Published
I've spent weeks on this blog telling you what doesn't work — scalping that fees eat alive, DCA that hides a loss behind a pretty average, 96% win rates built on martingale. So here's the other half of the story. After all of that, one idea kept passing every test I threw at it, and it's almost insultingly simple.
Hold the asset while its price is above the 200-day moving average. When price drops below, move to cash. That's the whole strategy.
The numbers, honestly
I ran it walk-forward — four out-of-sample folds, so the rule never sees the data it's judged on — on five years of daily SPY and QQQ, with a realistic 0.08% fee per switch. I let the moving-average length be picked by the in-sample fold from 50, 100, 150 or 200 days, then graded it on the unseen fold.
SPY: out-of-sample Sharpe 1.35. On the days SPY sat below its 200-day line — about a quarter of all history — the trend rule lost roughly −7% while buy-and-hold lost −26%. Across the full period its worst drawdown was −21.5% versus buy-and-hold's −51.6%. Half the pain.
QQQ: out-of-sample Sharpe 1.42. In the down periods, −4% versus buy-and-hold's −33%.
Read that again, because it's the whole point: in the good times this thing roughly matches just holding. In the bad times it loses about four times less. It doesn't beat the market by being smarter on the upside — it wins by not being in the building when the market burns down.
Why it works when cleverer things don't
Big crashes don't happen in a day. Markets roll over slowly, and the 200-day average is a crude but honest line between "uptrend" and "everyone's heading for the exits." The rule keeps you invested for the long grind up and quietly steps aside for the worst of the drops. You give up some return buying back in a little late after each dip — that's the cost — but you dodge the −50% holes that take years to climb out of. For anyone using leverage, avoiding those holes isn't a nicety, it's survival: a −50% buy-and-hold drawdown is a full liquidation at just 2x.
Where it doesn't work
I'm not selling a miracle. The same test on Bitcoin, Japanese equities and silver showed trend-following added nothing — those either don't trend cleanly at the daily scale or whip you in and out for no gain. Oil (Sharpe 0.42) barely paid on its own but did its job of protecting a portfolio in downturns. The edge is real, but it's specific: broad, liquid equity indices, daily timeframe, long horizon.
I'm running it for real
I believe this one enough that I put live money on it — a leveraged QQQ position that exits to cash the moment QQQ closes below its 200-day line. It's only days old, so I'm not here to wave a profit screenshot; the point is the rule, not a lucky week. If you want to think in these terms, the most important habit is still sizing: a strategy that loses 4× less in a crash only helps if you didn't over-leverage into it in the first place. Size every position with the position size calculator and know your liquidation price before you click buy. Boring, slow, and still standing — that turns out to be the rarest combination in trading.