The Big Picture

« The End of M3 Reporting? | Main | Historical M2 vs M3 »

Corp Bonds 1830-2005

Monday, November 14, 2005 | 06:45 AM

My colleague Charles Nenner has been painstakingly analyzing data, some of which goes back hundreds of years. (I will make these available as additional data is complied and analyzed).

Here's the first slice of historical data: Almost 2 centuries of corporate bond yields. The impact of WWII is apparent, as is the subsequent inflation spike (1960/70s) and reversal (1980/90s).

Corp Bonds 1830-2005

Source: Charles Nenner, Cycle Forecaster

This looks like one giant mean reverting system . . .

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c52a953ef00d8345ada4769e2

Listed below are links to weblogs that reference Corp Bonds 1830-2005:

» How to hide your point with a bad presentation of data from Belligerati
The Big Picture has a neat graph on the returns of US corporate bonds for the last 200 years. Unfortunately, interesting as the chart is, the aspect ratio doesn't make any sense. The scale is set to show return over... [Read More]

Tracked on Nov 15, 2005 3:29:34 PM

Tracked on Nov 15, 2005 3:32:21 PM

Comments

It would be sweet if he could overlay that on top of the corresponding inflation rate. I've read that the 1970s and 1980s saw higher rates as the market began to see inflation as more unpredictable and demanded greater premiums to compensate than they had in prior decades.

Posted by: royce | Nov 14, 2005 8:52:39 AM

How one sets the scale has a lot to do with the impression such a chart gives (as you know). If the scale were set to include just the range of returns and not zero plus an extra 100%, the impression of a mean-reverting system might evaporate. Similarly, if the chart showed only returns prior to the 1970s, you'd probably get the impression that returns were trending lower, rather than mean-reverting. A trading scheme based on the notion that returns would revert would have lost a lot of loot in the late 1970s and early 1980s.

Posted by: kharris | Nov 14, 2005 9:30:10 AM

It must be me - I don't see an impact from WWII, I see an impact from the Depression - the dip starts way before the War (early '30's), and the reversion seems to start around the time of the early 50's inflation scare - if there is anything connected to the War, it looks like rates were relatively stable from the late '30's thru the late '40's - or am I missing something....

Posted by: fatbear | Nov 14, 2005 12:38:46 PM

Fatbear -- you are not missing something. In WW II the Fed pegged the bond yield.

Posted by: spencer | Nov 14, 2005 1:13:51 PM

royce gets to my question: are these real returns? because if not, they don't tell us much of anything.

Posted by: howard | Nov 14, 2005 1:54:11 PM

howard, these are not "returns" - these are bond YIELDS.

Posted by: Alex Khenkin | Nov 14, 2005 2:57:00 PM

well, alex, yes, but those "yields" are returns, but ok, i phrased it sloppily (since returns could also imply cap gains on any bonds not held to maturity, but we digress).

are these real yields? the same problem holds.

Posted by: howard | Nov 14, 2005 3:31:24 PM

well, yields are always nominal - the coupon over price. Now, what returns would be after inflation, which is I believe what you're driving at, is a different matter entirely. This chart is very informative as it shows how bond investors responded to different economic environments, and how things can be "different this time" - from 1830s to 1950s bonds were and excellent investment, after which they turned into "certificates of confiscation" for a couple of decades. By the way, this chart pins the height of American economic power at right around 1950...

Posted by: Alex Khenkin | Nov 14, 2005 4:24:55 PM

yes, Alex, that's right: i'm after "real," meaning inflation-adjusted. i personally don't think we learn much if we simply look at the nominal, although you're right that the chart provides some suggestive patterns.

Posted by: howard | Nov 14, 2005 6:58:27 PM