US Cross-Asset Trend Backtest: 17.9% Max Drawdown, About Half of SPY's Annual Return
SPY lost 37% in 2008. This strategy lost 1% that year. The price was an annual return about half of SPY's, 5.7% against 10.7%. US Cross-Asset Trend splits the money into five equal parts: US stocks (SPY), developed markets outside the US (EFA), intermediate Treasuries (IEF), US real estate (VNQ) and gold (GLD). Once a month it checks each one, and any fund that closes below its own 210-day moving average has its share moved to cash. Backtested from 2005 to 2026 at 0.1% per side for fees and slippage, its worst drawdown was 17.9%, against 55.2% for holding SPY. The price is return, which trails the index by a wide margin: 5.7% a year against 10.7% for SPY. What cross-asset trend following does The idea is old and simple. Do not guess which asset will rise. Own a bit of five kinds, but only keep the ones still in an uptrend. 210 trading days is about ten months, and a price above that line means the long-term trend is intact. Stocks, bonds, real estate and gold rarely crash together. In 2008, as stocks and real estate fell, it moved them to cash one by one. By the end of the year only a fifth of the money was still invested. It only buys, never shorts and never uses leverage. On average about 72% of the money is invested and the rest is cash. We scanned the average from 150 to 270 days with a threshold of up to 2% either side, and Sharpe stayed between 0.62 and 0.88. Every year of the backtest YearStrategySPY 2005 (from Oct)+3.3%+1.7% 2006+20.6%+15.8% 2007+10.4%+5.1% 2008−1.0%−36.8% 2009+17.0%+26.4% 2010+11.6%+15.1% 2011+4.2%+1.9% 2012+6.0%+16.0% 2013+6.3%+32.3% 2014+4.8%+13.5% 2015−6.4%+1.2% 2016+0.5%+12.0% 2017+7.7%+21.7% 2018−3.5%−4.6% 2019+10.1%+31.2% 2020+0.8%+18.3% 2021+10.0%+28.7% 2022−10.5%−18.2% 2023+3.2%+26.2% 2024+10.6%+24.9% 2025+19.8%+17.7% 2026 (to Nov 19)−0.1%+5.8% It beat SPY in only 8 of 22 calendar years (the first and last are partial), including the three down years for SPY: 2008, 2018 and 2022. Backtest results do not promise future returns. Where it falls less In the 2008 crisis SPY fell 55% at worst, while the strategy fell no more than 7% over the same stretch. In 2022, when stocks and bonds fell together, SPY dropped 24.5% at worst and the strategy 11.1%. Its deepest drawdown came in March 2020, at 17.9%. Prices fell faster than a monthly check can react, and all five assets dropped at once, gold included for a few weeks. Where it loses It loses on return, and by a lot. US stocks have had a rare bull market for over a decade. SPY rose more than 25% in 2013, 2019 and 2023, while this strategy made 3% to 10%. With only a fifth of the money in US stocks, it gets a small slice of any rally. Sideways years cost something too. In 2015 and 2018 several assets hovered around their averages, so it paid fees without avoiding much. The test period is limited. All five funds only existed from late 2004, so after warm-up the backtest starts in 2005 and covers just one long bear market like 2008. Who it suits People who put not falling too far ahead of making the most. If beating the index is the goal, this is not that strategy. In the backtest, SPY's annual return of 10.7% was nearly twice this one's. Running it yourself Add it to Blave Agent to rerun the backtest, change the average or swap some of the funds and see what moves. Multi-asset US strategies cannot trade automatically yet, so for now it is a backtest. See US Cross-Asset Trend in the strategy library →
Strategies like this can run on your Blave Agent.
Browse ready-made strategies with real backtests →