If you want to understand science, you have to understand math. In business, if you're enumerate, you're going to be a klutz. The good thing about business is that you don't have to know any higher math.

We've got great flexibility and a certain discipline in terms of not doing some foolish thing just to be active - discipline in avoiding just doing any damn thing just because you can't stand inactivity.

Obviously if you want to get good at something which is competitive, you have to think about it and practice a lot. You have to keep learning because world keeps changing and competitors keep learning. You have to go to bed wiser than you got up. As you try to master what you are trying to do – people who do that almost never fail utterly. Very few have ever failed with that approach. You may rise slowly, but you are sure to rise.

With so much money riding on reported numbers, human nature is to manipulate them. And with so many doing it, you get Serpico effects, where everyone rationalizes that it's okay because everyone else is doing it. It is always thus.

It's a finite and very competitive world. All large aggregations of capital eventually find it hell on earth to grow and thus find a lower rate of return.

We may well have a competitive advantage buying decent businesses at decent prices. But they won't be fabulous businesses and fabulous prices. There's too much competition and money out there, with many buyout specialists.

All the equity investors, in total, will surely bear a performance disadvantage per annum equal to the total croupiers' costs they have jointly elected to bear. This is an inescapable fact of life. And it is also inescapable that exactly half of the investors will get a result below the median result after the croupiers' take, which median result may well be somewhere between unexciting and lousy.