1983-2003 Backing the losers would have been a win

I have had to run some figures for a uni tutorial on different asset classes and two opposite strategies.

Strategy 1
Invest the portfolio for the next year in the best-performing class of the current year

Strategy 2
Invest the portfolio for the next year in the worst-performing class of the current year

The result: the worst-performing strategy had an average return of 2.29% more and its std deviation on returns was lower (hence, lower risk I guess).

That brings me to the question about considering risk in quantitative terms. Does anyone here actually rationalise investment decisions based on risk calculations? I started a thread the other day about leverage vs. nominal return to gauge the importance of each in different combinations. But once you add risk into it, the greatest performer can be the one with the lowest nominal return.

Then of course, risk measurements (beta, Jenson alpha, Sharpe Index, etc) aren't exactly perfect. You are always only measuring past risk and whether you take that as a guide to future risk is up to you.

To be honest, I couldn't see myself at the moment choosing a much lower return investment if considering the alpha (for example) made it out to be a slightly better performer. Maybe once I become more confident with the risk theories I'll see it differently.

Anyway, just sharing a few thoughts