Efficient Markets: Scott Sumner vs. Keynes

     I've started reading Skidelsky's 2009 book: "Keynes: The Return of the Master." Obviously it was written just as the crisis was at its worst. Skidelsky-who's the definitive Keynes biographer-argues that Keynes' economics and his own personal history as a market participant are closely related.

     Keynes scandalized economics-and continues to scandalize it-with his repudiation of rational expectations and efficient markets. What's interesting about Sumner is with all the crying about how Krugman rejects conservative economists out of hand and doesn't give them credit where it's due, Sumner certainly has nothing charitable to say about Keynes.

    I think with all the differences between Keynes and today's New Classicals, nowhere is it more striking than regarding the market itself. I seem to recall Lars Christensen saying that he and Sumner don't play the market because they respect it's efficiency and reason you can't expect to beat the market consistently over time.

     Keynes, however, was one of the few economists who did play the market heavily and won regularly. I still remember when Mankiw wrote that rather silly piece "debunking" Keynes by talking about how he lost money during the 1920 slump and how since he got back in with financial help from his father, his market success wasn't real: anyone can make it in the market if someone else gives you the money to play.

     This makes little sense and even contradicts the idea that market efficiency means you can't beat it over time. Mankiw's catty comments would mean that you can beat it over time as long as you have a deep pocketed benefactor.

     I notice that Sumner tends to dismiss people like John Paulson and George Soros. We go from the extreme to worshipping success to completely dismissing it. For Sumner getting something right in the market doesn't matter because it was only luck anyway.

    I wonder if this is a defense mechanism on the part of Neoclassical economics. As it's predictions are so poor, simply argue that prediction isn't important anyway-mostly it's just dumb luck.