Arbitrage is an important aspect of trading. If you are a stock market investor or trader, you certainly have come across the term often; essentially it means the buying and selling of a particular asset or its derivative simultaneously in different markets. When arbitrage happens, the pricing difference of an asset in one market and the same (or its derivative) in another is exploited for gains. 

There are different types of arbitrages, from cash and carry to reverse cash and carry and statistical arbitrage. Also called stat arb, it is a term that defines a set of trading strategies where mathematical modeling is used to determine price differences between securities. The strategy makes use of the concept of short-term mean reversion. Statistical arbitrage is also bracketed under a set of algo trading strategies, where trades are executed on the basis of the algorithm that is preset.

If stat arb is employed, then price movement across several securities are tapped into after an analysis of price differences and patterns between these instruments. 

Stat arb is used by hedge funds and investment banks as well, as an effective strategy. 

What is short-term mean reversion and its relevance in stat arb?

This is a technique wherein buying occurs after prices have dropped below their average and leaving once they get back to normal levels. In a short-term mean reversion technique, these positions are held only for some days or weeks. This is the opposite of value investing where it is held on to for years. The principle where the price differences are expected to see reversion to the mean over the short term is at the core of this technique. The time leading up to this reversion is exploited for making gains.

The short-term nature of this model is employed in stat arb strategies, where hundreds of securities can be invested in for a highly shortened period of time, from a few minutes to some days. 

Types of statistical arbitrage strategies

There are many strategies that are bracketed under stat arb trading. Some of them are:

  • Market neutral arbitrage: This strategy is about going long on an asset that’s undervalued and taking a short position on an asset that’s overvalued at the same time. The long position is expected to go up in value while the short continues to drop, and the increase and decrease are at the same levels.
  • Cross asset arbitrage: This model taps into the price difference between an asset and its underlying. 
  • Cross market arbitrage: This model exploits the difference between the same asset across markets.
  • ETF arbitrage: This is also a cross asset arbitrage technique wherein the differences between an ETF’s value and the assets underlying are spotted. This is employed to ensure that an ETF’s price is in line with the price of the assets underlying. 

What is pairs trading and how is it different from statistical arbitrage?

Pairs trading is often used as a synonym for stat arb. However, statistical arbitrage is more complex than pairs trading. The latter is a simpler strategy and is one of the statistical arbitrage strategies. Pairs trading is a market-neutral strategy wherein stocks are bunched into pairs. It means two socks with similar price movements are found, and when the correlation comes down, a long position and a short position are taken on the two. The gap between the two is tapped into, till such a time that the two go back to their original or normalised level. 

Typically, traders look to pair stocks that belong to the same industry or sector because they tend to have strong correlation. 

Stat arb trading does not involve pairs and instead considers several hundreds of stocks, making up a portfolio. 

Not without risks

Statistical arbitrage plays a key role in ascertaining everyday liquidity and stability in the market. Also, traders benefit from such a strategy. However, it helps to remember that sometimes it also comes with a risk. One of them is that the mean reversion may not occur in some cases and prices may vary hugely from the normal level, as shown historically. The markets are constantly changing and evolving and sometimes don’t behave as it has in the past. This risk needs to be borne in mind while using statistical arbitrage strategies.

Conclusion

Statistical arbitrage is a strategy that uses extensive data and mathematical/algorithmic modeling to take advantage of price differences among securities. It relies on short-term mean reversion, wherein the price differences up to the point of reversion to mean levels are taken advantage of.