The Overreaction-Catch
A live, market-neutral limit-order strategy for Kalshi soccer "Advance" markets (World Cup 2026). Real money. No directional bias — profit from panic mispricing, not from picking a winner.
+$5.70NOR-ENG 11 Jul '26, 2 round-trips, net of fees
$0maker fee on entries
0directional bets required
1. The idea in one sentence
When a goal (or a big chance / VAR check) hits a live knockout, the binary Advance market gaps far past fair value for a few seconds as resting liquidity is pulled — then snaps back. Sit a deep passive maker bid in the zone the panic overshoots to, get filled on the spike, and sell into the reversion.
2. Why the edge exists
- Liquidity vacuum. On a goal, makers yank quotes instantly. The book briefly thins to a few hundred contracts and price gaps way beyond the new fair.
- Reversion. Within seconds–minutes liquidity refills and price returns toward fair. The overshoot is transient.
- VAR windfall. Modern goals are reviewed. A disallowed goal snaps the price all the way back — your panic-price fill becomes a large winner.
- You are the patient maker. Passive, $0-fee, free-to-cancel. You are paid to provide liquidity exactly when everyone else refuses to.
3. Mechanics
| Step | Action |
|---|
| Pick market | Prefer the Advance leg (survives ET/pens) over the regulation Winner leg. Needs a deep, fast book (WC knockouts qualify). |
| Set the catch price | Rest a post_only bid at the observed overshoot depth, not just below market. If fair ≈ .60, the panic can print ~.30 — bid there, not at .48. Depth of the catch is the whole edge. |
| Ladder | 2 rungs (e.g. 15 @ .40 + 15 @ .31) to catch both shallow and deep gaps. Deep gap fills both; shallow fills the top rung only. |
| Two-sided net | When the game is a coin-flip (e.g. 1-1 in ET), arm a mirror catch on each team's Advance leg (~.30 each). Whoever scores, the opponent's price craters into your bid. |
| Exit | On fill, sell into the recovery with an IOC priced at the best bid (often price-improves). Bank it; don't hold the binary variance once price ≈ fair. |
| Housekeeping | Once a decisive goal stands, cancel the now-stale opposite rung (free). At FT/settlement, flatten and disable the watch. |
4. Gap-safety rule (do not skip)
A deep resting bid mostly fills on adverse news. Only price it where you're happy to own it even if the goal is real: bid ≤ the post-real-goal fair value. Then downside ≈ breakeven and upside (reversion / VAR reversal) is convex. Example: England-advance post-real-Norway-goal fair ≈ .50–.55, so a bid at .31–.40 is value on a real goal and a windfall on a reversal. Never set the catch so shallow that a real goal makes it an instant loss (a "falling knife").
5. Risk management
- Small size, set at entry. ~quarter-Kelly; the edge is frequency + convexity, not size. Never average down.
- One position per game intent; the two-sided net is one contingent position (only one side realistically fills per goal).
- Maker-first / $0 fee on entry; taker only to exit fast. Kalshi taker ≈ 1–2c/contract.
- Cap downside: total resting exposure a few % of bankroll. A whole net was ~$9 on a ~$330 book.
- No directional thesis needed — but if you do form one (e.g. Advance leg trading above your ET-coinflip fair), take the free exit at/above fair rather than holding.
6. Worked example — Norway 1-1 England, QF, 11 Jul 2026
| Leg | Buy | Trigger | Sell | Net |
|---|
| ENG-advance ×30 | .48 (maker, $0 fee) | 55' Norway "goal" panic gap → later VAR-disallowed | .62 (IOC, price-improved) | +$3.71 |
| ENG-advance ×15 | .40 (maker, $0 fee) | transient ~86' dip (score stayed 1-1) | .55 (IOC) | +$1.99 |
Lesson banked live: the first catch (.48) was too shallow — the true overshoot ran to ~.30. Re-armed deeper (.31/.30) and two-sided for extra time.
7. Execution notes (Kalshi V2)
OPEN (maker rest): POST /portfolio/events/orders
{ ticker, side:"bid", count:"15.00", price:"0.3100",
type:"limit", time_in_force:"good_till_canceled",
post_only:true, self_trade_prevention_type:"taker_at_cross",
client_order_id:"..." } # $0 fee, free to cancel
CLOSE (take into recovery): same endpoint
{ side:"ask", count, price:, reduce_only:true,
time_in_force:"immediate_or_cancel", ... }
CANCEL (free): DELETE /portfolio/events/orders/<order_id>
8. When NOT to use it
- Thin / slow markets — no vacuum, no reversion, and no depth to exit.
- A goal that is clearly decisive and final late with little time left — reversion is weak; keep bids deep and size tiny.
- When your catch price would sit above post-real-goal fair — that's a knife, not a catch.