jpdoctor: it makes no sense to judge his long-term results based on a short stretch of under-performance. According to that same paper, the portfolio he managed for Cambridge University beat the market by 8%/year on average from the early 1920's until his death in 1946. That period includes the Great Crash of 1929 and the Great Depression of the 1930's!

PS. That would be like judging Warren Buffett's impressive long-term investment record based on the sharp decline in Berkshire Hathaway's share price during the financial panic of 2008!

PPS. If you think Buffett's record is explained by luck, check out the subsequent long-term record of the seven investors identified in this article he wrote for Columbia Business School's magazine in 1984: http://www4.gsb.columbia.edu/null?&exclusive=filemgr.dow... -- all materially outperformed the major stock indices after being identified by Buffett -- what are the odds of that?

> it makes no sense to judge his long-term results based on a short stretch of under-performance.

It makes enormous sense: He missed the biggest economic event of three generations. If you can't call the big ones, then there's very little proof his returns were something other than luck.

Look, both Keynes and Buffett (since you mentioned him) have long term positive results when there was a long term positive market. Choose a high-beta portfolio in such an environment, and voila! you beat the market. In order to show you have any actual insight and aren't just playing for luck, you need to show that you beat the market significantly in both up and down environments.

BTW, the entire mutual fund market is based on this approach, and when a fund seriously underperforms, it is conveniently removed from the portfolio of Fidelity/Janus/insert hucksters here.

Edit: Relevant http://www.marketwatch.com/story/americans-dont-understand-i...

Most people out there don't understand the inverse relationship of bond prices and yields, let alone beta.