Mean reversion · Intermediate · Weekly rebalance
Z-Score Mean Reversion
How it works, and when it doesn't.
The top 2 sectors by 3-month return have a 68% probability of remaining in the top 4 over the following month. Momentum persists longer at the sector level than at the stock level.
Z-score=0.98 — no extreme (±2σ required). RV=elevated. Cash.
Why it works
Sector rotation captures persistent institutional capital flows. When money rotates into technology or energy, it tends to stay for months. By ranking all 11 S&P sector ETFs by 63-day return and holding the top 2, this strategy systematically follows institutional momentum. The GBM model adjusts sector weights based on predicted leadership stability.
The signal logic
- Rank 11 SPDR sector ETFs by 63-day return: XLK, XLV, XLE, XLF, XLI, XLB, XLU, XLY, XLP, XLRE, XLC
- Hold top 2 sectors with equal weight (50% each)
- Only hold if the sector absolute 63-day return is positive
- Rebalance every Friday at market close
- GBM confidence score adjusts for sector leadership stability
When it fails
Fails at macro regime inflection points — when the strongest sector of the past quarter suddenly becomes the worst. Typically happens at economic turning points and rate cycle changes.
- Best market
- Bull regimes with clear sector leadership (2020 tech rally, 2022 energy, 2023 AI/tech)
- Worst market
- Whipsaw markets where sector leadership rotates every 2-3 weeks
Backtest
| Sharpe ratio | 0.26 |
| Annual return | 1.1% |
| Worst drawdown | -6.8% |
| Trades in the backtest | 64 |
Backtest on daily bars, positions taken the bar after the signal (no look-ahead).
Z-Score Mean Reversion, live.
Re-scored after every close.
The platform runs this and 32 other strategies against the live tape each day — with the evidence behind every number.