I'm not sure what Sharpe ratio signifies, but it seems to be some kind of risk-adjusted return. I don't see why the "excess return" should be divided by the standard dev--would the results be different if it divided by, SD squared?Anyway, it's interesting that the short-term seems to show that there is little "risk". Most of the data is from a period where there was not much short-term volatility, so that skews the results. I wonder if they did a similar study on the last 5 years, what would the curve look like.I'm not surprised to see the graph take that sharp upturn near the end - longer holding periods tend to mitigate final-value risk.However, my usual complaint applies: this assumes one has a lump sum and buys the portfolio all at once, then holds until the end. Some investors are going to be buying continually during the period, perhaps increasing their purchases as time goes by.So, sharpe ratio is a good measure of risk for some investors, but not others.
I just found the websitehttp://www.stanford.edu/~wfsharpe/art/sr/sr.htmwhere Sharpe talks about his ratio.He says that it gets complicated to compare sharpe ratios over periods of differing lengths. I'd like to know how these guys handled these issues (compounding conventions, etc.)
well you can drop out the risk free rate and you'll get a similar set of data points over the increasing holding periods.the point is that the data clearly shows, that despite the seeming unpredictability of short term price moves, that variance of those moves does not start out extremely high and then move progressively lower with time. in fact, variance increases with time until a holding period of 2000-3000 days, and then begins to drop off significantly (resulting in a "tighter" statistical disn). variance is highest at 100-1000 days, not 100-1000 hours. so if you define risk as price volatility, you should be aware of the fact that volatility, as defined by the price variance, is highest in that 100-2000 day time frame. most people do not understand this.it is particularly relevant for people who expect to be cashing out of investments in that 100-2000 day time frame, and for retirees who are withdrawing capital over time with a set asset alocation method such as the tri-partite method of Armstrong.tr
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |