What a crazy month! Swings of over 4% a day have been the rule and US equities closed the worst quarter since 1987…
Panic selling led to margin calls for investors with leveraged positions, so they were forced to raise money from all sources including gold, government bonds and crypto to avoid liquidation.
When everything is sold at the same time, the correlation between assets increases and nothing works anymore as a hedge except cash. This is a situation we haven’t even seen back in 2008, where bonds still provided a decent hedge.
Our PentaTrader strategies have as well been put to the test again in the last weeks.
But let’s look at things relative to the market based on year-to-day returns:
Benchmarks (YTD)
- Dow Jones: -24.07%
- S&P500: -20.68%
- Nasdaq: -15.31%
- SMI: -12.97%
PentaTrader (YTD)
- Pro: +3.37%
- Classic: -12.55%
- MiniMax: -22.04%
- Universal: +7.05%
While all strategies except MiniMax still overperformed all main markets, I was of course not happy with the results at all. So I did some extensive research in the past weeks on how this type of risk can be detected and mitigated in the future. I had already tried in the past to include algorithms for stop loss orders on the different components, but this always resulted in increased overall volatility and reduced longtime performance of the strategies.
Then finally I noticed the main error in my thinking: all PentaTrader strategies are based on the statistics of past correlations, which is very reliable in normal market situations, but becomes unpredictable in turbulent times. In crazy situations like the recent past, many asset classes suddenly fall (or rise) at the same time, resulting in very unusual daily returns.
The solution is again statistics: I have now implemented an additional “circuit breaker” that triggers when we have jumps >3 standard deviations of usual daily portfolio returns. If that happens, we exit all positions at the close of the day and stay in cash until the beginning of the next month. All other selection criteria and calculations remain the same, no need for any changes there. It doesn’t make much of a difference in the long run, as there are occasional false alarms, but short term this approach definitely avoids big dips.
To illustrate the special situation we are in: since Nov. 2005 and for 3’627 trading days this “circuit breaker” would have triggered only 16 days overall, of which 6 times just in the last three weeks…
If I had implemented this at the beginning of the year, things would look different today:
PentaTrader (YTD, w. circuit breaker)
- Pro: +14.15%
- Classic: +3.87%
- MiniMax: +1.10%
- Universal: +6.83%
But even without this, the performance over the last 12 months is still very decent:
PentaTrader (1Y)
- Pro: +86.38%
- Classic: +32.72%
- MiniMax: +18.09%
- Universal: +12.76%
Past 12 month performance remains close to the average annual performance over the last 10 years:
PentaTrader (10Y average)
- Pro: +76.33% (Sharpe Ratio 2.302)
- Classic: +45.09% (Sharpe Ratio 1,898)
- MiniMax: +42.20% (Sharpe Ratio 1.838)
- Universal: +14.25% (Sharpe Ratio 1.749)
No matter how hard we try to keep a long-term perspective, it’s human nature to judge things based on what happens in the recent past.
It’s easy to forget during long bull markets that there’s more to successful investing than just watching returns go up. Only in the midst of bear markets we can practice the hardest part: maintaining a disciplined automated strategy and keeping emotions out of the equation.
Stay happy, healthy and wealthy!
Omar
P.S.: As of this month, I am also listing my other two strategies Pro and Universal. I have developed the Pro strategy last December as a refinement of the Classic strategy and started to use it in January. Personally I have now active investments in Pro, MiniMax and Universal and have discontinued Classic. Unless anyone insists, I will exclude it going forward.