Here is the million-dollar question: When should we let go of an investment strategy?
Believe it or not, it is a lot easier to put strategies to work. Letting go of an investment strategy, however, is a much more daunting task. You might be wondering what is so difficult about it? If it is not working, just take it out. Isn’t that just common sense? Yes, the reasoning is sound. But the challenge is how do you know if the strategy is no longer functional?
The Painful Dilemma — To Let Go Or Not
Here is the scenario. You had a brilliant strategy which you used for quite a while and it made you good money. And you felt like you had the world in your hands. Then all of a sudden, things fell apart and nothing seems to go right. Your cash cow fell sick and you run into losses. But is this just a passing ailment or a terminal sickness? As hard as you try, you can’t uncover the root cause. So you are now hit with a painful dilemma. Should you cut the strategy? What if you cut and the strategy turnaround to make new highs? Or maybe you should just continue? But what if you continue and the strategy carries on its endless bleed? For those who have been through the process, I am sure this is all too familiar.
Each working strategy is the result of your hard work and effort. This can be months to years of effort starting from research, backtesting, and forward running before deploying it live. And if the strategy performed well after that for a good amount of time, it can be painful to consider dropping it when the situation arises. These are uncomfortable but necessary decisions professional investors got to deal with all the time.
Let History Be Your Guide
We don’t have answers to everything. More often than not, we got to make do with what we have and know. So if you did a comprehensive backtest, then that is the best starting point. And what is a comprehensive backtest? Besides realistic modeling, it should also cover a sufficiently long period comprising different market regimes. The reason is simple. We want to know how the strategy performs under different conditions. In particular, what are the risks, and what is the worst it has been through?
I have seen people showcasing backtests that are pathetically short like less than a year. I have also seen those who cherry-picked the periods to get a favorable result. This is not how backtests should be done. Backtests are meant to shed light on your strategy, both the good and bad sides. When you know your strategy inside out, you will be able to make more sensible decisions.
What Reference Can You Take From The Past?
There are no hard and fast rules when it comes to retiring a strategy. The considerations are usually a blend of science and art. Having said that, you can exercise the best discretions and yet still be wrong about it. That is the nature of the game. But what is more important is that you have a rational approach.
By no means exhaustive, here are some simple considerations:
1. Is the behavior of the strategy out of tune with similar periods in the past?
Holy grail strategies exist only in your dreams. There are only the right strategies at the right time. So all strategies will run into bad patches one day or another. A trend-following strategy suffers when the market swings sideways wildly. This occurs every now and then. Asset allocation strategies run into trouble when their correlations go to one and all head south together. Again, this is not new. It can happen when the market liquidity dries up under extreme pressure e.g. after the Lehman Brothers Collapse in 2008 and during the recent COVID-19 pandemic. Poor performance during periods unfavorable for your strategies are expected. As long it is within the norms, then it is not a cause for concern.
2. How deep are the losses compared against the past?
Now, it is one thing to expect losses during a bad patch. But that does not mean you should readily accept any level of loss. To gauge the level of acceptable loss, you can look at how much your strategy loses at its worst point in history. But since we are dealing with the future, it is always prudent to assume that you have yet to see the worst. So you might want to set a buffer over and on top of this worst loss. When that level is breached, it can be a signal to cut. Because you never know what can come next.
3. How long are the drawdown and recovery?
Another consideration is the inability of the strategy to recover from a drawdown. Of course, this might simply be a case of an exceptionally long bad patch rather than a fundamental breakdown. But in any case, we can never be sure and we can’t wait indefinitely. A slow bleed can be just as bad. Thus, after a certain point, it may make sense to drop it. Again, you can reference the longest drawdown in history to set the cut-off duration.
These are just some suggestions guided by quantitative measures. However, if you have a strong enough conviction that the strategy is still worth holding on, then there is nothing to stop you from doing so. Many long volatility funds underwent a period of slow bleed as volatility waned to record lows after quantitative easing since 2008. Their big break came after 2017 as markets went through a more turbulent period. But that said, they may still fall far short of recouping all their losses accumulated over the years.
Originally published at https://investmentcache.com on September 29, 2020.