80 year old DOW Record article
This guy is, "The glass is half empty type" :suspicious:
Much ado about nothing
Commentary: There's little significance in Dow's "up-days" record
By
Mark Hulbert, MarketWatch
Last Update: 12:11 AM ET May 8, 2007
ANNANDALE, Va. (MarketWatch) -- Psychologists have long known that the human psyche has grave difficulties contemplating randomness.
Consider what happens, for example, when human subjects have been presented with a series of random numbers and then asked by researchers whether there are any patterns in the data. Overwhelmingly, the subjects claimed to have detected meaningful regularities.
We should bear this in mind give all the hoopla Monday about the Dow Jones Industrial Average (
$INDU :
Dow Jones Industrial Average
News ,
chart ,
profile ,
more
Last:
13,312.97+48.35+0.36%
4:30pm 05/07/2007
<IMG class=pixelTracking height=1 width=1 border=0>
$INDU13,312.97,
+48.35,
+0.4%) matching an 80-year-old record for number of "up" days (24 of 27). That certainly may appear to our feeble cognitive faculties to be significant, but that doesn't make it objectively so.
To determine if a pattern is genuinely meaningful, we need to turn to the science of statistics. And upon doing that in this case, we find that the Dow's recent achievement is a lot less significant in reality than in our perceptions.
I mentioned this at the end of a column last week, when I wrote that the stock market's day-to-day fluctuations are largely random - "statistical noise, in other words."
See May 4 column
I'm revisiting the issue again, given the number of impassioned e-mails I received that objected to this conclusion.
Your e-mails prompted me to check in with Rudi Fahlenbrach, an assistant professor of finance at Ohio State University's Fisher College of Business. What are the odds, I asked him, of the market closing up in 24 of 27 trading sessions, assuming that whether today is an up or down day is independent of what the market did yesterday? This is equivalent to asking how many times 27 flips of a coin would come out with 24 heads.
The Dow has existed for more than 30,000 trading sessions since it was created in 1896. If you flipped a coin that many times in a row, he argued, probabilities would be low that you'd ever have 24 heads in a sequence of 27 flips.
But, Fahlenbrach pointed out, this coin-flip analogy isn't perfect. There actually is a slightly greater probability of the market going up on any given day than going down, given the market's long-term uptrend. To keep the analogy to a coin flip, we have to imagine a coin that comes up heads more often than tails. For example, the Dow has risen on 52.67% of the days since 1896.
When Fahlenbrach entered this 52.67% probability into his calculations, instead of 50%, he found that what we've seen over the last 27 trading days should happen between one and two times in an 111-year period. Since the actual number of occurrences is two, what has happened "is right in line with what you'd expect" given the assumption that the market's direction today is independent of its direction yesterday.
This finding doesn't in and of itself guarantee that the market's direction on a given day is independent of its direction the day before. But since we can explain what we've seen over the past 27 days in terms of a purely random sequence of day-to-day fluctuations in the market, Fahlenbrach's calculations suggest we should not be too quick to ascribe meaning to what has happened over the past 27 trading days.
Another reason to be skeptical that the Dow's recent achievement is all that meaningful: It has not been matched by the other major market averages. Consider the Dow Jones Wilshire 5000 index (
97199001 :
Dow Jones Wilshire 5000 Composite Index
News ,
chart ,
profile ,
more
Last:
15,222.62+28.54+0.19%
15,222.62,
+28.54,
+0.2%) , which reflects the combined value of all publicly-traded stocks in the U.S. In contrast to the Dow, which has risen in 24 of the past 27 trading sessions, the Dow Jones Wilshire 5000 has risen just 18 times.
Not bad, of course, but nothing to write home about.
Yet another reason to be skeptical came in an e-mail from Dan Seiver, editor of the PAD System Report newsletter, an emeritus professor of economics at Miami University of Ohio and currently a visiting professor of economics and finance at San Diego State University and the University of San Diego. Seiver pointed out that the stock market's day-to-day fluctuations are not completely statistical noise, since there is a greater-than-average chance the market will rise tomorrow if it rose today, and vice versa.
But, as Fahlenbrach pointed out, if we take this tendency into account, then what we've seen over the past 27 days becomes even less exceptional.
The bottom line? The Dow's achievement over the last 27 days may seem so improbable as to be highly significant. But a number of statistical tests fail to support that conclusion.
Did someone go on vacation? I was hoping I clawed my way out of the bottom decile.:suspicious:
I hope he is enjoying the weather.
