It’s been a while since I’ve write about TAA Strategy Performance so I thought it was high time for an update. I’ll compare mainly long term performance but also the consistency of that performance of about 90 TAA strategies. Also, I’ve made some some small but important changes to my core TAA models over the last couple of years which I’ll share in this post as well. Let’s jump right in.
Data sources for comparison are very important in benchmarking an investment model and TAA is no different. For this post I am using my own data and data from Allocate Smartly which keeps track of a few of my models as well as the others I’ll be using in my comparison. They have historical data and actively track about 90 TAA models. They are the best data source for TAA model data. In this post when I use data for models that are not mine I only show relative ranking for those models and not the actual data. For the actual data I recommend becoming a member of Allocate Smartly.
When I do my TAA comparisons my favorite metric to use is average returns across several time periods. First, returns and consistency of returns are among the most important metrics to consider in choosing a TAA strategy. To me they are more important than even risk metrics, like sharpe and UPI. Those are important too but not as much as returns. I look for TAA strategies with ‘acceptable’ drawdowns and risk adjusted metrics but among the ‘acceptable’ universe I favor high returns. Outside of returns I like to focus on ease of trading, number of trades, and ease of implementation. Just like buy and hold, many investors that try out TAA end up not being able to stick with it and end up switching among TAA strategies, chasing performance.
In the table below I am showing the top 20 TAA strategies ranked by average return over the full backtest period (usually from 1971), 20 year returns, and 10 year returns. A high ranking not only shows high returns but consistency of returns over those three time periods. Also, as mentioned above data is shown only for my strategies, those in bold, (which are part of the Economic Pulse Newsletter). I have also shown two benchmarks in the table; the traditional US centric 60/40 and a global version of 60/40. The green highlight means that the strategy outperformed 60/40 in that period.
Some observations from the data in the table. There are 38 strategies out of the 90 (~42%) that have beaten 60/40 on average over these three time periods. I am only showing the top 25 of those. 60/40 has been especially hard to beat over the last 10 years with the dominance of US large cap stocks. 21 of those strategies (~23%) have managed to beat 60/40 over all three time periods, not just on average, with 5 of those being mine. Only GPM has underperformed over the most recent 10 year period. Not too bad. Also, risk-adjusted performance for all these top 25 strategies is pretty good, or as I would say, ‘acceptable’. One of the reasons I say that is that with judicious strategy design and choice you can put together a portfolio of strategies that is more than the sum of it’s parts and it is the performance of the combination that matters more.
As an example of such a combination, in the table I listed an equal weight combination of three of the Economic Pulse Strategies; DM-MODCOMP, FAST, and GPM. The combination of those strategies has both good and consistent returns, and great risk-adjusted metrics. The improved outcome of the combo comes from the low correlations among these strategies which comes from the fundamental differences in how these strategies make their risk-on and risk-off decisions. The combo also happens to be pretty easy to implement with a maximum of 5 ETF holdings. These kinds of choices are as important as the individual strategies themselves. In fact, the changes I’ve made to the Economic Pulse Strategies over the last 2 years have lowered the performance of the individual strategies themselves, but have made them more robust for the future and has lowered their correlations to other strategies, thus making a portfolio of such strategies perform better. Let me use SPY-COMP as an example.
The SPY-COMP TAA strategy is my original TAA strategy and continues to be one of the best performing as you can see in the table above. But it wasn’t 100% robust. Half of the time the strategy was not price protected, it was basically buy and hold. This turned out to be fine over history but in the future left it exposed to some big exogenous event. Also it made it harder to stick to during tough times. So, I made a couple of changes to the strategy to make it protected 100% of the time. That is now the SPY-MODCOMP strategy. This lowered historical returns a bit but improved risk-adjusted performance, future proofed the strategy, and has given subscribers more confidence in the strategy. By the way, this exposure to exogenous events, which comes from dependence on fundamental factors, is a feature of several high performing TAA strategies which you should be aware of when choosing among them. In my opinion a TAA strategy should be price protected 100% of the time.
That’s about it for this update on performance. 60/40 has proven to be a tough benchmark to beat over the long haul, both in terms of returns and risk-adjusted metrics, but there are very good choices among TAA strategies that have been able to consistently outperform over the long term. Those strategies have a few things in common which will be the subject of another post. Also, I will address ease of implementation which becomes one of the major issues when using a TAA strategy over the long term.