Here are his recommendations.

> First, eliminate forward-looking earnings guidance. This practice forces companies to make public commitments about future performance, creating enormous pressure to meet those predictions regardless of changing circumstances.

Honest question - are companies forced to make forward looking projections? I would have assumed the whole reason they have a legal safe harbor for making forward looking statements that turn out to be wrong is because companies WANT to be able to make these projections as an IR exercise, but maybe I’m wrong.

> Second, create accounting treatments that allow companies to separate long-term innovation investments from operational expenses, giving investors clearer visibility into both current performance and future potential.

I am fairly sure that every accounting rule that exists is because someone has abused the numbers, but OK, let’s give companies more freedom. What could go wrong. (And wasn’t capitalizing opex basically what brought down WorldCom?)

> Third, develop new metrics and incentives that reward patient capital deployment and long-term value creation, not just quarterly financial performance.

Ah yes, innovative metrics like McDonnell-Douglas’s “return on net assets.” Last I heard, they had taken over their biggest competitor. Mission accomplished!

Wasn’t this the guy who wrote some ridiculous piece about how HP buying Palm was a terrific idea but it all fell apart because he was out of the office or something?