Weekly Systematic Trading Update
Week ending November 21, 2025
Stocks (S&P 500 index) were down 2.9% by the close of Thursday, November 20, 2025, primarily due to concerns about market breadth and whether just a few companies could provide support to the uptrend going forward. On Friday, November 21, 2025, a “verbal intervention” by the New York Fed President about lower rates triggered a rebound, and the S&P 500 index trimmed the weekly gains to 1.95%.
The rising volatility impacted all strategies in the ensemble, while gold (IAU) and bonds (TLT) failed to provide a cushion, with the former falling 0.5% and the latter rising just 0.7%. As a result, all strategies fell this week. Dow-30 long/short fell the least, by 0.2%, and did not provide a significant hedge due to the whipsaw. The primary loser with the highest impact on weekly performance was Dow-30 mean reversion with a loss of 3.4% for the week after two stocks hit a stop-loss.
Although there may be a regime change for weekly mean reversion, we hesitate to declare the strategy has failed due to a 0.5% year-to-date loss. However, we also think it is reasonable to assume that mean reversion is under stress in the current regime, although that may be a transitory effect. Regardless, we plan to reduce mean reversion exposure to equities next year by replacing the SPY ETF strategy with another strategy that also has exposure to TLT and GLD ETFs and is up 4.6% year-to-date, while the Sharpe ratio is about 0.9 since 2005. In general, we only make infrequent changes to the ensemble and only when we deem they are absolutely necessary. Note that we have not backtested any new strategies in the past two years to avoid the adverse effects of data-mining bias. For more details about the risk of frequent backtesting, see this article.
Disclaimer: No part of the analysis in this blog constitutes a trade recommendation. The past performance of any trading system or methodology is not necessarily indicative of future results.
The discipline to not backtest new strategies for two years just to avoid data mining bias is rare. Most traders would have already pivoted three times by now chasing whatever worked last month. That Dow 30 mean reversion loss is interesting thoug, feels like the whole market structure has shifted with concentration risk so high. When only a handful of stocks are driving everythng, traditional mean reversion setups probably don't work the same way.