As an investor my job is to figure out what will happen rather than what should happen.

One of the best investors around, Joel Greenblatt, has written a popular, charming and funny book about investing in great companies at low P/E multiples. To simplify an already simple book, great companies are generally measured as companies that can generate lots of profit without requiring a lot of capital. This means that they have high ROEs.

Microsoft could help Facebook with one of the biggest challenges, namely monetizing its traffic without reducing the user's experience. It's obvious that Microsoft needs traffic and Facebook needs search.

What is uncertain is how much further the bubble can expand, and what might pop it.

You have a certain set of facts and you are looking for situations where you have an edge, whether the edge is psychological or statistical.

Microsoft has one more shot at a role in smart phone software through its deployment on Nokia phones. Nokia is still the global market share leader in cell phones. Maybe it will work out, but this is hard to envision great success in the area coming on the heels of so much disappointment in missed opportunity in this important and visible category.

When you leave a good job to go off on your own and don't expect to make money for a while, you name the firm whatever your wife says you should.

Ive got a fantasy-baseball team with my brother. But I have to admit, he does all the work.

I will keep a substantial long exposure to gold -- which serves as a Jelly Donut antidote for my portfolio. While I'd love for our leaders to adopt sensible policies that would reduce the tail risks so that I could sell our gold, one nice thing about gold is that it doesn't even have quarterly conference calls.

In the real world, illiquid assets carry a discount.

What do you call a stock that's down 90%? A stock that was down 80% and then got cut in half.