Closing Bell / gaps-get-filled / sp500_10y_strict

The folklore is directionally true, but the naive fade is not a strong trading strategy.

This submission tests the old trading saying with a strict range-gap definition on U.S. equities, then separates the empirical question from the trading question. The short version: prices often do revisit yesterday's range, but that does not automatically create a good next-day fade.

60d fill rate
73.1%
Eventual fill rate
90.0%
Trade Sharpe
-0.60
Total PnL
$-594,165

The Definition

A strict gap up means today's low is above yesterday's high. A strict gap down means today's high is below yesterday's low. The gap is considered filled once a later trading day trades back through the prior day's extreme.

The trade test is deliberately simple: fade the gap at the next open, target the fill level, stop out at a fixed percentage from the gap day's close, and abandon the trade after a fixed number of days.

{
  "universe": "sp500",
  "tickers": null,
  "start_date": "2016-04-19",
  "end_date": "2026-04-19",
  "gap": {
    "direction": "both",
    "min_gap_pct": 0.01,
    "fill_horizons": [
      1,
      5,
      20,
      60
    ]
  },
  "trade": {
    "position_size_usd": 10000.0,
    "stop_loss_pct": 0.03,
    "time_stop_days": 20,
    "commission_bps": 1.0,
    "slippage_bps": 2.0
  }
}

What The Data Says

The claim gets stronger as the holding horizon expands. That matters, because folklore usually speaks in eventual terms, while a tradable setup has to care about timing.

The Trade Test

This is the more useful question for a trader: if you actually fade the gap instead of just counting later fills, do the profits justify the path risk, capital lock-up, and tail events?

Why The Story Breaks

A high eventual fill rate is not the same thing as a high-quality trade. These three frictions explain most of the gap between those claims.

Small edge, large stop

The median strict gap is only 1.8% while the stop is fixed at 3.0%. That is only 0.61x reward-to-risk before slippage.

Fills are often too slow

Among gaps that do fill, 35.8% take longer than the 20-day time stop. The folklore can be directionally right and still arrive too late for this trade design.

Tail losses dominate

Stops account for 58.6% of exits and the worst realized trade loses 36.2%.

Worst Trades

The losers below are the trades that keep a superficially high win-rate story from turning into a durable edge.

Symbol Side Entry Exit Reason Return
BLDRlong2020-03-132020-03-16stop-36.22%
NCLHlong2020-03-132020-03-18stop-29.84%
TDGlong2020-03-132020-03-16stop-23.97%
ROPlong2020-03-132020-03-16stop-22.08%
OXYlong2020-03-102020-03-11stop-21.26%
RCLlong2020-03-132020-03-13stop-20.72%
MPClong2020-03-132020-03-16stop-20.65%
TRGPlong2020-03-102020-03-10stop-20.52%
Generated from committed run data. Open this file directly in a browser; no server is required.