Cash out betting

[Plouffy]

Hi All,



Was wondering while I was watching football and saw some betting adverts. It seems that betting companies are increasingly advertising their cash out option for bets (e.g. if you've bet that a team will win, instead of waiting for the end of the match, you can exit the bet and earn some profit this way).



Clearly, these companies must have done some analysis that shows that providing this options is profitable. My only analogy would be with american/european options, and how it's not optimal to exercise early calls. Does anyone know of a paper or an article where this is discussed?



Thanks,

Plouffy

[AndyM]

Punter is not exercising an option; they're just crossing another bid-offer spread to close a position. How can that not be more profitable for the bookies?

[Plouffy]

If I bet 10£ that Chelsea beat Manchester Utd 2-0 with 10:1 odds, it's the 85th minute, the score is 2-0 but I feel that Manchester might score so I cash out and get £60 instead. If Manchester doesn't score the booky will have saved £40, but if the score ends 2-1, the booky has just lost £70 (the £60 he had to give and the £10 from the punter he would have received).

[AndyM]

Yes, sure. Now work through an example where the bookie is running a matched book and both punters cash out mid-match vs stay invested.

[misc2014]

Had a more serious look at this not too long ago. I thought there may be an market making opportunity here especially with exchanges like smarkets.



Algo based on market depth + live event information. All observable variables. Layered positions in the order book according to expectation.



The trouble is API's for betfair and betdaq are unreliable. Befair add a 6 second delay by design during in-play. Smarkets look promising, are working on a streaming API and are keen on MMs.