011, as will be seen in figure 2. Of the three strategies, the AA with low volatility target performs the best both in return and risk. It produces a steady return of about 6% annually with a low volatility of only 6% and a maximum drawdown of -11%. The performance of the equal weight strategy falls in the middle; it returns on average almost 4% with low volatility of 10%, but still rather large drawdown of -36%. The equal weight strategy suffered steep losses during the 2008-09 bear market. In figures 1a and 1b we show the historical allocation of assets for the adaptive allocation strategy. (click to enlarge) Figure 1a. Historical asset allocation for the low volatility target portfolio. Source: All the charts in this article are based on calculations using the adjusted daily closing share prices of securities. As can be seen in figure 1a, the portfolio was allocated to SHY about 50% over the entire time. It was also allocated about 25% each to SPY and TLT. There were only small allocations to gold, oil and real estate. (click to enlarge) Figure 1b. Historical asset allocation for the high volatility target portfolio. Here one sees that the high volatility target portfolio was allocated alternately to one asset only, the same as in the momentum strategy. Only for a few months in 2009 was the portfolio invested in two assets simultaneously. In figure 2 we show the equity curves of the adaptive allocation portfolios. (click to enlarge) Figure 2. Equity curves for the adaptive allocation (NYSE: AA ) portfolios. We see in figure 2 that the high volatility target portfolio performed well until the fall of 2011. Since then, the equity either went down or oscillated in a range. Recently the equity fell below the initial investment. In figure 3 we show the equity curves of the low volatility and equal weight portfolio. (click to enlarge) Figure 3. Equity curves of the adaptive allocation with low volatility target and the equal weight portfolios. We see in figure 3 that the equal weight portfolio suffered large losses during the 2008-09 financial crises. It performed well between 2009 and 2012, but it fluctuates in a range since 2013. Still, overall, the equal weight portfolio performed better than the adaptive allocation or momentum strategy, as can be seen in figure 4. (click to enlarge) Figure 4. Equity curves of the adaptive allocation with high volatility target and the equal weight portfolios. Source: All the charts in this article are based on calculations using the adjusted daily closing share prices of securities. Conclusion The adaptive allocation by portfolio optimization with low volatility target performs satisfactorily during all market environments. Over a long investment horizon, it beats the equal weight as well as the momentum strategies. Scalper1 News
Scalper1 News