In the April 21 2003 Barron's a letter to the Mailbag section was intriguing and I wanted to bring it to the traders here on ET and see if anyone has any thoughts about the simple market timing model that it suggested. For those that missed it, here is the gist of the system: Compare the yield on the best grade corporate bonds to the yield of the DJIA (both available in Barron's MarketLab pages). When the DJIA yield is 40% or more of the the yield on the best grade bonds>> BUY STOCKS SELL BONDS When the DJIAyield is 30% or less of the yield on the best grade bonds>> BUY BONDS SELL STOCKS I'm trying to get historical data to test this idea. Any leads or thoughts would be appreciated.
surfer, you're right as it is stated right now it is rather simple. But that doesn't mean you can't add your own parameters to narrow it down. Anyway, I think I've traked down a source for the historical data and will in my spare time do some backtesting for various constraints and see what it yields. You never know!
I wouldnt put too much stock into it - no pun intended. For a while the dividend yield vs interest raterelationship was a good proxy, however, for the last two decades, that hasnt really hold much water. That theory might suffer from spurious autocorrelation as well.
Testing the theory is not hard, but it takes considerable amount of time. Data for AAA goes back only to 1983 - if I am not mistaken. Thus you only have 20 years of annual data. Not much to rely on. Remember you have to detrend the DJIA ROR to get a good a proxy. Or else you have the same results as the AFC superbowl market timer. lol. In all seriousness you have to detrend the DJIA. Keep us posted on your results, thanx in advance ---Fish
...in any case, it sounds like this would comprise very long term signals, and not too tradeable in the short term...