Editorial: Smoke and Mirrors - the Betting Exchange Shell Game (Page 2)

4 October 2002
Back to Page 1

Table 2: Betfair, Yarmouth Race 1, 18/9/02 - A Heavily Backed Winning Odds-On Favorite

Betfair Matched
Trades
Betfair Avg. Odds (decimal) Betfair Backers' Stakes Betfair Layers' Liabilities Total Escrowed Stakes or  Real Turnover Max. "Net  Winnings" (if Winner) Maximum Gross Wagering Margin Betfair Commission @3.5% of  Max. Net Winnings Margin on .5 Matched Trades (70,741) Margin on Total 'Turnover' (547,938)
Camlet 423,706  1.56 211,853  119,457  331,310  139,290  27.1% 4,875  2.1% 0.95%
Frieda = Kahlo 26,450  4.93 13,225  51,963  65,188  270,424  52.6% 9,465  4.1% 1.84%
Calligraphy 7,050  10.70 3,525  34,184  37,709  262,345  51.0% 9,182  4.0% 1.79%
Masaader 4,692  18.64 2,346  41,378  43,724  270,718  52.7% 9,475  4.1% 1.84%
Our Lady 1,474  49.13 737  35,471  36,208  266,420  51.8% 9,325  4.0% 1.81%
Total/Expected Value by P(n) 463,372  100.98% 231,686  282,454  514,140  186,573  36.3% 6,530  2.8% 1.27%

We can see in Table 2 that in this case, the heavily backed favorite (in a small field) winning was the worst result for Betfair, although the alternate perspective, that any other result would have been spectacular, may be closer to the mark. This table does illustrate, though, that where all the trading is concentrated on one sub-event, the resulting real turnover, while still greater, is no longer a multiple of the matched-bets volume.

Table 3: Betfair, Sandown Race 1, 18/9/02 - A Real Roughie Won What Was Otherwise 'A Good Betting Race'

Betfair Matched  Trades Betfair Avg. Odds (Decimal) Betfair Backers'  Stakes Betfair Layers' Liabilities Total Escrowed Stakes or Real Turnover Max. "Net Winnings" (if Winner) Maximum Gross Wagering Margin Betfair Commission @3.5% of Max. Net Winnings Margin on .5 Matched Trades (70,741) Margin on Total 'Turnover' (547,938)
Blue Knight 19,588  5.58 9,794  44,826  54,620  105,773  19.3% 3,702  5.2% 0.68%
Strathclyde 21,048  6.21 10,524  54,831  65,355  115,048  21.0% 4,027  5.7% 0.73%
Esatto 50,805  5.56 25,403  115,754  141,157  161,093  29.4% 5,638  8.0% 1.03%
Belia Chica 18,936  8.77 9,468  73,570  83,038  134,843  24.6% 4,720  6.7% 0.86%
Jayanjay 12,669  8.92 6,335  50,197  56,531  114,603  20.9% 4,011  5.7% 0.73%
Bryano de Bergerac 5,589  9.51 2,795  23,774  26,568  91,720  16.7% 3,210  4.5% 0.59%
Matty Tun 8,211  11.09 4,106  41,442  45,548  108,078  19.7% 3,783  5.3% 0.69%
Three Days in May 2,113  25.07 1,057  25,432  26,488  95,116  17.4% 3,329  4.7% 0.61%
Kelsey Rose 1,565  33.27 783 25,253  26,035  95,211  17.4% 3,332  4.7% 0.61%
White Ledger 958 47.18 479 22,118  22,597  92,380  16.9% 3,233  4.6% 0.59%
Total/Expected Value calc P(n) 141,482  103.28% 70,741  477,197  547,938  118,802  21.7% 4,158  5.9% 0.76%

Table 3 demonstrates a pretty even betting race with no clear favorite, and despite the absolute roughie getting up, expected Margin was 21.7 percent (against an average of 20.3 percent) and expected commissions was £4,158 against an actual of £3,233. While margin on real turnover is lower (0.71 percent), this was compensated by the much higher multiple of matched-bets (3.8 x) or 7.75 times what would have been assessed as bookmakers' turnover.

These three examples show us that determining the average multiplier to make an assessment of real turnover is not easy, as it depends on both the number of runners and on the betting patterns for each race. However, it is reasonably estimated to be at least double the matched bets number.

Gross Profits Tax (GPT)

If Betfair (as it should be) is calculating its GPT liability as per the Finance Act 2002 amendment s5C BGDA 1981, in our first example the gross wagering margin* was £72,827 with a GPT liability of £10,924. But, if the favorite had won, its GWM would have been a loss (£96,510) or a carried forward GPT credit of £14,476. In our second example, the GWM was a loss (£99,624), and in our third example, the GWM was £48,144.

*GWM = sum total of all backers' stakes, liability on winner, returned stake on winner.

As GPT is paid based on the sum of all wins and losses from each event over a month, the GWM will logically fall to the overall retention rate (calculated from the traditional bookmaker's perspective), which is likely to be very low (and off a low base as would be calculated on half-matched bets (backers' stakes).

To consider the effect of this, we should look at the expected values based on probability (which, given the over-round has been eradicated, should be a reasonably accurate assessment).

In our first example, as we saw above, HMCE had a great win with Betfair's GPT liability exceeding commissions by almost £6,000, while Betfair's expected GWM was only £2,249; GPT was £337 (or 5.2 percent of £6,530 expected commissions).

In the second example, HMCE had a loss--a real soul destroying near £100,000--although Betfair's expected GWM was only £3,285; the GPT was £493 (or 9.4 percent of £5,235 expected commissions).

In the third example, while HMCE had a good win with GWM of £48,144 and a GPT of £7,222 (or a loss for Betfair of over £4,000), Betfair's expected GWM was only £2,248; the GPT was £337 (or 8.11 percent of £4,158 expected commissions).

The actual GWM over these three (randomly selected) events was £21,347 with a liability of £3,202, whereas the real revenue (commissions) was £13,210, so the effective GPT rate would have been 24 percent. But if all three favorites had won, we would have had a GWM loss of £266,550 to carry forward. Looking at the expected values, the total GWM was £7,702 for a GPT of £1,167 against commissions of £15,923, or 7.4 percent--about half of what it should be paying (ignoring the avoidance of Betfair's punters)

Therefore, in calculating GPT liabilities, the coin-toss shell trick works in the betting exchange's favor... Again!

It is likely that any exchange's long-run GWM could be below 2 percent of bookmakers' turnover or the half-matched-bets. It would certainly be unlikely to reach the dizzy heights achieved by Betfair, 4 percent to 6 percent of its race-by-race margin on 0.5 matched-bets--so let's be generous and say its half. Then Betfair's maximum GPT liability is probably less than 7.5 percent of its gross revenue (commissions), as our simple three-race example above demonstrates.

Of course, under the alternative two-bet approach advanced in Bet Unfair!, total received commissions would be counted as stakes received without any payouts, so the GPT liability could be locked in at 15 percent of commission revenue.

Where's that pea? Maybe the BHB knows where it is. At least it opted for receiving the data royalty on the exchanges' commissions. Just luck? It's only an educated guess, but my estimate of Betfair's monthly GPT payment (on racing alone) is approximately £325,000. But, perhaps it should be at 15 percent of commissions (which are estimated at £4.3 million) viz. £650,000. Either way, its a lot of money--up $2 million of my Aussie dollars (per month)--and probably more than enough for the HMCE not to contemplate auditing its operations. It was generating nothing in betting duty last year! But, it would appear that HMCE has robbed itself by assuming that a betting exchange would be profitable from a traditional bookmaker's perspective.

Betting Levy

Under the previous (40th) levy scheme 1.2 percent of turnover was levied on Internet wagering transactions. Betfair's matched bets on racing has been put at £45-£50 million pounds per week. However, we have seen above that rather than halving this number to determine turnover, we should, in fact, be doubling it (at least). On that back-of-the-envelope basis, Betfair's levy payment should be more than £62 million pounds per annum, but the levy is being paid at 10 percent of 1 percent (of true turnover), so only about £5.2 million will be payable. (The alternate calculation of 10 percent of 4 percent of half-matched bets gives the identical number, £5.2 million.)

Wait a minute! Both the BHB's and HBLB's forecasts for the 41st levy is only £110 million (including from bookmakers in Ireland). Obviously the £62 million per annum (from Betfair's turnover alone) is a ludicrous proposition. ... But is it?

What I am trying to highlight is that betting exchanges are operating under a different set of rules. So, while the bookmaker levy payments have gone up, they have gone up with both their turnover and revenue. (That is, the bookmakers have maintained their margin.)

Betfair is doing, on an escrowed funds basis, a very, very big turnover for a very fine margin. Its £50 million week (matched bets) multiplied by 2 (to get an assessment of total escrowed funds) multiplied by 52 (weeks) gives a real turnover estimate of £5.2 billion per annum. Now, to put that in context, according to HMCE, total stakes for the GBD (off-course bookmakers gross betting duty) in 2000-01 was £7.221 billion per annum (down from £7.295 in 99/00). By these numbers, Betfair alone is turning over 72 percent of 2000/01 off-course bookmaker turnover.

The point is, under the old levy rules (1.2 percent of turnover), they would have to pay £62 million per annum. Bookmakers are still paying this fee; in fact, on 10 percent of GWM of about 13.5, they are paying 1.35 percent of turnover.

Exchanges are happy to hide behind the half-matched bets because doing so does two things: It quarters their (estimated) real turnover (making them a less perceived threat to the bookmakers) and inflates their revenue (estimate £52 million) to betting turnover ratio to 4% as against 1% (making them seem 'closer to the real thing').

There is such a huge disparity between these numbers (£62 million v. £5.2 million) that something has to. I would argue that it may have been wrong for the BHB to abandon a turnover based fee for the use of its intellectual property. In the interest of a level playing field, it would appear more equitable to levy a cost-- as it was in the past--on punters' turnover.

The reasoning behind this is simple: It's ultimately the punter who pays for the intellectual property used to bet. Licensed wagering operators are simply a convenient collector of those fees (as is any retailer for VAT/GST). In the interest of keeping a level playing field (assuming the activities of betting exchanges in respect to horse racing go on unchanged), the only logical (equitable) method of charging is on true turnover--what the punter actually stakes.

Page 3

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.