Research note

Darvas Box, backtested: 12,456 trades say no

The famous box-breakout system managed 60d excess −1.34% (t −12.01) on S&P 500 stocks, 2010–2026. Verdict: no.

1 min read

Verdict: NO — the Darvas Box shows no edge on S&P 500 stocks, 2010–2026. The result is decisively negative.

The numbers

  • n = 12,456 trades
  • 60-day forward excess return vs SPY: −1.34%
  • t-statistic: −12.01 (significantly negative)
  • 72% of trades stopped out by day 60

What was tested

Box-breakout entries on S&P 500 constituents, 2010–2026, measured against SPY buy-and-hold. Two variations were run on top of the baseline.

  • A volume confirmation filter did not rescue it.
  • Restricting to bull-market regimes also failed.

One claim down. The audit continues weekly.

Not financial advice

Built for education. Backtests are historical; they do not predict future returns and are not a recommendation to buy or sell any security.

connecting…