Hi folks, Why do we Spread Trade? Is it really less risky than trading outright positions? Due to the lower volatility of hedged positions, and thanks to the reduced margin requirements, we end up trading correspondingly larger positions, which leaves us back where we started, risk-wise. Is it because spreads trend better (that is, stronger, longer, and with less variance) than outright positions? Isn't that the main claim of spread traders? Assuming it is, the question is whether that perception is valid. What do you think? Thanks, Steven
Some of the reasons you mentioned are indeed possible benefits of spread trading. Another attractive aspect of spread trading is that you are not betting on an absolute move in an underlying contract, but rather a relative move between two separate contracts. This allows for much flexibility. For example, say you believe X is undervalued compared to Y. You will make money if any of the following scenarios play out: 1. X increases more than Y. 2. X decreases less than Y. 3. X increases and Y decreases. However, it is important to realize that there is not necessarily less risk in spread trading versus outright trading. It is quite possible that the contract you are long decreases in value and the contract you are short increases in value. In this instance, the trade will lose you more money than if you were simply wrong on the direction of one outright contract.
Another point is the idea of correlation.... The logic of a spread assumes that there are two different price items that are highly correlated.... Such that if the differences in price were to seem abnormally high.....and if correlation was seemingly perfect....then the spread concept would work reliably.... The key point being "correlation change".....
It sounds like you're describing the mean-reversion type of spread. That's certainly well-thought of as a strategy, but for exactly the reason you point out--correlation change--it freaks me out to attempt to trade in this way. Seems a bit more comfortable to participate in divergence from correlation, rather than a return to it. Am I to understand that it's far more typical to play the mean reversion spread than the type that capitalizes on trends of divergence between the instruments?
it seems as if most pair traders on ET use mean reversion to get into and out of correlated and more importantly, highly cointegrated pairs. on risk management, the key is being able to identify if a pair has changed from its former cointegration to a new trending pattern and act accordingly (which is to say, GTF out). this happened with a lot of the financial pairs starting in summer of '07, with oil related pairs last summer or always, any one company getting a marked edge and taking market share from the other. IMO, the problem with looking to get into a trending pair is that if it is a somewhat correlated pair, then who is to say when it will start to come back in, which as in your original post, you may be better off just choosing a direction of one of the stocks. another option if you are more keen on pair divergence, if you have a knack for fundamental analysis and finding the under and overvalued stocks or which sectors will thrive and/or tank in a particular period of the economic cycle, then maybe just a paired portfolio of a variety of stocks or baskets of stocks with your capital balanced long and short could work. with that, you are at least theoretically mitigating systematic risk.