The Series
Algorithmic Trading Strategies for Prediction Markets
The ten strategies quant desks run on equities, rewritten for event contracts. Seven translate cleanly, one works only in a narrow corner, and two have to be rebuilt from scratch. Each page says which, and names the tool that runs it.
Quick Answer
most equities algorithms do transfer to prediction markets, but not all of them, and the ones that fail do so for the same reason: a binary event contract settles at 0 or 100, so it has a bounded payoff and no unlimited upside. Strategies that trade a spread, a schedule or a base rate survive the move. Strategies that need a trend to run indefinitely do not.
Transfer cleanly: pairs trading, smart order routing, market making, TWAP, seasonality, volatility, machine learning.
Works only in a narrow corner: scalping, and only on the 15-minute series where the spread is payable.
Must be rebuilt: momentum and VWAP — both assume a price that can keep running, which a bounded contract cannot.
Start here · The pillar
The 10 Algorithmic Trading Strategies, Translated
The overview: which strategies translate, which need rebuilding, which do not transfer at all.
The ten, in reading order
- 01Pairs tradingTrade the spread between two correlated names instead of either one outright.CLEANLYPublishing Sep 9
- 02ScalpingMany small edges harvested inside the day, paying the spread every time.WITH LIMITSPublishing Sep 12
- 03Smart order routingSend each slice of an order to whichever venue is quoting best right now.CLEANLYPublishing Sep 15
- 04Market makingQuote both sides, earn the spread, manage the inventory you end up holding.CLEANLYPublishing Sep 18
- 05MomentumBuy what is already moving — the version that breaks hardest on a bounded contract.REBUILD ITPublishing Sep 21
- 06TWAP executionSlice a large order evenly through time so you do not move the price against yourself.CLEANLYPublishing Sep 24
- 07VWAP and catalyst volumeBenchmark to volume-weighted price — which assumes volume arrives smoothly, and here it does not.REBUILD ITPublishing Sep 27
- 08Seasonality and base ratesTrade the calendar: what usually happens at this point in the cycle.CLEANLYPublishing Sep 30
- 09Volatility tradingTrade the size of the move rather than its direction.CLEANLYPublishing Oct 3
- 10Machine learningFit a model to history and trade where it disagrees with the market.CLEANLYPublishing Oct 6
One strategy publishes every three days through October 6. Pages appear here as they go live.
Series FAQ
What is this series?
Ten algorithmic trading strategies that quant desks run on equities, each rewritten for prediction markets, plus a pillar article that ranks them. Every strategy gets its own page explaining how it changes when the instrument is a binary event contract that settles at 0 or 100 rather than a stock with unbounded upside, and which tool on this site actually runs it.
Which strategy should I start with?
Start with the pillar, then read pairs trading. Pairs trading transfers to event contracts more cleanly than anything else on the list, because trading the spread between two correlated contracts sidesteps the bounded-payoff problem that breaks momentum and VWAP. Smart order routing is the other high-value starting point if you already trade on more than one venue.
Does this apply to Polymarket as well as Kalshi?
Mostly yes. The strategy logic is venue-independent, and smart order routing specifically depends on quoting differences between venues. The fee maths and the depth of the order book differ, so position sizing and the minimum edge worth trading change between the two. Each article says where a strategy depends on venue mechanics rather than on the strategy itself.
Which of the tools mentioned are free?
Most of them. The EV calculator, probability converter, Kelly calculator, Bayes updater, base rate scanner, KL-divergence tool, Kalshi fee calculator and the Combo Edge Builder are all free. The live edge scanners and Thee Oracle dashboard are Pro. Each article names the tool that runs its strategy and marks whether it is free.