Take-profits and stop-losses: every exit rule lost to holding
Lock in the double? Cut the loser? We tested both against our own real positions. Holding to resolution beat every mechanical exit we tried — including in the case built to favor it.
The hypothesisBinary markets should still reward risk management. Rule A: when a position doubles, sell — a bird in hand. Rule B: when a position drops through a floor, cut it before it goes to zero.
The test
Both rules were tested against our own live book — real positions, live bids. The take-profit probe watched every open position that could still double and, when the executable bid crossed 2× entry, logged the locked-in exit price, then compared it to what actually happened at resolution. The stop-loss test replayed 1,725 settled real positions against every cut threshold from shallow to deep, at the bids that were actually available on the way down.
The result
Take-profit lost to holding 33 times out of 33 in the armed band (entries ≥40¢). The mechanism is unforgiving: a position that doubles from 40¢+ now sits above 80¢, and things priced above 80¢ usually win — selling there systematically caps winners at the exact moment they've become favorites. Stop-losses fared no better at any threshold: in binary markets a price collapse is the information, the bid you cut into has already absorbed it, and enough crashed positions recover that cutting locks in the worst tick. Holding won the aggregate at every setting.
Conclusion
Intuition imported from equities fails here. A binary contract's price is the probability; mechanical exits just sell winners cheap and losers cheaper. The one exit hypothesis still open is informational, not price-based: exiting when the whale we copied exits — early evidence mildly favors it, sample still small. Until something passes, the rule is: pick entries carefully, then let resolution do the exiting.