Composer
200d Moving Average 3x
Robust
“In uptrends it splits between 3x tech and 3x S&P 500. In downtrends it buys oversold bounces.”
From Composer
As tested: Daily. SPY above its 200-day → 50/50 TECL/UPRO. Below: QQQ 10-day RSI under 30 → TECL; SPY's under 30 → UPRO; QQQ under its 20-day → the higher-RSI of SQQQ and SHY; PSQ's RSI under 30 → SQQQ; else TECL.
Holding today: TECL 50% · UPRO 50% · last switch Apr 8, 2026
Uses leveraged funds. The test covers 2016–2026, one long Nasdaq bull market; it has never seen a 2000–02 or 2008, where 3× funds lost 90%+.
Edge per month+4.54 pts/mopts per month vs SPY
t-stat3.9295% range +2.27 to +6.81
Months120Nov 1, 2016 – Oct 12, 2026
CAGR74.4%SPY 15.8%
Max drawdown-62.8%SPY -33.8%
Since published+3.21 pts/mo48 months from Oct 7, 2022; +5.29 pts/mo before
Last 12 months+68.24 pts12 months, summed
Why robust
- t-stat 3.92 clears 2.0
- the edge stays positive in 95% of resamples
- 100% of nudged versions keep the edge, so it isn't one lucky setting
- the edge held after publication: +3.21 since it was published in 2022-10, vs +5.29 before
Growth of $1
200d Moving Average 3xSPY
Log scale. Traded at the next open after each signal, daily rebalance, slippage on every trade.
Year by year
2016
+3.79 pts/mo2
2017
+4.28 pts/mo12
2018
-2.62 pts/mo12
2019
+5.97 pts/mo12
2020
+4.62 pts/mo12
2021
+4.56 pts/mo12
2022
+10.56 pts/mo12
2023
+3.69 pts/mo12
2024
+2.12 pts/mo12
2025
+5.18 pts/mo12
2026
+7.70 pts/mo10