Sports Trading (Option Pricing) Strategy

MrGecko

Senior member
Messages
2,778
Likes
789
A couple of people have mentioned that they do this now, and I wondered what types of strategies are used. I assume some people just work out what the implied probability is from the odds and compare this with the historical data - this, for me, sounds like a bad idea.

I was watching the national a while ago (whenever it was), and something of a lightbulb went off. I would like some feedback from anyone who can be arsed:

When you bet on a horse to win, your bet could be be described a vertical spread where the two strikes are infinitesimally close... (bear with me!)

Now, when you go to the races, the bookies are making markets in these vertical spreads, and the prices change. If we assume that all the bookies have identical information, they should all offer equal prices if the stastical liklihood of winning is a factor - it doesn't, it's a risk-neutral arrangement.

The bookies offer different (but similar) prices. Of course, the price the bookies offer does not represent the odds of whichever horse winning, but the size of the liabilities that the bookie has on that particular outcome - the bookie wants to secure his 107% (or whatever), which you can basically consider his spread for making a market. He doesn't bet on the race, he has no edge there - he bets on the size of bets he is going to get (?).

Now - here comes the interesting part; Would it not be possible to front run "retail betting flow"? Lets say the market for a particular race opens 2 hrs before the event... the prices at the open should differ from the prices 10 mins befre the race, because the "retail" crowd will:

i) only care about the next race on the card, then come 1hr 50mins later...

"Weighed in, Weighed in"...

ii) bet on the horse that Franki Dettori is riding
iii) bet on the 200-1 shot with a cool-sounding name

all while sinking the plonk w/ hot pork+apple sauce rolls.

... comments welcomed.
 
Major difficulty is that most of the volume on a typical ****ty meet in yarmouth or something will be about 5 bets by a whale or two... and here the bookies do have an edge over you as they have a much better idea who these guys are and what they're likely to put money on.

In general, front running on betfair is certainly possible. Favourites definitely seem to tighten... and you'll also see that favourites comprise most of the volume traded pre game to.
 
On football, but not enough to form a statistical sample, it's just an impression I have.

The only main thing I've used that impression for is the matched betting I'm currently doing... I'll use this impression as part of my execution strategy.
 
by matched betting I guess you mean securing yourself exposure to a result without any liability? Basically arb?

And football specifically because of the (I assume) volumes it gets?

The trigger for me was watching the price of some Donkey go from 200 (-1) to 50 in the space of 10 mins. The sport doesn't matter, the crux here is taking a view on the flow of the bets, not the liklihood of a result.
 
It's arb, but it's bonus whoring arb - the majority of it the bet will be laid off for a guaranteed small loss (although I've had a couple of arbs coincidently) but the bonus makes up for it.

Plenty of good bonuses around.

That isn't to say that there aren't arb opportunities around between the bookies, but I haven't yet run out of bonuses.
 
What I'm describing isn't arb - it is speculating on prices changing due to customer flow rather than the liklihood of a particular result.
 
I know, I was just letting you know where I was coming from...

Get yerself a copy of bet angel (you can pay by the day and I think they do a trial) and have a go at scalping a game pre match. Basic vanilla DOM based scalping works.
 
.... continuing the football theme...

What about the demographics of the people that bet (price takers) on a football game?

Initial thoughts:

* the team with the bigger fan-base will have more bets "to win" than their opponent (known in advance)
* those that bet on statistics can be discounted (the stats are in the public domain)
* bets on football = -ve expectancy, +ve utility
 
if you watch any england football match on betfiar the price is incorrect in the run up. precisely the effect described above.

to make this work you'd have to trade in decent size and on events with enough interest/flow. world cup, national or events with enough outcomes that you know stupid people will bet on the most popular.

bookies know this already-they just use it to build up their matched liabilities giving them more firepower to offset a bet elsewhere.
 
I do wonder if asian markets might be the best, given studies show something like 1 in 10 asian men is a problem gamblier (and anecdotally it seems to be even more)

Certainly Hong Kong's parimutual betting is supposed to be a goldmine if you're not a ****wit who bets on everything with the number 8 in.
 
I did an arb this morning. Backed Oratory for 5:45 goodwood 11/4 Skybet, laid him off 3.6 on betfair. At 5% comission that's 35 pence every £100.

Handsome!
 
I wonder if you could employ a similar strategy on Equity options in the run up to earnings?

(Not your Arb Dave, but the "front run flow" strategy)
 
on flow from the retail sector?

Yeah, say for example on Apple because of the iPad, or BP because of the spill.

The idea would be to try and identify early the stocks that people are going to take a punt on.
 
i guess so but there too many different factors in the equation when it comes to stocks. pension fund holdings/rebalances/hedges/stock shortages.

"The idea would be to try and identify early the stocks that people are going to take a punt on. "

and surely this is what every stock analyst on the planet tries to do...
 
Yeah, I appreciate there are many more variables in the Equity option example - just bouncing it around, thats all.
 
Top