Beating SPY with less risk: 16.5% a year, 15.3% volatility

The Orthogonal Portfolio is one idea. Hold assets whose returns come from different places, then use a leveraged S&P 500 fund to set how much risk you take.

Three things: a 2x S&P 500 fund, a trend-following fund that can go long or short across stocks, bonds, currencies and commodities, and gold. The exact funds and weights are for subscribers.

The 2x fund keeps stock exposure high while using less of the money. The rest sits in two things that don’t depend on company earnings.

I tested it from May 2019, when the trend fund launched, to November 2026. Rebalanced daily, total return.

It returned 16.5% a year. SPY returned 15.3%.

Volatility was 15.3% against 19.5% for SPY. The worst drop was 24.9% against 33.7%. The Sharpe ratio was 0.90 against 0.69.

The trend fund and gold moved almost independently of the S&P 500. Daily correlation to SPY was 0.18 for the trend fund and 0.14 for gold. That is what I mean by orthogonal. Three return streams that mostly don’t move together.

Risk is a choice here. I changed the size of the stock sleeve and kept the rest.

With a bigger stock sleeve: 18.1% a year, 18.2% volatility, 30.6% worst drop. Still ahead of SPY on all three.

With a 3x fund instead of the 2x: 19.8% a year. Volatility was 20.5% and the worst drop 34.4%. Same risk as SPY, more return.

It didn’t protect me in the March 2020 crash. The 3x version fell almost exactly as much as SPY. The help came in slower selloffs. In 2022 the 3x version lost 8.4% while SPY lost 18.2%.

The trend fund and gold also did unusually well in this window. The trend fund made about 9.6% a year and gold 16.6%. I wouldn’t expect that to repeat. The bigger-stock version beat SPY in 59% of rolling one-year periods. In the worst one it trailed by 10.6 points.

One window, May 2019 to November 2026. Daily rebalancing, before taxes. Fund fees and the cost of leverage are inside the fund prices. The trend fund didn’t exist before 2019, so I can’t test it in earlier crashes.

The edge came from adding return streams that don’t move with stocks. The leverage only sets the dial. What should I test next?

Historical backtest, total return. Not a promise of future returns.