A systematic strategy on the S&P 500: exposure from 200% long to cash and short, set once a month by leading U.S. economic indicators. Leverage works in good times; the strategy met 2008, 2020 and 2022 in full defense.
10 leading U.S. macro indicators — employment, jobless claims, housing, inflation, rates. The influence of each is learned from the last 10 years of data, not assigned by theory; an indicator whose link to the market is indistinguishable from noise is switched off.
Four states: +200% (clearly bullish macro), +150% (normal), cash in T-bills (bearish macro), 75% short (extreme case). Rebalanced once a month; the state changes 1–2 times a year.
Each type of bear market has its own detector: recessionary (2008) — the wave of layoffs; inflationary (2022) — price dynamics; price-driven (dot-com) — a trend filter. Defense is lifted when its cause fades, not on hope.
Executed with S&P 500 futures — the most liquid instrument in the world. Capacity is practically unlimited; turnover costs are negligible.
| Year | Strategy | S&P 500 |
|---|---|---|
| 2008 · global financial crisis | +3.3% | −36.8% |
| 2020 · pandemic | +45.1% | +18.3% |
| 2022 · inflation | +11.2% | −18.2% |
Defense is the source of the result: the strategy carries more exposure on average than a constant 1.5×, yet met the three largest crashes of the era in cash or short.
| 2000–2026 | CAGR | Max DD | $1 → |
|---|---|---|---|
| Strategy (leverage with an off switch) | 19.2% | −31% | ×108 |
| S&P 500, constant ×2.0 | 11.9% | −85% | ×20 |
| S&P 500, constant ×1.5 | 10.4% | −72% | ×14 |
| S&P 500, no leverage | 8.4% | −55% | ×8.5 |
Constant leverage does not scale returns honestly: volatility and financing eat the return while drawdowns grow faster — after −85% it takes +550% to recover. Double leverage without timing loses to the strategy on both return and risk.
Only 20% of the outperformance over the S&P 500. No management fee, no other charges: if we don't beat the index, you pay nothing.
The core of the capital sits in an Irish UCITS S&P 500 fund — held for years, dividends accumulate, outside the scope of U.S. estate tax. Positioning is done with a futures overlay; no withholding along the way.
The world's most liquid market, one rebalance a month, capital transparent and liquid on any day. Total implementation cost ≈ 0.5 pp/yr versus model figures.
Worst day −11%, worst month −16%, worst drawdown −31% on daily data (long-run monthly model shows up to −37%). If those numbers are too much, the same strategy scales down to a smaller size.
A typical weak year is not a loss but lagging a booming market after a crisis: 2023 returned +16.8% against +26.2% for the index. Defense is lifted on data — and data sometimes comes late.
The signal was re-run on first-print macro data — as a manager would have seen it in real time, before revisions. The core result survives; the edge does not rest on three lucky years: without them, +7.4 pp/yr.
Results are based on historical simulation and are not a guarantee of future returns. The methodology is fixed and frozen; changes are made only as documented appendices.