Wednesday, March 3, 2010

Playing the prediction markets

This article provides an introduction to both utilising implicit predictions in binary options, as well as some pointers on trading. Prediction markets provide an interesting venue for gauging the probability of an event based on the trading of both informed and uninformed participants.

As an investor, trader or someone with a general interest in things like economics related events you can use predict markets to help get an idea of the probability of the event occurring. You can then take it a step further and take potentially profitable positions if your view on the probability contrasts with the markets’ view of the probability.

Prediction Markets and Binary Options
So what are prediction markets? Prediction markets generally involve the trading of binary options. You may know what options are (the right but not the obligation to purchase a given thing), and the benefits of using them (nonlinear payoff profile i.e. fixed premium/cost vs variable profit). But binary options work a little differently, the standard binary option pays $1 if a specified event occurs by or on a specified date – otherwise it pays $0.

For example ipredict has a contract on the US Fed increasing interest rates by November (here): “FED.INCR.NOV10”. This contract pays $1 if the Fed increases interest rates on or before the 4th of November 2010.

You can both sell and buy binary options. So using the previous example, if you believe the probability of the US Fed increasing rates on or before the 4th of November is greater than 0 then you would buy contracts (e.g. if you bought a contract at $0.50 and the Fed increased rates before expiry you would receive $1).

Likewise if you believed that there was no way the Fed would lift rates this year then you could sell short the contract. So for example if it were trading at $0.50 then you would sell a contract for $0.50 and on expiry if the event did not happen you would get to keep the $0.50. But of course the converse is true, if the event did occur then you would have to pay $1 to the holder of the contract, but this would be offset by the $0.50 you sold it for.

How to Read the Market Predictions
So by now you probably can already answer this question, you gauge probabilities based on the price. If the price is $0.50 then on average the market is pricing in a 50% chance of the outcome.

To look at an example, on ipredict there is a suite of contracts on the RBNZ (Reserve Bank of New Zealand) interest rate decision on the 11th of March. At the time of writing the approximate prices of the contracts were as follows:

OCR.NC.11MAR10 Price $0.9650 …therefore probability = 96.5%
OCR.25.11MAR10 Price $0.0300 …therefore probability = 3.0%
OCR.OTH.11MAR10 Price $0.0050 …therefore probability = 0.5%

So at the moment you can see the market is overwhelmingly expecting no change to the Official Cash Rate on the 11th of March. You can see how the probability aspect works with this suite too by noting that all outcomes (all mutually exclusive and exhaustive) add up to 100%.

How to Make Money Playing Prediction Markets
So now, down to business! I bet most of you have already started plotting how to make money, but this section sets out the basic strategies for making money by playing the prediction markets.

1. Positioning
The first way is what I call positioning. Basically you find contracts where you’ve got a strong opinion on the outcome and position yourself to profit from that view. For example with the Fed rate increase contract, if you believe there’s about a 70% chance of it happening then you would buy it all the way up to $0.70 (possibly higher).

The key things to think about are: a) risk and reward profile e.g. if the price is $0.70 then your profit if it happens is $0.30 ($1-$0.70=$0.30), but if it doesn’t happen you lose the $0.70 you paid to buy the contract. This is a sub-1 trade; the ratio of profit divided by loss is 0.43. When you play sub-1 trades you need to get it right to make money.

A 10+ (ratio) trade on the other hand gives you a higher margin for error, but at the same time is the result of the market ascribing the outcome a low probability. For example if the price is $0.05 the proftit-loss ratio is 20, so in trading for an expected breakeven you could afford to be wrong 19 times out of 20 if the position size is the same on all trades. But again, the probability according to the market is 5% so you need to study up and try and foresee what the market doesn’t.

2. Trading
Another way of making money is utilising the fact that prices are determined in a liquid market with at least some informed participants. So as new information is revealed the price will change. For example if the price of the US fed rate increase contract is like $0.20, and then an inflation report comes out showing a huge upside surprise then the price could rise to, just for example, $0.30. In that case you could close out your position at a $0.10 profit.

You can play these sort of trades quite speculatively around interest rate announcements e.g. if you have a tightening bias you could buy a later contract e.g. April, as a play on any hints for a sooner increase coming from a sooner interest rate announcement. Anyway there are all sorts of usual market trading strategies you can use for actively buying and selling binary options.

3. Arbitrage
Sometimes arbitrage opportunities will open up in suites of contracts or between similar contracts. There will be times when you get quasi-arbitrage from contracts that have very similar payoff profiles, but might be structured slightly differently.

Anyway the easy way to arbitrage between contracts where there is more than one out come and you have mutually exclusive and exhaustive contracts is as follows:
a) Bid side: add up the bids, if the total is greater than $1.00 then sell an equal amount of all of the contracts.
b) Ask side: add up all the asks, if the total is less than $1.00 then buy equal amounts of all contracts in the suite.

Example: a suite of 3 contracts has the “ask” as follows; $0.9005, $0.0890, $0.0075, the total of the “asks” is $0.9970, so you could buy all the contracts for $0.9970 and receive $1 at expiry/event date, thereby netting $0.0030 (or $0.30 for 10 contracts, or $3.00 for 100 contracts, etc – to illustrate).

So there you have it, you can use prediction markets (aka binary options) to get a reading on the views of a market of informed and uninformed participants of certain events happening. But more importantly you can use these markets and instruments to make profits by taking positions on certain outcomes you have knowledge about, or by actively trading in the market, or even through arbitrage.

Article Source:

I have only used this provider: But I am aware that there are other providers out there, and a quick google search should reveal them. Also note this is not to be construed or relied upon as advice. I have no association with ipredict, but I do trade on the platform and from time to time I do take long and short positions in contracts that trade on that platform.

No comments:

Post a Comment

What do you think?