How a sports betting hedge fund works(businessweek.com)
businessweek.com
How a sports betting hedge fund works
http://www.businessweek.com/magazine/content/10_29/b4187069936116.htm
35 comments
The legality of betting is interesting. Why is it illegal to financially speculate on the outcome of sports games? Why is it illegal to pay someone to play badly? I can see how sports leagues would prohibit the latter, but I don't see how the Federal government has the authority to control that. As long as you pay taxes, I don't see a problem for society. (Except that professional sports would suck, but that's not the government's problem.)
The main reason betting itself is illegal is for moral reasons. In many societies, betting is considered a vice just like drinking alcohol, presumably because of its non-serious nature and the general atmosphere of lewdness that usually surrounds it.
Apart from that, the addictive aspect of gambling is in many places a major argument to prohibit or severely restrict gambling.
As to why paying someone to play badly is illegal, it's because the information asymmetry distorts 'normal' trade relations. It's a consumer protection mechanism. Of course it could be dealt with by a civil claim, but that may not have enough of a deterrent effect.
Sports leagues don't have much power in this matter. They can only influence the player being bought off, not the one paying the player.
Apart from that, the addictive aspect of gambling is in many places a major argument to prohibit or severely restrict gambling.
As to why paying someone to play badly is illegal, it's because the information asymmetry distorts 'normal' trade relations. It's a consumer protection mechanism. Of course it could be dealt with by a civil claim, but that may not have enough of a deterrent effect.
Sports leagues don't have much power in this matter. They can only influence the player being bought off, not the one paying the player.
While in other societies drinking alcohol is not considered a vice, speculating on the outcome of sporting events is considered a relatively harmless pastime and "gambling" on arcane financial events (and the associated lewd behavior) is considered by most to be immoral when those who benefited get bailed out with taxpayers money.
I think the reason is that you can launder money this way.
What is also interesting to me is why the internet betting is prohibited in USA, but buying stock or forex trading is not prohibited. So you can't bet on sport outcome, but you can bet on companies stock prices or currency fluctuations.
What is also interesting to me is why the internet betting is prohibited in USA, but buying stock or forex trading is not prohibited. So you can't bet on sport outcome, but you can bet on companies stock prices or currency fluctuations.
Money laundering is a big problem for betting exchanges, but it's not an unsolvable problem. Countries that issue betting exchange licenses generally have very tight anti-money laundering requirements.
Betfair has an api, http://bdp.betfair.com/, if you're interested in writing your own trading bot - although the free version doesn't allow you to place bets.
There's a group of Python developers in Hammersmith who are Betfair's number 1 client by volume. The founders are an ex-trader and an ex-professional gambler.
They write their own market data and execution feeds from various sports betting facilities, abstracting everything, and handle it exactly like they're running their own exchange.
They were looking for people a little while ago - I had an interview there and bombed after spending an hour trying to find their office - but they're friendly guys.
They write their own market data and execution feeds from various sports betting facilities, abstracting everything, and handle it exactly like they're running their own exchange.
They were looking for people a little while ago - I had an interview there and bombed after spending an hour trying to find their office - but they're friendly guys.
>> They write their own market data and execution feeds from various sports betting facilities, abstracting everything, and handle it exactly like they're running their own exchange.
What exactly do you mean? The have their own betting agency and selling their customers bets to betfair?
What exactly do you mean? The have their own betting agency and selling their customers bets to betfair?
They create things - pricing feeds, execution APIs - that in the non-sports betting world you'd normally expect the exchange to provide.
It's prop trading only AFAIK.
It's prop trading only AFAIK.
Yes it does, but it limits you to 100 per minute. (http://bdp.betfair.com/index.php?option=com_content&task...)
We (http://www.smarkets.com) are about to launch our own free RESTful trading API.
We (http://www.smarkets.com) are about to launch our own free RESTful trading API.
Thanks for the correction. How do you plan to compete with Betfair?
We want to make betting on a betting exchange easier to understand for the average punter. Loads of people (at least in the UK) still use high-street bookmakers with bad odds. Betfair is a bit of a mess (don't view source if you value your sanity).
1) Is it legal for me, as a US citizen, to use your product? If it's not, are there any alternatives available?
2) When will the RESTful API be available? Can you email me when it's ready?
2) When will the RESTful API be available? Can you email me when it's ready?
I don't think it's an offence for a US citizen to gamble over the internet, but it's not legal for us as a betting exchange licensed in Malta to let you use our product. So you can register on Smarkets, but you won't be able to deposit from a US IP or a US credit card.
Also the Smarkets founders are US citizens and they'd quite like not to get arrested when they go back home.
For API news, follow our blog: http://blog.smarkets.com/
Also the Smarkets founders are US citizens and they'd quite like not to get arrested when they go back home.
For API news, follow our blog: http://blog.smarkets.com/
I don't want to have to follow an entire blog to learn when your API is going to be ready. Can't you just email me?
But where does the bot obtains current live score? Or will the bot "watch" only the odds movements? I started recently to trade tennis on betfair, but not sure how it could be automated without the bot knowing the current score.
Depending on the market you are betting on, how quickly you need updates and what you can afford, you have a choice of subscribing to a live data feed (e.g. from http://www.enetpulse.com/ or http://www.pressassociation.com/sport/data.html), scraping a live score website or updating manually from the venue or live TV (although this does have a delay!). You do not always need ultra fast updates since:
Some markets are suspended and all unmatched bets cancelled when a significant event happens (such as when a goal is scored in football).
Any in-play market places a 5 second delay between placing a bet and having it matched, giving the market makers a chance to get out.
Some markets are suspended and all unmatched bets cancelled when a significant event happens (such as when a goal is scored in football).
Any in-play market places a 5 second delay between placing a bet and having it matched, giving the market makers a chance to get out.
Did you try to write some bot yourself? If so, for which sport? You can also PM me if you want, I do have interest in this area.
I have written a few bots before but only in certain soccer markets such as over/under goals.
That's cool, was it long-term profitable? I traded under 3.5 and 4.5 (manually) in a few matches with quite a good results.
They have been profitable, although I don't use a fully automated strategy yet, the bots show me where to bet based on statistical analysis (Monte Carlo analysis) based on past games, the current situation in each game and the available odds - I place the bets myself combining information from the bots and my own subjective analysis. The fact that profits are tax free helps! I'm not sure how feasable full automation is, maybe in certain sports like cricket that can be analysed well through stats alone, but for sports like soccer it is difficult to build an accurate statistical model that can outperform a market that can make subjective judgements based on the precise circumstances of a game. Reading the article, it seems like this hedge fund is using experienced traders to provide judgement backed by statistical modelling too.
I've been chatting with a few people and they confirmed to me that outperform football market long-term (by statistical model or any other way) is indeed very difficult, and most professionals currently do tennis and horses. The reason is that predict a goal in football is difficult, while the tennis match can be partially predicted by watching the game.
It's absolutely untrue that Galileo is the first sports betting fund. There are quite a few market makers on Betfair.
But I would imagine a lot of those market makers are betting their own money. This is a fund i.e. they're betting other people's money. I won't be investing!
So they simply bet on sports results after thoroughly researching the odds? Where is the hedging in that?
It works like this: for example, you bet before the football match that the result will NOT be a draw. Then, when favourite scores a goal, you can place a second bet, this time you bet that it WILL be a draw. But after the favourite scored a goal, the odds for draw moved up (since it is now less likely that the game ends in a draw). So no matter what will be the result, you win (it's exactly the same principle as with stock trading or forex, where you sell when the price is high or buy when the price is low).
Of course, it might happen that as the match going on, the draw is more and more likely, so the odds for a draw is getting lower. Then at some time, the trader can still bet on a draw (like in the first example), but this time he will have a loss, but a lot lower loss as if he would place just only one bet and lost. And that's I think is called hedging, you bet on all outcomes in different time points.
Of course, it might happen that as the match going on, the draw is more and more likely, so the odds for a draw is getting lower. Then at some time, the trader can still bet on a draw (like in the first example), but this time he will have a loss, but a lot lower loss as if he would place just only one bet and lost. And that's I think is called hedging, you bet on all outcomes in different time points.
The process you've described isn't really hedging, it's effectively closing out a trade, either through booking a profit, or mitigating a loss. The equivalent in the equities world is roughly these two cases; today you buy some shares in Acme Oil Company, tomorrow they announce that they've discovered a massive oil well and their share price rockets, so you sell and book a nice profit; or today you buy some shares in Acme Oil Co, tomorrow they announce they've sprung a leak off the coast of Florida and their share price falls, you sell your shares to mitigate the losses you've suffered (and the potential for more losses).
That is not hedging; hedging is the process of considering (at the point of purchase) how you can protect your investment against unquantified events. For example you might decide that Apple as a company is undervalued so you want to buy some of their shares, but the markets seem volatile at the moment so you want to hedge yourself against large movements in the index.
You can do this by selling short on the index at the same time as you buy your Apple shares. So suppose you're correct about Apple (that they're undervalued) and the market realises this and their share price increases by 1% relative to the index, without your hedge this could see your investment say gain 4% if the index goes up 3%, or lose 2% if the index falls by 3%, so you get either +4% or -2%.
Now suppose that you have your index hedge in place, if the market rises by 3%, you get 4% on Apple -3% on the index = 1%, if the market falls by 3% you get -2% on Apple and 3% on the index = 1%. So either way you get the 1% profit through your Apple intuition. By selling the index you've hedged yourself against movements in the underlying market.
So to hedge a bet on something like Roger Federer winning against Rafael Nadal in a match today, you need to try and find something else to bet on at the same time. I can't really think of how to do this though.
That is not hedging; hedging is the process of considering (at the point of purchase) how you can protect your investment against unquantified events. For example you might decide that Apple as a company is undervalued so you want to buy some of their shares, but the markets seem volatile at the moment so you want to hedge yourself against large movements in the index.
You can do this by selling short on the index at the same time as you buy your Apple shares. So suppose you're correct about Apple (that they're undervalued) and the market realises this and their share price increases by 1% relative to the index, without your hedge this could see your investment say gain 4% if the index goes up 3%, or lose 2% if the index falls by 3%, so you get either +4% or -2%.
Now suppose that you have your index hedge in place, if the market rises by 3%, you get 4% on Apple -3% on the index = 1%, if the market falls by 3% you get -2% on Apple and 3% on the index = 1%. So either way you get the 1% profit through your Apple intuition. By selling the index you've hedged yourself against movements in the underlying market.
So to hedge a bet on something like Roger Federer winning against Rafael Nadal in a match today, you need to try and find something else to bet on at the same time. I can't really think of how to do this though.
Thanks. That was what I was looking for in the article.
You can short a bet (known as a laying) as well as go long (known as backing).
http://vimeo.com/8995197 (done in 2006, so a little dated). Disclaimer - Sean Park is one of our investors.