« Sorry About That | Main | Mulling the Fed Discount Cut; Damn You Ben Bernanke »
Latest Stories
- Excel Wankers and Recession Averages
- Sorry, New York is Closed. Check Back Later.
- Catching Falling 2009 Earnings Estimate Knife
- Survivorship Bias in Global Markets
- Talking Positions on a Lazy-ish Retirement Portfolio
August 16, 2007
Updated: Records for Largest Single-Day Snapback Rallies
With the Dow having been down as much as 340 points partway through the day today, only to end the dayAnswer: Today's meager 2.5% bounce off the intraday low puts it it in around 4,000th place in terms of snapback rallies on the Dow since WWII. Here are the recordholders, and the market's return over the subsquent 6- and 12-months.
| Date | Bounce | 6-months | 12-months |
| 10/20/1987 | 13.9% | 8.9% | 15.9% |
| 7/24/2002 | 9.4% | 6.2% | 11.8% |
| 5/29/1962 | 9.1% | 5.5% | 18.9% |
| 10/28/1997 | 8.1% | 20.9% | 12.5% |
| 10/9/1974 | 6.6% | 18.4% | 29.9% |
| 10/16/1989 | 6.4% | 2.8% | -11.0% |
| 10/22/1987 | 6.1% | 3.0% | 10.7% |
| 9/1/1998 | 6.1% | 18.9% | 39.4% |
| 1/3/2001 | 5.6% | -3.5% | -6.8% |
| 8/14/2002 | 5.4% | -10.1% | 6.5% |
Only twice did large intraday snapbacks not presage great returns over the next twelve months.
[Update] Two quick updates. First, in the original version of the above post I had the wrong closing figure on the Dow. The right figure doesn't change anything material, but it's fixed above anyway.
Second, a few emailers, and a couple of commenters here, seemed to think I was proposing a snapback trading system. Maybe I should start footnoting things (i.e., * No trading system expressed or implied unless otherwise specified), but, as I say in a comment below, anyone who would contemplate building a statistical trading system from a middling data point in a small sample a week after large sample variants utterly broke down needs their head checked.
Me? I was mostly curious where today's intraday rally fit in historically, and now I know -- as do you.
Sphere It
| 
Digg it
| 
Bookmark it
| 
Stumble it
| 
Facebook it
Note that the two times that intraday snapbacks did *not* lead to great returns over the next twelve months were both at the bottom of the range (#9 and #10) suggesting that perhaps only in cases of extreme snapbacks does this presage great returns...
...so the rally we've just had, which ranks #4000 by your estimation, is probably meaningless when it comes to predicting future market direction.
As it stands, a sample of ten instances is probably not enough to merit true statistical significance anyway.
Perhaps more relevant is the old trader's chestnut: "The strongest rallies take place in bear markets".
Actually the Dow ended down 16, not up 8, but what's a minus sign and factor of two between stock analysts.
"hack" -- Easy big fella. When I wrote the piece (between meetings) Google Finance's realtime quote said the Dow closed up slightly. Later the correct number turned out to be a slight decline on the day. The difference changes diddly in turns of my broader point: This was only a small intraday snapback rally in historical terms. Feel free, of course, to fixate on the narrower issue.
"speedlet" -- Wrt statistical significance, this is purely expository, not a trading system. I was simply intrigued by where today's rally fit in historical terms, and the answer was nowhere. Anyone who would build a trading system from a middling data point in a small-sample example a week after we watched large-sample statistical stock distributions break down needs their head checked.
Too bad that word of the Fed's rate cut looks like it leaked out yesterday and sent the Dow up on that "snapback" - otherwise, that 325 gain combined with today's (at one point) 300 point surge would have presented a more attention-grabbing 600 point gain this morning. The FT had some speculation on today's Fed action in yesterday's print version, but I don't think anyone saw it coming so swiftly - ya gotta love Big Ben's decisive action in a seemingly hopeless and futile situation, which reminds of the great Vince Young in the Rose Bowl one year and seven short months ago...
shouldn't this be put in the context of the decline that preceded said snap back? the reason this one is meager is because the decline has been meager.
Typeset your blog post in Impact Bold, put in some exclamation markets, and you could make a fortune selling a new trading system via direct mail.
SECRETS TO RICHES!!!! Our Snapback Signal proven to produce 10,000% profits!!!