Article

Range Day or Trend Day: What the OI Gamma Regime Did on 170 Sessions

We replayed every touch of every level on ES and NQ, faded and broken out, in positive and negative gamma. Buying dips paid in positive gamma; selling rallies paid in negative gamma; breakouts paid nobody.

By GammaTape7 min read

The sign of the net open-interest gamma is the one options signal with papers behind it: when dealers are net short gamma, their hedges chase the move, and on our archive 30-minute realized volatility ran 1.4 times higher below the flip than above it. That much was known. What we wanted to know was simpler and more useful: at a level, do you fade or do you go with the break, do you buy or do you sell, and does the answer change with the regime?

How it was measured

  • 88 SPX sessions replayed on ES and 82 NDX sessions on NQ, one-minute bars, the level mapped with the previous minute’s basis, costs in, stop checked before target inside a bar.
  • Every contact of every level — the two majors, the walls, both zero gammas, prior day high and low, the expected move, the opening range — played twice: as a fade (limit at the level, against the move) and as a breakout (the first one-minute close through the level, with the move).
  • One stop, 0.10 % of spot (about 8 ES points, 28 NQ), a target of 1.5 times the stop, and a control that held to 15:55 with the stop only. No grid, no parameter search, a placebo at the same distance for every line, 30 % of dates held out.

The regime decided the side

In positive gamma, buying the dips at a level paid on both indices and selling the rallies lost: on NQ the long fade ran a profit factor of 1.34 against 1.05 for the short, and held to the close 1.28 against 0.79. In negative gamma the picture reversed — the long fade fell to 0.61 held to the close, while the short paid 1.33 at 1.5 R. The placebo short paid too, which is the honest reading: on a trend day the side mattered more than the exact strike.

Fade, not breakout

Breakouts were rare — one combination in the whole study reached a hundred events — and they did not beat a short taken anywhere else. They produced about 1.3 trades a day against 2.7 to 3.4 for fades. A one-minute close through a level tells you that level is not to be faded again today; it did not, on this archive, make a trade of its own.

The first read

The single best cell was the volume zero gamma reached from above in positive gamma — the dip back to it, bought: 1.68 against a placebo of 0.90, 56 % of trades reaching 1.5 R, about 1.2 of them a day. The majors and the walls followed the regime rule but did not have a hundred events each on their own, so they are the second read, not the first.

Read the regime on the last 30 sessions.

Open the desk →

Keep reading