From Chapter 1 of The Alpha Masters by Maneet Ahuja:
Dalio estimates that if the firm can make money on 60 to 65 percent of its bets in any given year, the odds are very high that the fund will meet its return targets. In 2010, as the D-process continued to unfold, about 80 percent of Dalio’s bets made money.
Bridgewater sees endless opportunities to do this because spreads are uncorrelated. In doing this, the most important rule is not to compare the correlations against each other in a quantitative sense, but according to their drivers.
“But the truth is, as you get to 15 or 20 you start to reach diminishing returns,” says Dalio. “So the issue is, ‘Do you really know what you’re doing? Can you be confident it’s good?’ I used the word good. I didn’t use excellent. Can I be confident it’s good?”
Dalio thinks the best mix of assets is an amalgam of things, and advises to derive your top alpha generators from a combination of currencies, bonds, commodities, stocks, and so on, and calibrate them properly against each other, in terms of their size.