Editorial: Bet Unfair - Flaws in the Betting Exchange Business

11 September 2002

There has been a lot of debate in the United Kingdom in recent weeks over betting exchanges. All manners of (reasonable and unreasonable) arguments have been given a good airing, if not in articles and interviews then in letters to editor of the Racing Post.

The humorous thing at the moment is that U,K. bookmakers, especially the "big three," are "wagering" (sic) war on the betting exchanges, but they are not game to come out and attack the exchanges on the obvious abuses that they are avoiding, or minimizing--the BHB/Betting Levy Fee (and the GPT)--because (1) they fought it for so long and (2) they still want it reduced.

No one has been brave enough to come out and say what the real issue is: The punters are getting too good a deal.

But no one has been brave enough to come out and say what the real issue is: The punters are getting too good a deal...

Oh no! I've said it out loud.

In Web Storm Rising (August 2002), Bear Stearns' Jason Ader analyzes the effect of downward pressure on prices being exerted by online travel booking sites, firstly on airfares and latterly on hotel rates. Christiansen Capital Advisors' Eugene Martin Christiansen more recently reviewed this securities research in an article "Online Booking Exerts Downward Pressure on Travel Costs" (www.cca-i.com/np/EC/9-4-02EC.htm ).

What has travel to do with betting exchanges?

Well, more than you think.

The conclusion reached that "the explosive growth in online travel spending validates one of the Internet's early promises: lower consumer prices through more perfect information online" translates directly to wagering--with atomic bomb-type detonation.

This pressure is on the price of physical goods--something that has a real cost so that there is a floor to the price that can be offered. But with wagering there is no physical good. So, perfect information can create a perfect market (100 percent, in bookmaker board terms, "zero overround"). However, because the market continues to operate (each bookmaker still standing his own book), you are more likely to end up with market failure (100 percent, in bookmaker board terms, "negative overround").

The Big Problem

The big problem in the United Kingdom comes with betting exchanges, which by definition operate a "perfect market." That is, two punters are matched without risk to the "wagering operator," one punter backing a "runner" at 3/1 and another punter backing "not-runner" at 1/3.

The United Kingdom's legislation to introduce the GPT had specific clauses to assess the gross wagering margin of a "bet broker" (betting exchange) on which the 15 percent GPT would be paid. What the rocket-scientist drafter came up with was that the "bet broker," at the time of the matched/brokered bet, stands to the position of layer for one punter, viz. be a "bookmaker." But this has been easily circumvented by exchanges standing in the position of a layer (bookmaker) for both transactions and writing two bets 1/3 and 3/1 for zero gross wagering margin. Too easy?

Don't believe me?

Ian Davies, chief executive of BackAndLay (Racing Post, August 2002) chastised Betfair (and other exchange operators) for trying to convince the government it was a (real) "bookmaker" for this exact reason. So, we know that BetAndLay operates in this manner.

Even Betfair's Mark Davies admitted that Betfair "takes the place of their individual users for regulatory and tax purposes." Its also why Betfair's "turnover" has to be halved to compare it with bookmakers' "turnover."

The whole problem is then exacerbated by the fact that the exchanges then effectively shield their users from both the GPT (15 percent) and the "betting levy," which is now 10 percent of the gross wagering margin license fee. That's an awfully big competitive edge to give someone with product that is otherwise low cost.

Betfair's Davies said in an IGN article (Aug. 6, 2002), "We hope we will be regulated, but we are making it clear to government why it would be a mistake to regulate our individual users, whose place we take for regulatory and tax purposes. The precedent is the stock exchange: The NYSE is regulated, its brokers are regulated, but you and I trading shares do not need to be regulated, because the exchange ticks all the boxes."

But that is the nub of the problem. Stock brokers are facilitating the exchanging of a physical asset; they are paid a fee for providing that service. But in wagering (on racing) there is no asset; the whole of the industry is dependent on the profits made by the wagering operators (because that is the punter's losses, or "expenditure"). Now the users are being shielded by the betting exchange, which makes no wagering profit.

To take the analogy further, companies creating stock (via an IPO, capital raising etc.) are a class of brokers that are strictly regulated. So, why shouldn't the true "layers" on a betting exchange be regulated?

In that context, Warwick Bartlett's (chairman of the BBOA and CBA) demands that the "layers" on the exchanges must be licensed bookmakers (as required under current U.K. laws) is not unreasonable. More important though is an argument that online exchange "layers" have the same obligations as bookmakers imposed on them , that is, offer a market for every runner, have betting limits that they must accept. For example, rails bookmakers in Australia must accept bets, at the price on their board, to lose $5,000, or other rings $1,500.

Therefore, betting exchange wagering is causing a huge distortion in the U.K. wagering market (on racing). What's more, it is being fuelled by bookmakers (both on-course and off-course) using it to both lay off and soak up additional wagers. This opens up what should otherwise be a wholesale market to retail customers.

It is widely accepted that a reasonable amount of "turnover" on the exchanges is by virtue of the fact that they are operating as an efficient clearing house for bookmakers. In fact, it is so efficient that the Racing Post's Graham Green reported on Sept. 6, "Major bookmakers are demanding that on-course layers be banned from using online betting exchanges while on duty at the races over dangers it could artificially distort the starting prices which determine payouts on millions of bets each year."

It's Not Only Betting Exchanges

Not all the blame can be laid on the betting exchanges for creating an environment for market failure. Other "virtual betting ring" Web sites, such as bestbetting.com, allow the savvy punter to get "top odds." But even bestbetting.com excludes Betfair (despite displaying its prices) when calculating "best price percentage."

The fact that a race is going to start at a specific time means that as "off" nears, there will always be good value betting opportunities as each bookmaker tries to square up his book. There is a real risk that without proper controls, fixed-odds wagering will fall prey to the avalanche betting phenomenon in pari-mutuel wagering, where the pool doubles every 30 seconds over the last few minutes. At least with a pari-mutuel totalisator, the operator has a guaranteed margin, and the punter a guaranteed return.

There is varying elasticity and there is a point where reducing the "cost" of the bet will not increase total consumption.

I don't want to open up another can of worms, but.... The continuing dominance of starting-price (SP) betting in the United Kingdom off-course betting market (and off-course betting's increasing dominance over on-course betting) means that it is essential for the off-course bookmakers to manage their risk by betting into the on-course ring to ensure that the SP price reflects the overall weight of money. Some people call this manipulation of the market, but that suggests some malice, and really, they are only trying to make money.

Of course, off-course bookmakers would prefer to set their own SP. How's that for an understatement?

The off-course bookmakers go to all the trouble of maintaining a very expensive distribution infrastructure to promote their product (betting) on BHB's product (racing) and are snookered because either they (the bookmakers) are too lazy or inefficient to build a real book or they just can't get the British punter out of his entrenched SP betting habit. I say "lazy" because they have found it easier (and more profitable) to aggregate wagering data and bet on-course to move the SP toward a better result.

The Challenge

The weakening of the on-course rings over recent years has enabled them to do this more easily and at less cost; hence the latest shot over the bows by the Betting Offices Licensees' Association (BOLA) calling on the National Joint Pitch Council (NJPC) to enforce rule 18.1 so that on-course bookmakers can't use the betting exchanges to access to the off-course market, nor to deaden the effect of the chains betting to effect the SP (Racing Post, Sept. 6). The real target of this attack is the removing the depth and liquidity (and with the credibility and confidence in the prices offered) from the exchanges.

Mind you other countries have successful pari-mutuel totalisator systems to do this automatically and accurately.

That leads me to my next point: The challenge for bookmakers and the racing industry is for the horse racing wagering market to operate at an optimum, that is, to maximize revenue (or punter expenditure). Global (historical) data puts that (the optimum retention rate) at about 13.5 percent takeout for win/place wagers (its higher for exotics).

Without wanting to delve into an economic argument, the HMCE studies conducted prior to the reform of the betting duty highlighted a number of issues. But importantly, there is varying elasticity and there is a point where reducing the "cost" of the bet (or the gross wagering margin) will not increase total consumption (or expenditure). In fact, it is likely to lower it.

The nature of wagering, though, is the punter will always want the "best price." But, the "best price" is incompatible with the broader operation of wagering, most important of which is payments to the racing industry from the wagering nexus.

There is a reasonable argument that the BHB's decision to move to percentage of profit (gross wagering margin) from a payment based on turnover may have been a flawed decision (from the perspective's of both the racing industry and, apparently perversely, bookmakers). This doesn't mean, however, that the United Kingdom's HMCE decision to move to a consumption tax (GPT) based on gross wagering margin was similarly flawed (provided it doesn't let anyone escape the tax net).

The Effects

As recently as Monday, William Hill's chief executive, David Harding, claimed that that the betting exchanges weren't hurting his business. "I don't feel commercially threatened," he said. "At the moment, 95 percent of my business comes from the £5 punter in the betting shop, and he's hardly likely to bet with an exchange. ... Exchanges have probably taken some of my high-staking telephone and credit business, but I'm not worried for the totality of William Hill's operation" (Racing Post).

Harding left out the crucial qualification: today.

At the same time William Hill posted its results. Reuters reported, "Turnover at licensed betting offices leapt 48 percent to £1.12 billion, while telephone betting sales rose 36 percent to £243.5 million. William Hill's telephone betting business is the biggest in the world. Online betting sales leapt 78 percent to £197.6 million."

These figures highlight the dramatic growth in account wagering (to £441.1 million, or 40 percent of betting office turnover)--a far more efficient low-cost distribution channel--and online betting is growing at double the rate of "shop-front" betting (although with some apparent cannibalization of the telephone betting).

The same firms that operate the betting offices have to be successful in the account wagering environment to subsidize their expensive infrastructure to remain competitive overall.

Betting offices will undoubtedly continue to be an important distribution channel (a combination of cash betting, credit betting and the social aspect of wagering will ensure that), but one thing is for sure: The same firms that operate the betting offices have to be successful in the account wagering environment to subsidize their expensive infrastructure to remain competitive overall.

If the downward pressure on the prices phenomenon continues in wagering we will see a reduction in the gross wagering margin without a compensatory increase in overall turnover. That is, overall expenditure will fall. This fall will be all the more dramatic if there is leakage from the contributing operators (bookmakers, tote and attheraces) to non-contributing operators (betting exchanges and their "shielded" users).

The Solution

Without wanting to be seen to be a heretic, mandatory industry payments and/or taxes effectively create a "cost" or a floor price that the ensures the wagering operator operates on a level playing field. Bookmakers have long argued for profit-based taxation and industry levies but maybe it was a necessary evil. The United Kingdom, by having an excessive betting levy that was managed by deductions (at about 9 percent, which is broadly equivalent to the tax and levy paid in Australia by the off-course tote operators in each state), has been somewhat isolated from that and perhaps is having difficulty operating in an environment where the tax/levy component has to be encapsulated in the price (odds) offered.

That's not to say that the betting exchanges are not contributing to racing; Betfair has made a point of paying the BHB's data royalty in its revenues. But this is only on the "commission," which is levied at a rate of 2 percent to 5 percent of the punter's net winnings for a "market" (eg a horse race). Net winnings, not turnover!

So, while Betfair shields its users' profits from GPT and BL, it makes its revenue by assessing it and then charges only 10 percent to 20 percent of what the combined GPT/BL should be--25 percent of profit (net winnings).

However you look at it, the bottom line is the exchanges aren't paying enough on their revenues, and their users are paying nothing. Therefore, there isn't any tax or levy component built into the price each user offers/sells (or is prepared to pay/bid).

If this logic is applied to U.K. betting exchanges, then a minimum turnover fee of the order of 1.35 percent (based on 10 percent of the 13.5 percent "industry" average gross wagering margin) should be paid to the Levy Board so that the users of such exchanges would not be shielded from the requirement to make industry payments (and logically GPT). Thereby the distortion to the markets would be reduced.

As for HMCE's GPT, that's about another 2 percent of turnover, which is probably the only way to ensure that exchange users do not escape the tax net (unless it is calculated on a per-event basis with no offset for losses in other events you would end up with a secondary loss trading market).

Now, I suppose Betfair could pay that 3.375 percent of turnover out of their commissions, with licensed bookmakers who use the exchange systems getting a rebate benefit toward their GPT and betting levy liabilities if they properly account for their laying and hedging activity in the determination of their gross wagering margin.

The point of hedging is in itself a seminal issue because it also highlights that it is probably inappropriate to try and assess the profit of the "users" of the Betting Exchange for GPT and the levy. Gross wagering margin is based on the assumption that the person is a licensed bookmaker and is building a book taking bets on all runners in a manner to "ensure" that the total winnings paid out is less than total stakes received (which again demonstrates that a bet broker is not a bookmaker). Where credits are given for hedging, they are only permitted on bets back to another licensed bookmaker of stakes already received (that is, there is no credit for "punting" with the bookmaker's own money). The hedging bookmaker reduces his stakes received by the amount of the bets back; if the hedge wins then the winnings (including the stake) are accounted as "stakes received." The bookmaker that has received the bet simply accounts for it as he would for another bet.

The murky cesspit of the betting exchange creates all sorts of "transactions." A licensed bookmaker could be using it as a platform to lay bets. Providing he is accounting the bets received and paid out (for GPT & BL), the "punter" should escape any penalty. A bookmaker who hedges through the exchange (places bets) cannot receive any credit (unless the bets are with another licensed bookmaker). Punters betting with each other should be levied a fee of about 3.4 percent, which is decidedly cheaper than the old 9 percent deduction, but is based on the same principles.

One could reasonably argue that any B2B risk-management platform between bookmakers should not be open the "retail" punter, as it creates a false impression of what the "true" odds should be and creates a muddy environment with different classes of transactions. Call me a purist, but any system that encourages punters, and for that matter bookmakers, holding off betting to the last seconds by inducing avalanche betting (in search of the "best odds") is counter the interest of the sport. Hell, I want to be up in the stands watching it. Racing is not just about the betting!

At the same time, in the interests of the long-term health and vitality of racing, it is critical that on-course bookmakers have access to off-course punters via account wagering without causing compression in the prices, particularly the SP.

A 'Fourth Force'

In Australia--where fixed-odds betting is basically confined to racecourses only with an industry average 5.5 percent retention rate and there is an effective off-course monopoly for the tote with an average 16 percent takeout (14.5 percent win, 14.25 percent place)--there is currently significant discussion for the introduction of a patented "virtual bookie" system to provide a single off-course fixed-odds betting at inflated margin (approximately 13.5 percent) with the differential (8 percent) being used mainly to make racing industry payments commensurate with the tote (about 4.5 percent to 5 percent of turnover as against the current 1 percent paid for on-course bets).

However you look at it, the bottom line is the exchanges aren't paying enough on their revenues, and their users are paying nothing.

It can be seen from the above that compared to the United Kingdom, the Australian racing industry is on a pretty good thing with both the share and magnitude of its revenue from the wagering nexus, or alternatively, the U.K. betting industry gets off pretty cheaply.

How such a "virtual bookie" system would work in the United Kingdom is not immediately apparent; there is no need to make a turnover payment to the industry under the current "rules" based on 10 percent of GWM. Nor is there a compelling reason to encourage cooperation among on-course bookmakers (like gaining access to the off-course market in Australia). But there are a lot of arguments that support the same premise that all U.K. on-course bookmakers should combine together to create a single account wagering system operating in competition with the "Big Three" off-course bookmakers (with the system's operating profit distributed among the owners), thereby creating a real fourth force in U.K. bookmaking.

Perhaps attheraces could combine its tote betting systems with an on-course bookmakers' "virtual bookie" system to create a single account wagering platform that offered both tote, in particular exotic bet types, and a single competitive fixed-odds betting board that distributed bets back on-course. There is certainly a lot more synergy between on-course bookmakers and attheraces than with the "Big Three" from both their individual perspective. Like Australia's proposed system, atttheraces could then promote a day at the races where you accessed bookmakers, with all the color they add to racing, in a more competitive environment.

In summary, though, there is no doubt that, for racing--I don't want to touch the sports betting issue because its non-contribution to the sports that are wagered on creates a whole different scenario--betting exchanges and the so called virtual betting ring (or best price betting) business models for wagering are flawed, as they minimize both industry revenues and bookmakers' revenues, which are inextricably linked. They might be great for the punters that currently use them, but as they gain market share, they will slowly strangle the whole industry.

We have seen the inverse effect in Australia, where as the off-course tote gained market share the overall wagering margin grew over the optimum and as a result expenditure began to fall. But the totes didn't care as they kept making more money by virtue of their market share growth, thereby causing a net contraction in wagering (both turnover and expenditure) on racing.

The United Kingdom has undertaken a major structural reform in both its taxation and its industry payments via enforcement of intellectual property rights and it is leading the way for the global racing industry. It would be a tragedy if those reforms stumbled at the first hurdle.

Its time for the BHB, as one of the architects of the reform, to step up to the plate and contribute to this debate by throwing a very harsh light on the issues and articulating the importance of a level playing field for the racing industry, even if it's going to cost the punters a little bit more. It might be helpful if the bookmakers were a touch more forthright as well.

In Australia, a number of regulators in major jurisdictions are gloating (and with the benefit of hindsight, rightfully so) that they resisted the temptation to allow bet broking in their jurisdictions. Australian legislation, while allowing individuals to bet with each other, specifically prohibits anyone but a licensed bookmaker from advertising that they are willing to make or accept a bet, or to publish any odds. I know, as I was involved in the design and development of one such exchange system, and it is because of that rather painful experience that I know that betting exchange systems are not the panacea that their operators claim them to be.

How the big ugly genie is put back in the bottle in the United Kingdom I am not sure, but in there he has to go. The punters (particularly the "professionals") who have benefited from the operation of betting exchanges (not to mention the operators themselves) are going to scream like stuck pigs. The only easy solution I can offer is to stick an apple in their mouth and rotate slowly over a low fire.

The views expressed in this editorial do not necessarily reflect the views of Interactive Gaming News or the views of the Australian Bookmakers' Association, which has not examined the U.K. issues in any detail.

Tim Ryan is a graduate of Agricultural Science, though he is probably best described as an innovator. He has worked in consumer online information and transaction systems since their inception and was the Australian pioneer of television data broadcast and is still the Consultant General Manager of the Seven Network’s teletext and datacast services – the principal provider of real time wagering information in Australia. Tim was an innovator of integrated wagering information and transaction systems. He has also consulted to a number of wagering system developers on the internationalisation of their systems. Tim became a NSW licensed bookmaker early in 2000 for regulatory reasons associated with the development of wagering systems. Despite not being a ‘working’ bookmaker Tim was appointed to the Australian Registered Bookmakers Advisory Council, Australia’s peak body representing all its 800 bookmakers. In this capacity he was instrumental in lobbying the independent Senators and the Government to ensure the exclusion of wagering from the Interactive Gambling Act.