Ahhh...I see where SnoopRIP is coming from. The recent reverential talk of Sharpe ratios and Nobel prizes in Economics and the comparisons of the market to dice and coin-flipping suggests that he is an Efficient Market Theorist. The Efficient Market Theory [EMT] is basically that no one can ever beat the market, because news travels quickly and all information about a company is already priced into the stock. Furthermore, future price changes depend on presently unknown facts (those that will not be known until some time in the future), and since those facts are unknown and unknowable, no one can predict the market. Stocks will therefore follow a random walk, and the only way to play is to index. "Capital Ideas" by Peter Bernstein is a very interesting book about the history of ideas about the stock market, and is accessible to non-experts.First a joke. Two economists walking across campus see a $20 bill on the ground. As one of them bends to pick it up, the other says "Don't bother. If it were really worth $20, it wouldn't be there." Then a suggestion. Read "Buffett" by Roger Lowenstein, "The Warren Buffett Portfolio" by Robert Hagstrom, and some of the letters to Berkshire Hathaway shareholders. In 1988 Buffett wrote: "Observing correctly that the market was frequently efficient, they [EMT theorists] went on to conclude that it was always efficient. The difference between these propositions is night and day." [italics are in the original]As I said before though, the problem I have with SnoopRIP is that he has already decided what he wants to prove, which is that the FF can't work. Why is he so sure? Because under EMT, no stock picking strategy that relies on easily knowable data can ever work unless it is undiscovered. After it is discovered it will always fail, because the Efficient Market will include the discovery in its pricing. Then, he designs experiments around his conclusion, dismisses ideas that would test the conclusion (e.g. insisting on out-of-sample tests, and then insisting that the BSP30 is not an out-of-sample test), uses obfuscation as a debating tool (obvious to anyone who tries to read through the threads or sees the frustrations of others trying to ask for clarification over and over again), uses hyperbole (suggesting that TMF has suppressed the history of the FF and implying a coverup), or overinterprets statistics (failure to reject a hypothesis does not prove your favorite alternative). You start with a hypothesis, then do an experiment and either reject or fail to reject the hypothesis. That's scientific method. When you start with the conclusion, do an experiment, and then accept or reject the experiment in order to keep your conclusion, that's pseudoscience.None of this proves that the FF works, nor does it prove that EMT is wrong. But I think it may explain why SnoopRIP is so adamant. EMT has as much, if not more, to lose if the FF works than TMF has to lose if the FF doesn't work. If the FF works well after its discovery, then EMT is wrong. That would explain why SnoopRIP refuses to accept BSP30 as out-of-sample - the only out-of-sample test he cares about is after the discovery of the strategy.However, the data mining issue is addressed VERY well by the BSP30. To test the hypothesis that the RP4 method is just data-mining with no more validity than picking stocks starting with D, an ideal experiment would be to construct an alternative group of stocks with similar characteristics to the Dow. Ethan's strategy of sorting the S&P500 by market cap, discarding utilities and transportations, and picking no more than 3 from any industry group as defined by Business Week, is as good a way to find an outgroup as you could imagine. This list would include many of the Dow stocks if he hadn't removed them first. That step of removing the Dow stocks when constructing the BSP30 may actually hurt the stock-picking value of the BSP list, but it makes it an ideal out-of-sample test for the FF, because NONE of the stocks are the same. I think SnoopRIP conceded that the BSP30 did beat the market in the past. Maybe he will accept the hypothesis that dividend-based strategies worked before their "discovery" in 1988. But I suspect that hypothesis is largely irrelevant to him. I predict that where he will really fight hard is the issue of whether the strategies worked after 1988-90 (whenever his cutoff date is), for the abovementioned EMT reasons. The problem, as others have pointed out, is that there aren't enough data points yet to say anything about it. I also predict that he will reject the argument that Graham emphasized the importance of stocks with high yields beginning in the 30s, so the FF and BSP30 results even from before 1988 suggest that dividend-based strategies can be successful in spite of their being based on easily knowable facts. Next he will use the Sharpe ratio to argue that even if the strategy works, it is unacceptably risky. That's the next part of EMT. The conclusion will always be the same: you can't beat the market, and if you beat the market it's because you took unreasonable risks and were lucky.An interesting corrolary to EMT is that strategies may exist that beat the market without increased risk, but they will stop beating the market after they are discovered - the Sharpe ratio will revert to the mean. However, if the majority of investors don't believe a strategy will work, that ought to be just as good as having it remain undiscovered, it won't be arbitraged out of existence. So the publicity and hyperbole the Gardners have used ("Crush your mutual funds..." makes me cringe) is actually likely to prevent the future success of the FF, while the BYU study, Zweig's article in Money, and SnoopRIPs posts are likely to enhance its success. I like the irony.
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |
BATS data provided in real-time. NYSE, NASDAQ and NYSEMKT data delayed 15 minutes.
Real-Time prices provided by BATS. Market data provided by Interactive Data.
Company fundamental data provided by Morningstar. Earnings Estimates, Analyst Rat