So I get this email and it provides a link to a paper and one of the authors is Sornette and I somehow recall that name and ...

>You recall something? Now that's a first, eh?
Pay attention!
It deals with market crashes and what to look for to anticipate them and ...

>You're kidding. Predicting crashes? I doubt if ...
... and I recall playing with this a while back, here, so I thought I'd play again and ...

>And now you're predicting a crash?
Huh? Me? I'd never do that! However, the spreadsheet does this:

You enter a stock symbol, click a button to download a bunch of recent prices, apply the magic analysis given here.
Then click another button to run thru' a bunch of parameters to see if, for some set of values, it predicts a crash.
Of course, you get to define what you mean by a "crash" ... like a drop of, say, 20%.
>And it predicts a crash?
Yup!
>And I click on the picture to download the spreadsheet?
Yup!
>And you guarantee the validity of the prediction?
Nope!