Last updated: 14 May 2026
You have decided to lay a horse. The selection work is done. You like the angle, the price feels wrong, the form does not stack up. Then the screen asks you a question you were not expecting.
Stake or liability?
It looks like a bookkeeping detail. It is not. The choice between fixed stake and fixed liability decides how much of your bank you put at risk on every single lay you ever place. Get it wrong and one bad afternoon takes a fortnight’s profit. Get it right and you can sit through losing runs that would otherwise end your laying career.
Should you lay to a fixed stake or a fixed liability?
For almost every backer, fixed liability is the safer choice. Fixed stake gives you the same small profit on every winning lay but exposes you to a much larger loss when the lay loses. Fixed liability caps the loss per bet at a known fraction of your bank, which is the property that survives variance. Across over 126,000 UK Betfair-priced favourites between July 2013 and March 2026, the favourite trades at an average BSP of 3.41 and wins 35.7% of the time. A £50 fixed-stake lay on that average horse risks roughly £120 to win £49. A £50 fixed-liability lay risks £50 by definition. Choose fixed liability unless you are running a tightly defined, low-odds-only system with a tested edge.
What Laying Actually Means (And Why Staking Choice Matters More Than Selection)
To lay a horse on the exchange is to take the role of the bookmaker on a single runner. You agree the price, you accept somebody else’s stake, and you pay out only if that horse wins. If the horse loses, finishes second, refuses to enter the stalls, or pulls up before the second flight, the stake stays with you.
That is the part most articles cover. The part most articles skip is what happens at the moment of placing the bet. The exchange asks you to enter a number. You can enter it as the amount you want to take from the backer taking the other side. That is fixed stake. Or you can enter it as the maximum you are prepared to lose if the horse wins. That is fixed liability. Same selection, same price, same race. Different shape of risk on every single bet.
This is why the choice matters more than people think. The selection edge in any lay angle is finite. Even a strong angle, like the ones we look at in the wider laying the favourite guide, only generates a few pence per pound across thousands of bets. Lose half your bank on a single bad outlier and the angle does not matter. The bank is gone before the edge has time to show up.
Fixed Stake Laying: Small Wins, Big Losses
Fixed stake means you set the same number every time as the amount you would take from a winning lay. If you lay at £20 fixed stake, you keep £20 (minus Betfair commission) every time the horse loses. The size of that £20 is constant. The size of what you risk is not.
Take a £20 lay on a horse priced at 6.0 on the exchange. If the horse loses, you collect £20. If the horse wins, you owe the backer who took your offer £20 multiplied by 5, the net odds, which is £100. Same lay at 11.0, you owe £200. Same lay at 21.0, you owe £400. The amount you can win is fixed at £20. The amount you can lose climbs with the price.
The Race Advisor dataset shows what this looks like over real UK racing. Across over 126,000 UK Betfair-priced favourites between mid 2013 and early 2026, a £50 fixed-stake lay on every favourite would have averaged a liability of £120 per bet, with a worst single bet of £1,423 (a Sedgefield runner who started favourite at BSP 29.46 in April 2023 and won). One bet of that shape wipes out roughly twenty-eight clean lay wins. The shape of fixed stake is small wins arriving frequently and rare losses arriving with size.
People accept this because it feels normal. Backing backers use fixed stake every day, and the maths is symmetrical when you back: you risk £20, you can win £80 at 5.0. They keep the habit when they cross over to laying. The maths is no longer symmetrical. Lay at £20 fixed stake at 5.0 and you risk £80 to win £20. The pattern is reversed. The habit no longer fits the bet.
Fixed Liability Laying: Capped Downside, Smaller Wins
Fixed liability flips the question. Instead of fixing the amount you take, you fix the amount you can lose. If you set liability at £50 and the horse wins, the bet costs £50. The actual stake the exchange shows you is calculated backwards from that figure: liability divided by net odds.
At 6.0 with £50 liability, the actual stake is £50 divided by 5, which is £10. So a winning lay returns £10 (minus commission). At 11.0, the stake is £5. At 21.0, the stake is £2.50. The amount you can win shrinks as the price climbs, but the amount you can lose stays exactly where you set it. You will never wake up to a Sunday morning -£1,400 for a single horse you barely thought about.
The same UK favourite sample makes the contrast obvious. A £50 fixed-liability lay on every favourite would have averaged an actual stake of £33 per bet, with a worst single loss of exactly £50 (the cap). The largest liability under fixed stake is roughly twenty-eight times the largest liability under fixed liability for the same nominal £50 bet. Bet by bet, the variance of profit and loss under fixed stake is roughly twice that of fixed liability. Run the same 13-year chronological sequence under both methods and the worst peak-to-trough drawdown is about 2.6 times worse under fixed stake.
The trade-off is real. Fixed liability gives you smaller winning lays at any price above evens. The horse you laid at 8.0 returns less than half what fixed stake would have returned. The horse you laid at 21.0 barely registers. That is the cost of certainty about your downside. Most backers who price the trade out properly take it.
What The Maths Looks Like At Real Betfair Prices
It is one thing to compare fixed stake and fixed liability in the abstract. It is another to see what they actually do when you point them at thirteen years of UK racing.
Race Advisor’s full UK dataset contains over 126,000 BSP-priced favourites between July 2013 and March 2026. The favourite wins 35.7% of the time at an A/E ratio of 1.04, meaning the market is, on average, slightly undervaluing the favourite (laying every favourite is therefore a losing strategy in either method). The point of running the comparison is not to recommend lay-the-favourite as a system. It is to show what happens to your bank under each staking method when you put it through a realistic distribution of prices and outcomes.
Bet £50 nominal on every UK favourite for the whole period:
- Fixed stake at £50: total profit of -£476,861, average loss per bet of £3.76, worst single bet -£1,423, worst peak-to-trough drawdown -£477,814, standard deviation of bet outcomes £80.
- Fixed liability at £50: total profit of -£184,347, average loss per bet of £1.45, worst single bet -£50 by definition, worst peak-to-trough drawdown -£185,179, standard deviation of bet outcomes £41.
Both methods lose money over the period (favourites are slightly value), so neither is a system. What the numbers reveal is the shape of the risk. Fixed liability halves the standard deviation, more than halves the drawdown, and caps the single-bet damage at the planned amount. Fixed stake delivers consistent £49 wins until it does not, at which point a single bad afternoon clears most of a year.
The gap widens with price. Bands of UK favourites between BSP 7.0 and the maximum lose roughly £27 per bet under fixed stake against roughly £3 per bet under fixed liability for the same nominal £50. At BSP 1.5, the two methods are within a pound or two of each other per bet. The decision to use fixed stake is, in effect, a bet that you will never accidentally lay the wrong long-priced horse. Most backers are not in a position to make that promise.
How Big A Bank Do You Actually Need?
Bankroll size is the first practical question and the one most often skipped. The honest answer is: enough that no single bet, and no normal losing run, can move your overall position by more than a couple of percent. The numbers translate cleanly into both methods.
Fixed liability. Set your maximum liability per bet to between 1% and 2% of your bank. A £1,000 bank gives you £10 to £20 of liability per bet. A £2,500 bank gives you £25 to £50. The constraint is the £2 minimum stake on Betfair. At £20 liability and a top lay price of 11.0, your actual stake is £2, which is the floor. If you want to lay at higher prices than that, you need either a bigger bank or a tighter cap on the price you will lay at.
Fixed stake. Set your stake to between 0.25% and 0.5% of your bank, because the relevant figure is liability and average liability sits at roughly 2.4 times stake on UK favourites. A £1,000 bank therefore supports a £2.50 to £5 fixed stake. A £5,000 bank supports a £12 to £25 fixed stake. To match the £50 stake used in the dataset comparison, you need a bank of around £10,000 to £20,000 if you want average liability inside 2%, or close to £70,000 if you want even the largest historical liability inside that limit.
This is why fixed liability is the default for most layers and fixed stake is the preserve of bigger banks running tightly priced angles.
As a rough floor under fixed liability, working back from the £2 Betfair minimum stake and a 2% liability cap, a bank of about £500 covers lays up to BSP 6.0. £1,000 stretches the same discipline to BSP 11.0. £2,000 is the realistic minimum to lay anywhere up to BSP 21.0 sensibly. Below those numbers you are either betting too small to be worth the screen time or laying too large a chunk of your bank on every horse.
Which Staking Method Should You Use?
Pick fixed liability if any of the following is true. Your bank is under £5,000. Your lay angle includes prices above 5.0. You have not yet measured the historical drawdown of your selections. You bet across multiple race types. You have ever talked yourself into a lay at a longer price than you intended.
Pick fixed stake only if all of the following are true. Your angle is restricted to a known low-price band, normally under BSP 4.0 and ideally under 3.0. You have a tested A/E ratio for the angle and the sample is large enough to trust. Your bank is large enough that the average liability under fixed stake is comfortably under 2% of bank. You actively want a flatter, more bookmaker-shaped P&L curve and accept the larger single-bet risk that comes with it.
For everyone else, fixed liability does the job. The smaller winning lays are the price you pay for being able to keep betting through the bad runs that any honest lay angle goes through. The discipline that protects the bank is the same discipline that allows the edge to compound. The same logic shows up in the wider Kelly Criterion staking piece: survival is the precondition for growth.
Once the staking method is in place, the next question is which favourites are worth laying in the first place. The disparity method for laying favourites covers the price-and-volume side of that. The Race Advisor racecards layer PR Rank, VDW Rank, Race Predictor probabilities, and pace ratings on top of the form so you can see exactly which favourites the data does not back. Today’s racecards show all of it on one screen.
