Skew is a concept from statisitcs that is highly relevant to finance, and should be considered alongside kurtosis. Many formulas and equations used in finance assume normal distribution of returns (among other variables), that they fit the bell curve. While this in and of itself may not be a safe assumption to make, an important consideration, assuming normal distribution, is the shape of that distribution. Skewness is one way we describe the shape of normal distributions, and refers specifically to its symmetry. Skewness places a number value on this symmetry, with negative values indicating more data points to the left of the mean, and a positive skew indicating more data points to the right, When veiwed graphically, Negative skewness show distribution curves with a fat left tail, and positive skewness show distribution curves with fat right tails.
Q. A viewer tweeted to us about skew in response to last week’s buzzword, Sigma, and you agree he’s got an important point. What is it?
A. Yes, it is a really crucial point. To see why, we have to take a step back, and recall that most of the statistics we use in analyzing investment performance depends on one key idea: that outcomes are normally distributed, say for example in a symmetrical bell curve. That’s a foundational principle of modern portfolio theory, and is critical in computing things like Sharpe ratios.
Q. Alright, then, where does “skewness” fit in to this?
A. Well, it turns out that in the real world many investments don’t actually generate results that are so evenly distributed, especially for hedge funds and other more sophisticated investments. There might be many more data points on one side of the average than the other. That’s skewness: negative skew if more data points are on the left, positive skew if they’re on the right.
Q. And, for investors, that means what, exactly ?
A. With negative skew, odds are increased of an event that involves a loss. For a positive skew, odds are increased of an event that generates a gain. You won’t be surprised to find out that we therefore like positive skew more than negative skew. Also note that if you see a group of stats on a fund, and there’s a big skew, the more traditional measures might not be so relevant.
Q. So what kinds of investments tend to show skewness, and which show more normal distributions?
A. Well, big picture, you think of normal financial market returns as being pretty equally distributed, so that’s where our friends the Sharpe ratio are most helpful. But a lot of hedge fund strategies show skew, and some alternative strategies show much more skew than others. Two that usually have positive skew are global macro and managed futures. Strategies like that can be very helpful when markets get turbulent…their skew can provide more fat tail risk protection.