Not Volatility

KYUR uses advanced mathematical algorithms and quantitative screens to pick investments and construct portfolios customized to your personal risk tolerances, to limit the risks of potential losses to your capital.

Just like for the wealthy and institutions, we quantitatively measure your risk tolerances by the probabilities and amounts of possible losses to your capital —  not just the ups and downs of investments as used in Mean-Variance Optimization (MVO) of Modern Portfolio Theory (MPT) or guesses like “conservative” or “aggressive” as in traditional ad hoc approaches.

What’s the Difference?

Stocks are risky investments; they go up and they go down. Some investors view rallies as time to sell and down days as opportunities to buy. Others take the opposite views.

KYUR philosophy is, what matters is the amounts of wealth you end up having for education, buying your dream home, retirement, or preservation of your life styles.

Do you know that during the period between August 1982 and December 1999, the Dow Jones Industrial Average generated total gains of 1,409 percent – 14 times of your investments?

In comparison, between January 2000 and February 2009, if you had invested in the Dow, you would have lost -44 percent, almost half of your capital? One would think the bear market of 2000 to 2009 was far riskier than the almost two-decade long bull market of 1982 to 1999 –as graphically shown nearby.

But if one uses volatility to measure risks as in Mean Variance Optimization, the two periods would be deemed equally risky. Both periods have the same annualized monthly volatility of 15.2 percent. For explanations of volatility as the measure of risk in Modern Portfolio Theory, see writings by the father of MPT and Nobel Laureate Harry M. Markowitz , “Portfolio Selection,” The Journal of Finance, March 1952, and his Portfolio Selection: Efficient Diversification of Investments, John Wiley & Sons, (reprinted by Yale University Press, 1970.)

Screen Shot 2017-04-10 at 3.00.13 PM

How come?

It has been well recognized that volatility is a poor measure of risks. Volatility (or standard deviation, the square root of variance) is the measure of risks in MVO. Volatility cannot effectively differentiate the risks of bad from good investments. Thus, MVO may result in unsuitable advice to investors and expose them to potentially large losses. For a modern, up-todate understanding of volatility, and downside risk, see, for example, Pablo Triana, The Number that Killed Us, John Wiley 2011, available on

In the final analysis, how much does it matter if stocks fluctuated up and down by not very much, if you ended up losing almost half of your capital?