Passive holders earn when their terminal forecast is right. Momentum traders earn when price path continues after information—successive higher bids, thinning asks, volume lifting offers without a full retrace. This chapter is for shorter horizons, more tickets, and stricter execution discipline.
Momentum is not a belief that prices drift forever. In prediction markets, trends end at releases, rule clarifications, and arbitrage that snaps mispriced siblings back in line.
Momentum is the first active strategy in the module: more fees, more screen time, more ways to confuse flow with edge. It rewards preparation on order books and news dynamics more than it rewards reflexes.
What trend looks like on a real book
On a central limit order book, momentum usually means the bid stack rebuilds higher and sellers stop leaning on the offer. The spread may widen after a spike—that often means uncertainty, not a free ride. On hybrid venues that pair a book with a pool, the pool price can jump before depth returns; momentum on the pool and fade on the book is a common two-step pattern.
Read volume with skepticism. A $200 print is not a trend; compare to typical depth at the touch. If YES plus NO sum drifts away from parity without bots correcting it, you may be watching illiquidity, not informed flow. The order-book chapters taught that headline mids lie; momentum traders feel that truth first.
Momentum versus passive and contrarian
If your fundamental probability is stable for months, passive hold probably beats scalping. If order flow after a verified headline is your edge, momentum is on the table. If two venues disagree by more than fees, that is execution arbitrage, not trend. If the crowd overshot your Bayesian update, contrarian fades belong in the next chapter—not every green candle is momentum.
| Cost | Passive | Momentum |
|---|---|---|
| Fees | Lower | Higher |
| Time at screen | Lower | Higher |
| Skill emphasis | Forecasting | Timing + execution |
| Tail risk | Settlement | Whipsaw + settlement |
When momentum is rational
Momentum pays when you also moved your probability on the same evidence, not when you are renting volatility. A poll that shifts your forecast from 48% to 55% can justify buying after a move from 41¢ to 49¢ if 49¢ is still below your updated fair value. Buying 66¢ when your post-news fair value is 63% is negative expected value even if the chart keeps climbing for an hour.
Use momentum after timestamped information you can defend: filings, official data, debate moments—not vibes from a clip you have not verified. A cooling period—no market order for thirty minutes after a headline—filters many fake trends.
A worked example that failed well
A gubernatorial YES contract jumps from 41¢ to 49¢ in eight minutes on a poll headline. The spread is 3¢; you market-buy at 49¢. Three hours later, after a rebuttal clip, the bid is 44¢. You lost 5¢ per share plus fees—not because momentum is fake, but because you chased without a fresh fair value and paid spread twice.
The lesson is procedural: wait two minutes for the spread to settle, recompute probability, prefer a limit in the upper forties, and define an invalidation level (for example, prior session low) before entry. Momentum without a post-news probability is gambling with better charts.
A worked example that respected edge
A court filing hits; YES rips from 55¢ to 62¢ in twenty minutes as latecomers pile in. Your read of the filing moves fair value from 52% to 61%. You buy 57–60¢ on limits during the climb and scale out 65–68¢ as flow exhausts. The cascade follower who buys 66¢ with 63% fair value donates the last nickel. You traded information arrival with a plan; they rented the chart.
Entry and exit habits
Size smaller than passive Kelly suggests—momentum clusters violate independence. Many disciplined traders risk half a percent to one and a half percent of bankroll per scalp and cap daily loss at a few percent before stopping for the day. Concurrent momentum legs in the same election cluster should stay at one or two, not five.
Exits often scale out into strength rather than hope for the last cent. Time stops matter when the thesis was “today’s flow” not “November outcome.” If sister venues flatten while yours still trends, you may be late to arbitrage, not early to genius. Hard price stops are blunt on gap news; invalidation levels plus size caps often beat stock-style percentage stops.
Multi-outcome and timeframes
Winner-take-all slates trend relative to siblings: one candidate spikes while others drift. Sometimes two candidates rise together—that is party-level shock, not idiosyncratic momentum on one name. Do not momentum-trade bundle legs without understanding combo liquidity.
Scalps from five to ninety minutes need deep books and tight spreads. Multi-day swings can work when spreads stay stable. Pure automated market maker long tails punish market exits—avoid size there entirely.
Failure modes
Chasing the top tick after a headline is the classic error. Overtrading twelve tickets to scratch boredom burns edge through fees. Narrative trading without a source timestamp looks like momentum in your journal but reads as noise in the data. Confusing a crypto-native spike with a regulated flatline is cross-venue lag—arbitrage territory, not trend continuation. Paying spread on every partial entry when one limit ladder would do is death by a thousand cents.
Daily loss limits and journaling
Many momentum traders stop new risk after three to five percent daily bankroll loss. The stop is on activity, not on one position’s price—momentum is a business of small edges repeated. Log catalyst time, entry VWAP, and exit reason so monthly review can show whether you bought information or rented charts.
Link to execution and sizing
Momentum is where market orders sometimes pay for themselves once per thesis; every additional market clip needs the same post-news probability check. Size at half to one percent of bankroll per scalp unless portfolio cluster already holds election risk—then cut again. The limit-versus-market and position-sizing chapters are not optional extras; they are how momentum survives fees.
When not to momentum trade
If your edge is months-long macro, passive hold beats renting intraday variance. If spread is wider than your expected continuation, pass. If manipulation prints are lifting the book without verifiable news, wait for flow to stop. If YES and NO sum already violates parity, bots or arb may snap price before your trend matures.
Volume filter in prose
If twenty-four-hour volume is usually fifty thousand dollars and today’s print is two hundred dollars, ignore the tick. If volume is five times baseline and your probability moved, trend hypotheses deserve limits—not market chase.
Key ideas
Momentum needs updated fair value after news, tight spread discipline, and small size. Trends end at information and arb, not at your target price. Execution quality determines whether path trading is a skill or a fee donation.
What comes next
When the crowd runs too far on the same headline, a different strategy applies: fade with evidence, not attitude.
Related markets
When a parent election contract moves, child state markets may lag. That lag can look like momentum on the child—it may be stale pricing. Check whether the move is sync across siblings before you chase one leg.
Misconceptions
“Trend equals truth” confuses path with forecast. “Volume up so buy” without a source is herding. “I scalped green” without fee math is illusion. Momentum is a skill tax on impatience—pay it only when post-news edge is real.
If you cannot state the catalyst timestamp in one line, you are not momentum trading.
Next: Contrarian Strategy: Betting Against the Crowd