CI Staff .


John Desantis from CA posted over 3 years ago:

The problem with both the Sharpe and the Treynor ratios is that they treat upside price movements the same as downside price movements. In my opinion, the denominator of the ratios should not increase due to upward price movements.

The Ulcer Index has the same numerator as Treynor & Sharp; but the denomonator includes only the amount of the price movement that is negative.

The concept is described in a book by Martin & McCann called "The Investor's Guide to Fidelity Funds".

Rene Barrios from FL posted over 3 years ago:

helpful for a new subscriber, thanks

Wayne Thorp from IL posted over 3 years ago:


I agree with you. Treating upside risk and downside risk is one flaw of traditional risk measures. In the Q2 issue of CI we will be discussing the Sortino Ratio, which focuses on downside risk and in a future Spreadsheet Corner I am going to apply the Ulcer Index to some of the more successful AAII stock screens.


Glad you enjoyed the article. Welcome to AAII & CI!

Wayne A. Thorp, CFA
Editor, Computerized Investing.

Ronald Longhofer from MI posted over 3 years ago:

A question about the use of the Treynor ratio for investment decisions. You write:
"Looking solely at return figures makes the choice very clear: One should invest in Wynn Resorts. However, the Treynor ratio paints a different story. . . . The Treynor ratio actually points to Pfizer generating a better risk-adjusted return."

I have calculated the ratio for 32 stocks I own or am following, and find that stocks that I, as a value investor, would consider buying (because they appear to be undervalued) tend to have lower Treynor ratios than the median for the market as a whole (which I calculated both for a universe of 1170 stocks screened for positive earnings and other measures such as positive book value, and separately for the components of the S&P 500). This is no doubt because, being undervalued, the stocks I follow do not have exceptional recent price performance. Conversely, stocks that have performed well in the recent past tend to be fully valued, and are therefore not necessarily good candidates for future returns, although they may have high Treynor ratios based on their historical performance.

I wonder whether the historical Treynor ratio isn't a better after-the-fact performance measure than it is a stock-picking tool? At least for value investors who want a margin of safety. You write that the Treynor ratio measures "how well an investment vehicle compensates the investor for a given level of risk." At least for the historical version of the ratio, wouldn't it be better to say "has compensated"?

Now, the "expected return" version would be a different story. I hope you are planning to do an article showing how this version might be implemented, as I, for one, would find it very useful in comparing potential investments.

Sorry, you cannot add comments while on a mobile device or while printing.