We often think of investments as playing roulette, with “laws” that somehow can be discovered using statistics or machine learning. Yet many investment problems we face actually fall in a completely different category. Firm valuation, determining what to look for when predicting markets – even portfolio construction – are not statistical problems, but complex decision problems that require a very different, methodic approach. The best investors are those who follow disciplined approaches that resonate with the theory of computation. But what about markets? I show that markets should treat these problems as if they were statistical ones, and as a result, should underperform the average investor. Experiments confirm this prediction.