Wednesday, March 25, 2009 0 comments ++[ CLICK TO COMMENT ]++

A common flaw with charts: not going back far enough

One of the things one has to be really careful about is, avoiding the comparison of the near-future to the last 30 years. I am of the view that the near-future will be more like the 30's to 50's (in an investing sense) than the 80's to 2000's. I see a lot of people, including the mainstream media, using charts that only go back to the early 80's and I always shake my head. Although it is worth looking at the last 30 years, it is quite possible that the future will be nothing like that period.

Consider the following chart from The Globe & Mail purporting to show that junk bonds outperform stocks for a few years after a recession:

Conventional wisdom seems to hold that bonds outperform stocks in the immediate period after a recession. This may very well be true; but it's dangerous to form that opinion from that chart above. On top of only having 3 data points, one also needs to keep in mind that bonds have been in a bull market since the early 80's (yields have continuously fallen.) It's true that junk bonds, which is what is plotted, behaves more like equity than bonds but nevertheless, the fact that yields were declining would have boosted the junk bond. I'm just speculating here but it's not clear if such a marked outperformance will exist if yields were rising (this would be bearish for bonds.)

Even if the conclusion is valid, I would be more confident if the data went all the way back to the 1920's (modern junk bonds didn't exist but you can probably use defaulted railroad bonds and the like.)


No Response to "A common flaw with charts: not going back far enough"

Post a Comment