Start a new topic

arbitrage

If I own Yes shares am I allowed to purchase No shares to play a spread?


9 people like this idea

Well for example, there are two markets asking the exact same question: who will win the whitehouse in 2016. There's the "will the Dems win" market and the "will the reps win" market. A no in one is a yes in another. So you should be able to sell a "no the republicans won't win" to cancel out a position of "yes the democrats will win" otherwise the markets start being dumb. As it stands I can buy the same position in different markets and just hold them till the end and make a guaranteed profit (if I buy both side for less than 50 cents) that opportunity should not exist for longer than a split second in a functioning market.

2 people like this

Yeah, it makes more sense when you're looking at multiple, mutually exclusive events. Unfortunately, when there's more than two options (as there would be, otherwise it would just be a single YES/NO event), you'd need to invent the concept of turning one YES share into many NO shares (and vice versa), and that might be tricky for the devs to implement. What do you do when someone puts in a limit order to buy a YES, it generates a sell order on a spread of NO shares, and then one NO gets bought and the others don't? Probably best to keep things simple and inefficient until the market gets a bit more liquid.

What you describe is not arbitrage, it's hedging.


1 person likes this
If the goal of this site is to produce price discovery - i.e. accurate/efficient market predictions - then Predictit should allow arbitrage. The combination of not allowing it and a confusing setup of separate YES and NO markets, is making for awfully large inefficiencies, which means the pundits won't be particularly interested. If the odds for all the Republican candidates adds up to 150%, for instance that obviously makes no sense. More money coming in will help, but so would the ability to arb, and so would a simpler system (a la the late Intrade).

As an aside, there are a number of posters who seem confused as to what constitutes arbitrage.

 


1 person likes this

The YES and NO markets are not actually separate though, so this question only applies to markets with multiple mutually exclusive options.


How did Intrade handle such cases?

The YES and NO markets are both separate and intertwined at the same time, which is why it's not terribly intuitive. Intrade simply had a single market, which you either went long or short. The people behind Predictit seem to think that the shorting part was confusing, so they created NO markets so no matter what you believe, you are essentially going long. My own view is that this is far more confusing, and this view is borne out by the fact that NO markets appear to be way more inefficient than YES markets.

 

What do you mean by saying that the NO markets are less efficient? YES and NO orders write to the same order book, and so they always have the same spread, and represent the same information...


1 person likes this
Here's what I mean:

I don't think many people fully understand the NO markets. They come to the site and place money on YES bets they favor. Thus, YES prices are routinely inflated, something you can really see in multiple proposition markets like who will win the GOP nomination. If you add up all the probabilities, they routinely add up to way more than 100%, which obviously makes no sense. It's way better to bet NO in these markets.

So, let me amend what I said. It's not that NO markets are more inefficient. YES and NO markets are equally inefficient, it's just that the inefficiency tilts exclusively in one direction, i.e. NO.

This systematic bias isn't ideal for price discovery, something fans of prediction markets are keen on. Perhaps it's something that will sort itself out with more liquidity, but that shouldn't have to be the case. In the meantime, I will take solace in the trading opportunities that the inefficiencies provide.

 


1 person likes this

It seems more likely to me that the YES price inflation is due to the much lower dollar tie-up of buying YES shares in a multi-proposition market.


1 person likes this
Candidly: Your answer 3 years ago still applies now, IMO, that PI is (allegedly!) for research purposes. People who want PI to mimic Stock Markets need to go play in them, if that is what they enjoy the most. Personally, I enjoy watching and learning the psychological factors involved in the fluctuations and resolutions of PI markets, by both the PI players and the public at large (Politicians, Constituents, Mass Media, etc) as much, if not more than the opportunity to make a buck or two profit here and there. re:"...The concept of arbitrage seems to me to be contrary to what the site was intended to do -- measure the public perception of whether a political event would occur. If someone is more interested in "making money" from the shares here than registering an opinion, it's not that hard to corner the market..."

1 person likes this
Login to post a comment