Wonk Warning – the following post may be boring and probably appeals to an audience of maybe two, if that. But I needed to get it out there. But don’t leave just yet – the take away here is that the on-peak WESM daily average price to-date has been about 15-18 times more volatile than California.
I tend to use a short-hand, informal measure of volatility – but that’s probably misleading because you can’t really compare that to published volatility measures of other commodities. So I’m going to convert over and use a bit more of a conventional measure. Although, frankly, there is no “standard.” Volatility measurement procedures tend to be a bit loosey-goosey and not consistent.
Conventionally, volatility is an annualized measure of the standard deviation in the geometric percent change in the value of a quantity from one period to the next (e.g. the day to day price movement). The geometric percent change is measured as the natural log of, say, the ratio of one day’s price to the previous day’s price (or hour or month or whatever period you’re using).
The annualization process is just an estimate based on random walk theory. I won’t go into the math here, but by convention this annualization is done by measuring the standard deviation in ln’s over some period (say two weeks, or three months, or whatever). Then multiplying that standard deviation by the square root of the number of number of such trading intervals in a year. So if your measuring the StDev in ln’s of daily WESM prices, you would annualize by multiplying by the square root of 365.
So if I measure the ln’s over two weeks and annualize that, I will most likely get a different volatility measure than if I measured the ln’s over three months. See? Loosey-goosey.
My measures here and here were not annualized. As a shorthand measure, I tend to look at the StDev of the LNs over some period. But if I annualize the 66% number (based on 6-day weeks since I eliminated Sunday), I get a volatility of 1,166%. I looked at an arbitrary three month period of on-peak prices in Southern California for 2005 and calculated an annualized volatility measure of about 65% or so. (this paragraph has been updated).
So that leads me to the statement that WESM prices have been 15-18 times more volatile than California.