Topic: Intrade arb bots

I briefly thought about adopting my bot to do intrade/betfair arbitrage earlier this year (when the Clinto/Obama contest was quite volatile), but decided I couldn't be bothered when I realised there's no free intrade API.  To quote from their website:

"A monthly fee for API access may be charged which may in turn be refunded in full if the member generates sufficient trade volumes ... Market makers may be entitled to free use of the API if they reach an agreement with the exchange and are designated as a market maker."

I would be amazed if there aren't a significant number of people doing it already though.

fred I would have to disagree with your statement

"if there is any significant money to be made then bot users have the will and ways."

Ive seen a few simple (even possibly obvious) things that make a significant amount of money. Im of the conclusion that the gambling markets are still far far from efficient. My conclusion is there more than likely opportunities there if you are smart and dont do the std. (best back > best lay) > commission.

fred, not meant as a flame, I more meant it as words of encouragement to intradeuser.

Ive seen a few things that i was convinced before hand couldnt work. I think people to often make the assumption that if it would work then someone would already be doing it.

No i really dont. I have a good understanding of the financial markets. This is why i remain confused by some of the dynamics on bf.

One of the main things Im amazed works are strategies like buy when price is x. It makes no sense how that could ever work. I'd agree its not supendous in terms of return, but its something that just logically shouldnt work. An efficent market would presume that that E[profit] = 0, i'm sure at most of the people here have some stat arb bot that proves this principal wrong.

There are two points I can see,

The prices that you have shown are only £0.04 different  1.5 in bf and 1.54 in itrade. This outright you could not make money from due to commision. (assuming 5%).

                                     Win           Loss

lay    +5000    @ 1.5      £-2500        £5000

back -4870.12 @ 1.54    £2629.87    £-4870.12

ex comm                      £129.87      £129.87       

comm @5%                  £-1.67         £-120.13 

comm @ 2%                 £77.27         £29.87

At standard commission rates you're looking at a loss on these trades.  

The second point would be that you're locking up £10,000 to make max £77.27 this is assuming you can get a deal and get decent commission rates.

The main reason that ppl would not take advantage of this opportunity is that arbitrage is in general very capital intensive. The profit per trade is pretty low compared to other techniques so youll need to make sure you do a lot of it; this makes your liquidity the most important thing if you'd like to make a lot from aggressive arbitrage.

Out of interest, I just checked back over my spreadsheet of bets, to find out what made me think about writing a BF/intrade arbitrage bot (even though I didn't do it in the end).

On 2 February (according to my spreadsheet) I was able to sell McCain on intrade to win the Tennessee primary at 55 (i.e. 55% implied probability) and back him at Betfair for 2.12 (i.e. 47.2% implied probability), which was enough to give a pretty good profit even after commission.  So at that point there definitely were profitable arbitrage opportunities that weren't being hoovered up by bots.  Worth noting though that the amounts I bet were pretty small - I guess I would have done more if it had been available.

Of course (as stated in the above posts) arbitrage is a lot more attractive in a market like that, with only a few days to run.

The comm you pay is relative to the price you take the arb at.

basically,

comm  = (back_price <= 2.0)?lay_stake*0.05:back_price*0.05; // assuming 5% commision rates

if the price is lower than 2.0 then the most commission youll pay is on the stake in the exchange where you had the lay.

else the maximum commission you'll pay is if they win and you'll pay commison in the exchange where you had the back.

you want to make sure that the arb spread is bigger that this and you'll be ok.