Selling Gordon and buying Jordan

Brett Mends reveals who the real winners were in last week's budget

While half the City was at Cheltenham last week losing its shirt on Rooster Booster, colleagues left behind at their desks were doing rather better on the budget.

The Square Mile runs a lively market on this annual ritual, but it has little to do with the measures announced. At the specialist 'spread betting' bookmakers, traders take positions on how long the speech will be, the number of times the Chancellor will mention certain words, and even on the number of sips of water or interruptions. Traders have to watch the speech anyway, but these days many of them care less about what the Chancellor says than the way he says it.

This year was an upset for the bookies. In a sly speech designed to take the wind out of the Tories' electoral sails, Gordon Brown mentioned 'tax' just 35 times. The spread betting 'price', or prediction, based on Brown's previous budget speeches, was for 75 mentions. Plenty of traders thought the price too high and 'sold', betting that the danger-word would occur less often this year. They cleaned up.

Spread betting is a particularly British enterprise, combining the financial ingenuity of the City with the sporting eccentricity of P.G. Woclehouse's story The Great Sermon Handicap, in which young bloods ran a book on which local vicar's sermon would last the longest. It gives gamblers access to markets in almost everything. Those not content with dealing in Marks & Spencer shares or cocoa futures can trade in Test match wickets, election results or 'reality television' contestants. The City had a bad I'm a Celebrity . .. Get Me Out of Here!, by the way: traders backed ex-punk John Lydon and top-heavy Jordan to last longest; Kerry McFadden's win was a disaster.

These non-financial markets operate much as the real thing does. People trade in and out of positions, capitalise on swings of opinion, and 'arbitrage' between prices. And, just as in the more traditional financial markets, if they take the wrong view they can lose their shirts, buttons and all.

The technique was thought up 30 years ago by a young out-of-work stockbroker called Stuart Wheeler. The business he founded, IG Index, became hugely profitable and was taken over last year for £143 million, enabling Wheeler to become the Conservative party's biggest individual donor and play a part in the downfall of lain Duncan Smith, as well as buy himself a castle in Kent.

Wheeler's great innovation, in 1975, was to open financial markets to people outside the City. He began by letting punters bet on the gold boom without owning any gold. His book quickly expanded to include other commodities such as copper and sugar, and then to shares. And in the 1980s a rival company, City Index, run by Jonathan Sparke, began to apply the same principle to sports — to be followed by IG and a third competitor, Sporting Index.

The idea was simple enough. The cocoa futures market, for example, tries to predict the price of cocoa on a given day: traders buy and sell, depending on whether they think the prediction is too low or too high. Sparke asked why the mechanism couldn't be applied to the number of goals in a football match or the length of the budget speech. The bookies make a prediction, then traders pile in, 'buying' or placing an 'up' bet if they think the forecast too low, or placing a 'clown' bet if they think it too high. How much they wager per point, goal or minute depends on how rich, confident or foolhardy they feel. It caught on like wildfire in the City, where there is no shortage of people who are rich, confident and foolhardy.

The first spreads market on a general election, in 1992, let people bet on the number of seats each party would win. While the opinion polls were proved famously wrong, IG Index was the first to predict the surprise Tory victory — and did so several days before polling began. The general election spreads market remains a more accurate and informative barometer of the parties' strength than the opinion polls, although it demonstrates a slight City bias towards the Conservatives. If the market is right this time. Labour is on track for a majority of 30 next time.

Spread betting companies continue to run markets in shares, currencies and commodities. The stock-market slump was quite good for them: for private investors, spread betting is the only realis

tic way to bet on shares falling. Likewise, 1G does a roaring trade in its market on house prices, which tracks the Halifax index. Many home-owners are trying to insure against a slump by betting on it. Yet this is an expensive undertaking — bets have to be rolled over every six months, and you have to cover the bookmakers' commission. And when the bet goes against you, you have to stump up real money.

On the sporting front, spread betting firms did not always know what they were doing in the early days. They were taken to the cleaners when Brian Lara scored 375 in the Antigua Test in 1994. Many shrewd gamblers had 'bought' his runs, so each additional run beyond a modest score made them money. Compton Hellyer, the founder of Sporting Index, recalls, 'It was horrendous. He was in for three days. Every time he hit a four it cost us £1,500. We were sitting there bleeding to death.'

And just like any other market, spreads are open to insider trading. Managers at one firm recall an incident at the start of another Test, A client placed a large 'down' bet on the time of the first 'no ball'. The sooner in the match it came, the bigger his profit would be. England's opening bowler then delivered clear no balls with his first two deliveries. The first went unnoticed by the umpire; for the second, the bowler's front foot landed half a yard beyond the crease. He got the call, and the gambler a very handsome profit: the gambler was the bowler's brother-in-law.

The Financial Services Authority, which apparently lacked the resources to keep a close eye on Equitable Life's reserves, has even been known to intervene to close down a market in early throw-ins in football matches. Players with a financial interest kept hoofing the ball into the stands with their first kick.

Today, beating the bookies is harder, whether by fair means or foul. They have teams of specialists analysing every sport. But errors still occur: last year, before the arrival of the Russian billionaire Roman Abramovich and his bankroll of stars, Chelsea won 67 Premiership points. Yet when the market for the new season opened in August, the bookmakers had only moved the prediction up to 70. Chelsea's predictably impressive performance has since moved the price to 80, and the bookmakers are sitting on heavy losses.

Your correspondent foolishly closed out last autumn, far too early, and made a fraction of what he should have done. To make matters worse, he then placed a big 'down' bet on season points for Arsenal — who raced to the top of the table. So 'the house' won anyway.

Brett Arends is the author of Spread Betting: A Football Fan's Guide (Time Warner, 2002, £5.99).