The Hidden Power Of Back To Back Extreme Up Days
Published March 18th, 2008 in Market InternalsIn mentioning a few reasons why we were seeing an intermediate bottom, I wrote last week, after Tuesday’s 90-90 day (March 12th, 2008):
The best scenario for the bulls would be another extremely strong day which would be as or even more lopsided than yesterday’s. If we get that within a reasonable time, like a week or two, the chances of a solid bottom increases exponentially.
Today the market gave us that second nine to one up day. On the NYSE 88% of traded securities advanced vs. 10% that declined. On the Nasdaq, advancing volume was almost 92% and the index itself jumped +4.2%.
Believe it or not, it was almost exactly a year since the last time we saw bunched up extreme up days. As you’ll recall, after last year’s February stumble, which many blamed on the Chinese market, we got two “nine to one” days very close together in March 2007, just as today when the market was staring into the jaws of the bear:
Previous to that, a double 90-90 signal was triggered on June 29th, 2006. Which you may recall was also near a major low:
If you’re thinking there is something to this pattern, you’re right.
David Aronson, a professor of finance at Baruch College looked at instances in the market (from 1942 to present) when we have these double 90-90 days. His time frame for the second is much wider than what we just witnessed - 3 months. But the results are intriguing nonetheless.
After the special circumstance of a double 90-90 day, the following 60 (trading) days have historically provided a return of +22% instead of a paltry 4.5% annualized otherwise. It is more remarkable when you consider that that return comes with the assumption that you enter the market on the close, after a double 9-to-1 signal was triggered and without adding any dividends!
But there’s nothing special about the 60 day time period. The trend continues, although somewhat weaker, up to 90 trading days after.
If you’re keeping a count down, 60 trading days from today would take us to late June 2008. And 90 trading days, close to the final days of July 2008. This coincides with the calendar countdown for the AAII +50% bearish sentiment.
If you bought the S&P 500 after the March 21st, 2007 signal and held for 60 trading days, you had a ~40% (annualized) return. While if you bought after the June 30, 2006 signal and held for 60 trading days, you had a ~19% (annualized) return.
If you enjoyed this, take a look at the book: “Evidence-Based Technical Analysis” by David Aronson.
Enjoyed this? Don't miss the next one, grab the feed or
7 Responses to “The Hidden Power Of Back To Back Extreme Up Days”
- 1 Pingback on Mar 21st, 2008 at 10:51 pm
- 2 Pingback on Apr 1st, 2008 at 8:23 pm
Trader’s Narrative is powered by
Babak,
Last week you mentioned that it wasn’t quite a 90-90 day yet because the % up stocks fell short of 90%. Is David Aronson’s backtest based on a number greater than 90%? If so then one will have low confidence in applying the study results to the recent two signals, which fell a bit short of 90%.
Also, how does one track 90-90 days using Stockcharts.com?
Just looked into the Lowry’s report further. I’m wondering if you use the points gained measure as well?
Well, it looks like the same points were gained with both 90-10 gains, unlike previous times when one day added to the previous day’s points.
Henry, the important thing is not if they get exactly, absolute, 90%. There is no magic in a specific number. What we are after is a very lopsided market event. A 90-90 is just one way to measure that. AFAIK stockcharts.com doesn’t track it. You can subscribe to Lowry’s or my feed to keep up with them
wes, Yeah but I don’t really thing it matters. We’ll see what happens in a few weeks/months.
Thanks for the reply Babak.
The points gained does matter according to Lowry’s original paper. They said, “Investors should be wary of upside days on which only one component (Upside Volume or Points gained) reaches 90.0% or more, while the other component falls short of the 90% level. Such rallies are short-lived.