When Strategies go Bust
What would you do if I told you that you could have earned an average of about 50% annually over the past twenty years? Would you believe me? Would you go out and invest in the strategy? Above we have a screen that has increased by over 600,000% over the past 20 years. We’ve gone through and verified the trades and these results are actually legitimate.
I haven’t been building investing strategies for long – roughly two years. And when you first begin learning about Quantitative investing, you’re told one thing more than anything else. One day, your strategy will die. I didn’t believe that back in 2015 when my portfolio took a hit because I was heavily invested in biotechs at the time, and now, a lot of people who are using our system don’t believe it. Well here is a perfect example of a Screeners death.
What’s the Screener?
This screener is actually a variation of our most recently featured screen, “Ranking the Technicals”. Hoewver, we’ve changed the order by away from the original
(income statement score + (num times large change*-1))
So we are taking the top 5 companies that have the highest SCTR and investing in them for three months.
Looking at the backtest by time, you can see just how well this strategy did up until 2014. The average year is almost up to 50, but once 2014 hits, it begins to underperform the S&P. Yes, it does well in 2015, but I think we would all like something that stays consistent throughout most every year.
Isolating the recent years
If a strategy looks like it might not be performing as well in recent time, you’ll want to isolate that time to analyze the current viability of the screener – especially if you’ve validated the longer term aspects of the strategy.
In the case of ranking the technicals with the adjusted order_by, we can see that over this time period, our strategy did not beat the market. Worse yet, is that we didn’t beat the market and had a standard deviation of double the market making our strategy extremely volatile for the minor return that we got.
It actually looks like it works so much more poorly than the S&P that is has a sharpe ratio of half. Remember, you have a sharpe ratio as high as possible when identifying the risk/return of the strategy.
Basically, you want to go through an entire validation process for the recent years to check every possible spot for viability.
The Strategy is broken, what next?
There are two simple solutions to your investment strategy going dead.
1) Delete it and start from scratch
This is the most common procedure. Once you identify the reason that this strategy stopped working, you want to make sure that your next strategy isn’t correlated at all. Say you have a tech strategy that has been doing poorly for the past three years; you don’t want to build another tech strategy because the error may be the industry rather than how you are picking companies.
2) Adjust the existing strategy to eliminate the failing parameters
A bit more complicated, but a lot of screeners have really good fundamental reasoning behind them, and it isn’t likely that this reasoning will ever fail. A company that is making money and making the right business decisions will typically continue to do so unless they jump head first into the fan. So identify the part that is making the strategy fail and adjust that according to current market conditions.
None of us ever want our strategies to stop working. We are usually invested in it when it happens, and losing money is never fun. The best thing you can do is be prepared.