Do festival favourites make money?

There is a particular confidence that arrives with festival week. Pascal, the Lab's eternally hopeful punter, spends fifty weeks of the year taking chances on wet Wednesdays, and then Cheltenham or Royal Ascot comes around and he straightens his tie. This, he says, is when betting stops being a lottery. The big meetings are where the best horses in training actually turn up, prepared to the minute for a day circled in the calendar since autumn. The form is the deepest of the year, argued over in public for months. Nobody is hiding anything at a festival: the trials have been run, the gallop reports filed, the ante-post market has chewed every possibility since Christmas. So the festival favourite, the horse still standing at the top of the market after all that scrutiny, must be the most informed bet in racing. On the biggest stage, back the best horse.

It is a lovely theory, and unlike most punting folklore it starts from true facts. Festival favourites are, on average, better horses than everyday favourites. The form is more exposed. The preparation is more serious. The question is whether any of that survives contact with the price.

So we tested it the blunt way. We defined the big days by money rather than by name: every British race in our database with prize money of forty thousand pounds and up, which captures the Cheltenham and Aintree features, Royal Ascot, and the big Saturday heats besides, 2,095 races in all. In every one we backed the favourite, one flat stake, settled at Starting Price and counted honestly. One warning before you read on: 2,095 races is our smallest headline sample, and we will be straight with you about exactly how much it can and cannot prove.

No corner of betting has better marketing than the festival banker. Every preview night anoints one, every office sweep leans on one, and the word itself does half the work: a banker is not a bet, it is a deposit. Memories supply the other half. Everyone who loves jumping can name a banker that rolled home up the hill with the ears pricked, and those memories get replayed every March. The bankers that fell at the second last survive only as hard-luck stories, and hard-luck stories are never filed under evidence.

Then there is the feeling of informational safety. For most of the year a punter half-suspects the connections know something he does not, that the real race was decided on a private gallop last Tuesday. At a festival that fear evaporates. Everything is public: the trials, the going debates, the jockey bookings, the trainer's every cough. Backing a festival favourite feels like betting with the lights on, and the lights are the whole appeal.

The favourite obliges often enough to keep the story alive, too. It wins 32% of these big-money races, close to one in three, and festival wins are the most vivid wins in the sport. A banker landing at the festival is an anniversary you celebrate; a banker landing at a Tuesday all-weather card is a coupon.

Pascal adds his own flourish: at festivals, he reckons, the once-a-year money floods in on names, colours and last year's winner, so the sharp punter riding the genuine form horse is finally on the right side of the crowd. It sounds shrewd. It has the market exactly backwards, and the next section is about why.

The problem with betting with the lights on is that the lights are on for everyone. All that public information, the trials, the gallop reports, the months of argument, does not sit waiting for you to use it. It is already inside the price, put there by the deepest and most professional betting markets of the racing year. A festival market is the opposite of a soft target. The liquidity is enormous, the shrewdest money in the game is at its most active, and the odds have been rehearsed since the ante-post lists opened. Thousands of opinions have already traded against each other by the time you step in, and the price you take is the residue of that fight. When everything is known, the price knows it first, and a price that knows everything, minus the bookmaker's margin, loses by exactly that margin.

The margin itself is the second engine, and festivals do not shrink it. Add up the chances implied by every price in a race and they come to more than 100%, about 12% per race extra on a typical field and climbing towards 30% in big fields. Big fields are a festival speciality: the twenty-runner handicaps that anchor every Cheltenham card are precisely where the overround runs fattest. So the festival punter is often paying the market's biggest toll of the season for the privilege of betting into its most accurate prices.

Notice what this argument does not say. It does not say festival favourites are bad horses, they are usually outstanding ones. And it is not Pascal's mug-money story in reverse. The efficient-market problem is worse than any mug-money problem, because there is no side of the crowd left to stand on. Every horse is priced almost exactly right, and almost exactly right minus the margin loses at the same patient rate on Gold Cup day as on a quiet Tuesday. The occasion changes everything except the arithmetic.

We defined the festival by money, not by name. A calendar list of festivals invites endless bickering about what counts, so we drew the clean, checkable line the theory actually means: every race with prize money of forty thousand pounds and up, the days when the best horses turn up because the prize is worth turning up for. That captures the spring festivals, Royal Ascot, and the big Saturday features year-round. The filter leaves 2,095 races out of the 26,839 real British races we hold. In every one we backed the favourite: one flat stake, official industry Starting Price, non-runners voided, joint-favourites split, a recorded winner required, and over jumps the fallers and pulled-up horses counted as the losing bets they are. The same honest settlement as every experiment on this board.

Now the caveat, placed in the middle of the method where it cannot be missed: 2,095 races is a small sample by our standards. It cannot be otherwise, because big-money races are rare by definition, a few hundred a season against thousands of ordinary ones. Small samples do not lie, but they speak with a wide margin of error. The 95% range on this result stretches from clearly painful to nearly break-even, and honest reporting means showing you that range rather than quoting the headline figure as if it were carved in stone. Our reference experiments elsewhere on the board rest on tens of thousands of races and pin their figures tightly. This one wears its uncertainty on its sleeve, and we would rather you saw the sleeve.

What the sample can do is answer the yes-or-no question it was asked. If festival favourites carried the profit the theory promises, even a couple of thousand races would be expected to show it, because a real edge does not spend that long hiding below zero. It does not show it. The verdict is no. The small print is about how emphatic a no it is.

Here is the number, and this time the range matters as much as the figure. Backing the favourite in every big-money race returns -8.37% across 2,095 races, meaning about £91.63 of every £100 staked comes back over the long run. The favourite wins 32% of these races, right in line with favourites everywhere. The 95% range is [-14.5,-2.2], and you should read it before the headline: it is wide, because the sample is small, and it matters that both ends of it sit below zero. The kind end creeps towards break-even without getting there. The painful end is worse than the favourite's everyday figure.

For context, the favourite across all 26,839 races returns -8.6%. The festival figure sits in the same neighbourhood, and with a range this wide we will not claim it is better or worse, only that it is about the same. That, quietly, is the finding. The most scrutinised, best-informed, most confidently backed favourites of the racing year lose money at roughly the rate favourites always lose money. All the extra information, the trials and the reports and the months of argument, bought the punter precisely nothing, because the market ate it first and charged the usual fee on the way out.

What we refuse to do with this sample is over-read it. A couple of thousand races leaves room for luck in either direction, and you will meet people online who insist festival favourites are secretly profitable, and others who insist they are a bloodbath. Both stories can be squinted into a range like [-14.5,-2.2], which is exactly why we publish the range instead of a story. This is also the page that will move most as seasons pass. Every festival adds races, the range tightens, and these figures live-pull from the current data, so what you are reading is the present state of the answer, not a snapshot of an old one.

So, do festival favourites make money? We can find no sign of it, and we have looked the honest way. Across 2,095 big-money races the favourite returns -8.4%, about the same leak favourites show all year round, with a 95% range that never pokes above zero. We hold this verdict more lightly than the board's big-sample results and we say so plainly: the sample is small because big days are rare, the range is [-14.5,-2.2], and the answer will sharpen season by season as this page updates itself. But a profit that exists tends to show itself, and this one has had a couple of thousand chances.

The deeper lesson is the efficient-market one, and it is worth carrying to every big meeting you ever attend. The festival favourite is not a weak bet because festivals are chaotic. It is a weak bet because festivals are the most orderly markets of the year. Maximum information, maximum liquidity, maximum professional attention: everything the banker theory counts as safety is exactly what squeezes the value out of the price. The bigger the day, the sharper the market, and the margin never takes the day off.

None of this is a reason to sit the festivals out. They are the best days the sport has, and a bet can be worth its cost as entertainment when it is sized as entertainment. If you take one practical rule from this page, take the board's oldest one: strike rate is not profit, and a banker is a story about a favourite. Past performance is not future returns, our SP settlement flatters the figure, and no staking plan turns a negative edge positive. Bet the big days with a flat stake you can afford to lose, or a fraction of the Kelly criterion if you genuinely hold an edge. On this evidence, the festival favourite is not that edge. It is simply the most enjoyable way there is of paying the margin.