Steve Roman (who works in NYC’s financial industry) attracted my attention to this line from The Los Angeles Times about the world’s #1 betting exchange (real-money prediction exchange), BetFair:

[...] the Internet gambling colossus Betfair, which handles more daily trades than the New York Stock Exchange. [...]

I contacted BetFair about this statement, and Robin Marks (Head of Media at BetFair) sent me the data on their transactions since February 2005. I forwarded the BetFair data to Steve Roman for analysis. He has researched the transaction data for the main financial exchanges, and has compiled them along with the BetFair data, for comparison. You can download his spreadsheet file here —XLS file (BetFair data + Financial Markets data). If Steve Roman and BetFair are interested, we may update this spreadsheet on a monthly basis and see how things evolve in the future.

Here’s Steve Roman‘s analysis:

- BetFair’s trade volume is twice as volatile as the NYSE, LSE, and ICE. The CME’s volume volatility is in-between BetFair and the NYSE.

- As one might expect, there is a higher correlation in volume between financial exchanges than with BetFair and the financial exchanges.

- From June 2005 to December 2005, BetFair’s volume by number of trades was higher than the NYSE. It hasn’t been higher since that time.While the NYSE has shown steadily increasing volume, BetFair’s volume is more erratic, having a less discernible trend.

- BetFair’s August 2005 volume wasn’t bested until May 2007. LSE, ICE and the CME more than doubled in that time and the NYSE number of trades more than tripled.

Jason Ruspini remarks:

The ‘# of trades’ metric might be meaningless because it doesn’t reflect the dollar amount of the trades.

Robin Marks (the Head of Media at BetFair, which is located in London, United Kingdom) explains the high volumes experienced in the summer of 2005:

The 2005 large monthly figures were during the cricket ashes series between England & Australia which takes place every two years and England won for the first time in ages and the nation went wild !!!!

Jed Christiansen comments:

What’s interesting to me is that the Betfair data consists largely of trades by individuals, whereas the same cannot be said for the large stock exchanges. They must have a wealth of data on users and markets.

—

ADDENDUM:

Previous: 2005: TradeSports-InTrade’s Heyday

Price Discovery Among the Punters: Using New Financial Betting Markets to Predict Intraday Volatility – (PDF) – by Eric Zitzewitz – 2006-07-XX

—

External Links:

- BetFair – Wikipedia

- TradeSports-InTrade – Wikipedia

- Chicago Mercantile Exchange – Wikipedia

- New York Stock Exchange – Wikipedia

- InterContinental Exchange – Wikipedia

- New York Board of Trade – Wikipedia

- London Stock Exchange – Wikipedia

- Steve Roman‘s new site, Civil Rights Action Corporation

—

Thanks:

- Robin Marks of BetFair.

- Koleman Strumpf for a little piece of advice.

- Eric Zitzewitz for taking a quick look at our draft.

—

Related Page:

- Betfair results eagerly awaited

Eric Zitzewitz’s comment:

Just to challenge a little the conventional wisdom that commercial prediction market exchanges need sports to make money: in my paper on Tradesports/Intrade’s binary Dow options, I note that only 54% of Tradesport’s volume comes from sports. 35% is financial contracts, 7% politics/current events, and 4% is entertainment.

This suggests that a real-money exchange like Intrade could probably survive without sports (obviously, it would prefer not to have to). Take away the financial contracts too though, and things don’t look so good.

(Of course, a caveat to this market share-based analysis is that ignores possible economies of scope. For instance, it may be that running sports markets attracts the “dumb money” that you need for the financial and politics prediction markets to work.

For the data see: