A correlation swap is a particularly complicated form of financial derivative that is not based directly on the price of an underlying asset. Instead it is based on the relationship between the prices of two or more assets. Because of this complexity, a correlation swap must be arranged privately and is not available through mainstream financial exchanges.
The basic form of derivative is comparatively simple. The derivative is an asset in itself, but it derives its value from a separate underlying asset. A simple example is a futures contract, in which one party agrees to buy a set amount of a stock at a set price on a set future date from the second party. This may prove a good or bad deal depending on the market price of the stock on the agreed completion date: if the market price is higher, the buyer of the stock can immediately sell at a profit. Because a futures contract is an asset in itself, the buying party can sell on the rights to complete the deal before it comes due. This is known as selling a position.
A swap derivative goes one step further as it is based on two or more underlying assets, one from each party in the agreement. It involves the two sides agreeing to swap the revenue from the respective assets. For example, in a bond swap, the two sides each own a bond, but agree to swap any coupon payments they receive from their bond. In effect, the two sides swap the risk involved in their own asset, for example, the risk that a bond issuer may not pay the expected coupon payment. Such deals can be done purely as speculation, or they may be used to mitigate risk, a tactic known as hedging.
The correlation swap is based on the correlation between two assets on a future date, not the price. For example, the first party in the deal may predict that the stock price of company A may be twice the stock price of company B in three months, and pay a flat amount to the second party. In return, in three months, the second party will pay a variable amount that depends on the actual correlation. For example, if the stock price of company A turns out to be three times the stock price of company B on that date, the second party may have to pay a larger amount back to the first party.
The correlation swap process is relatively complicated as those involved not only have to predict how each price will change, but also the comparative changes of the two. In turn, this makes it much more difficult to figure out a fair price for buying or selling a position in the deal. At the moment, there aren't any formulas that are widely accepted as giving an accurate and fair valuation in a way that cuts out the possibility of arbitrage. This is where a trader can exploit differences in pricing between two deals, such as two correlation swap agreements, to greatly increase the likelihood of making a profit, or even make doing so theoretically certain.