GEX Around CPI and FOMC: Plan for a Broken Regime

Distinguish direction from volatility

Positive modeled gamma may suggest a dampening feedback under simplified hedging assumptions. It does not specify the correct direction of a surprise. Negative modeled gamma may suggest amplification, but it does not say whether the initial move should be up or down. The sign of a portfolio sensitivity cannot substitute for understanding the information shock.

Research on options hedging and market impact concerns mechanisms and estimated aggregate effects. It does not promise that one event will respect a public indicator. The Cboe discussion of gamma hedging is useful for understanding feedback; the scenario framework here is our own practical inference from that mechanism.

Also separate realized movement from implied volatility. An announcement can produce a large initial price move while option implied volatility falls as uncertainty resolves. That combination affects option values and hedge sensitivities in more than one way. Vanna and charm help explain why a gamma-only story can be incomplete.

Liquidity changes the cost of being wrong

Before a release, resting orders can thin and spreads can widen. After it, fast updates can make a visible snapshot old almost immediately. A stop level is a risk intention, not a guarantee of a specific fill. If a strategy only looks attractive assuming zero slippage during the fastest minute, its apparent edge may disappear in practice.

Build a separate event cost assumption into review. For an invented example, a normal-session study might assume one tick of adverse slippage, while an event stress scenario uses several ticks or a skipped fill. Those numbers should come from observed records for the instrument and order method, not from a generic promise about liquidity.

Correlated exposures can also increase together. Long index futures and bullish short-dated options may represent different tickets but similar directional risk. Aggregate the scenario loss across instruments before judging whether any single position is small.

Refresh the map after the market changes

Once fresh prices and volatility inputs arrive, compare the new exposure view with the saved pre-event view. Explain whether differences came from repricing, time, new transactions, or a changed data universe. Avoid editing the original prediction to match the new map.

Then examine the new auction: where did meaningful volume accumulate, which reference was rejected, and which area gained acceptance? If the post-event distribution is far from the morning range, an updated volume profile may be more relevant than the old center of balance.

Review event sessions separately from ordinary sessions. Mixing them can conceal a strategy that works in calm conditions and loses disproportionately when information arrives. Keep the no-trade decisions, data interruptions, and invalidated setups in the record alongside attractive examples.

Use OrderMap as a read-only analytics workspace for the views and datasets available to your account. Combine the model with explicit position-sizing rules and a calendar you verify from the official release source before the session.

Review the first move and the second move separately

In an invented announcement sequence, price jumps above the morning range, reverses through it, and later builds a new distribution below. Calling the whole session a failed breakout obscures three different states. The initial repricing may occur before reliable fills are available. The reversal may invalidate a continuation idea. The later distribution may support an entirely new research question.

Record when each state became observable and which inputs were fresh then. A later exposure update should not be used to justify an earlier entry. Likewise, the first price reaction should not permanently determine the day's narrative when later activity contradicts it.

For a practical review, mark the pre-event state, immediate reaction, first stable range, and eventual accepted region. Compare those transitions across a fixed group of releases. That chronology is more informative than asking whether one morning gamma line happened to be near the day's high.

Explore OrderMap research tools