The Physics of the Dead Cat Bounce

Includes: DIA, SPY
by: Andrew Butter

On March 14, 2009, Professor Nouriel Roubini wrote a piece: "Reflections on the latest dead cat bounce or bear market sucker's rally."

Well he could be right, certainly the rally looked busted on Friday and it tanked Monday. Plus Tom Dryden pointed out that those bank profits were thanks to a scam; someone called it a "fraud bounce" (but aren't bank profits always thanks to some sort of scam?).

On top of that, where we are with Secretary Geither's plan, which supposedly created a surge of optimism based on the idea that there WAS a plan, is starting to look less good now that people get around to reading it (and working out how long it will take to implement). And still no one can agree the value of the toxic assets, and more important the CDS "problem" still didn't go away (not much different from last September), although the good news is that the CDS at Lehman got cleaned up quite painlessly (that's one good argument for liquidation rather than the miss-mash of semi public-semi-private).

Two arguments against that thesis (there are others);

  1. The stock market is now valued exactly the inverse of how much it was over-valued in 2000 (but that's old news).
  2. I did a bit of investigation about dead cats bouncing.



Definitions: A dead-cat-bounce is when the next low after a rally is lower than the preceding low (note mini-rallies less than 25% of the previous drop that then drop more than 10% are not considered dead cat bounces, those are defined as dead mouse bounces).

Thesis to be investigated:

  1. Can the height that a dead cat bounces be reliably predicted from the height you drop it from?
  2. Can you work out if a cat is dead from how high it bounces as a function of the height you dropped it from?

Materials used for experiment:

Seven dead cats (representing the most elastic dead cats observed in the period 1909 to 2009 on the DJIA (from one hundred years and 26,234 points of data).


"Bounce" was compared to the previous drop.


  1. On average, 91.35% of the "bounce" could be explained by the height from which the cat was dropped.
  2. This is consistent with Newton's Laws of Physics (and the General Theory of Relativity...).
  3. The purple dot on the picture representing the "bounce" from 9th March is 4.2 Standard Deviations away from the best-fit line.
  4. In other words the probability that the cat observed bouncing is a dead cat is one in 16,966. That is a much lower probability than a AAA bond rated in 2006 by either Moody's, Fitch or S&P, will default; in fact about 3,000 times less likely.
  5. Unless of course it's a "Black Swan" (there have been a lot of those flying around these days, and if you can claim to have seen one of them, well you can say black is white). That's extremely unlikely given the number of Black Swans that have allegedly been sighted (a Black Swan is the explanation used by the rating agencies to explain their little faux pas "not my fault 'Gov, it were a Black Swan" - last time I tried that line the Policeman breathalysed me).


The rally was not caused by either the janitor or the chimpanzee who are all that's left at AIG selling plutonium and anything else not screwed down out the back door, as was so graphically explained in this article (my money is on the chimp).

So I'm sticking to 6,547 DJIA or thereabouts as THE bottom, which is more or less what I said in January (then I was branded a pessimist, now I'm supposedly part of the "chorus of optimists" - can't win).

Not that I have any hope that there will be a "Raging Bull" for quite some time. One way to know when that's imminent is that I won't be whiling away my time writing silly articles about dead cats.

Stock position: None.

About this article:

Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here