Volatility or Annualized Returns - What's More Important
- Thread starter CPTrader
- Start date
Once again, I think it depends on your personality. If you are not a risk taker, big volatility might lead you to give up on the fund at the WRONG time, just when the system is about to recover from a low. Can you stomach the volatility is the question. If not, then a more prudent approach is appropiate.
My spreadsheet divides leverage with drawdown, to tell me if I am ontrack...all things must be equal and the smooth upward sloping NAV equity curve is desired. The rate of slope is nice, but not as important as being smooth...
The problem is that academic analysis has equated volatility with risk. Perhaps that is valid, but many investors are quite happy with upside volatility. It seems to me that any trend following or volatility breakout strategy would tend to generate spikes of upward volatility, while downside volatility should be fairly even. Maybe I'm wrong.
Clearly for most of us, return:drawdown is the most important ratio. I guess the problem arises in that the manager may be lucky for a few years and dodge a big drawdown, even as he generates a positive but erratic return. So Sharpe is used as kind of a canary in the coal mine to warn of the potential for disaster.
Is there academic research that shows a correlation between Sharpe ratios and subsequent performance? Or blowups?
Clearly for most of us, return:drawdown is the most important ratio. I guess the problem arises in that the manager may be lucky for a few years and dodge a big drawdown, even as he generates a positive but erratic return. So Sharpe is used as kind of a canary in the coal mine to warn of the potential for disaster.
Is there academic research that shows a correlation between Sharpe ratios and subsequent performance? Or blowups?
My impression is that Sharpe Ratios are a cousin of standard null hypothesis testing, the boring kind you learn in Statistics class and you find in most academic papers.
The central limit thereom tells us that no matter what the underlying distribution, the distribution of sums of repeated independant trials will approximate the normal bell curve.
So one can say Sharpe Ratios are trying to tell you how likely a manager's returns are due to chance, assuming a null hypothesis of returning the risk free rate, which presumably any one can do.
Of course there is no perfect measure of risk/return, and the Sharpe has its weaknesses. Actually Bill Sharpe has probably written the best stuff on the topic. Not for nothing did they give him the Nobel Prizein 1990!
The central limit thereom tells us that no matter what the underlying distribution, the distribution of sums of repeated independant trials will approximate the normal bell curve.
So one can say Sharpe Ratios are trying to tell you how likely a manager's returns are due to chance, assuming a null hypothesis of returning the risk free rate, which presumably any one can do.
Of course there is no perfect measure of risk/return, and the Sharpe has its weaknesses. Actually Bill Sharpe has probably written the best stuff on the topic. Not for nothing did they give him the Nobel Prizein 1990!