Daily Speculations The Web Site of Victor Niederhoffer and Laurel Kenner

Home

 

 

The Chairman
Victor Niederhoffer

 

 

2005 All content on site protected by copyright

 
Write to us at: (address is not clickable)

01/21/2006
A Note on Friday's Decline, by Victor Niederhoffer

Our readers know that when a big decline occurs, like the 23-pointer last Friday, it's generally time to take out the canes. Such declines have been incredibly bullish for all foreseeable periods of the future since 1850.

For example, seven of the last eight times a 20-pointer has occurred, the market was up by three days later, with the one decline's being 1% and the modal rise of 2% occurring three times. I enumerate the dates so that those with a pencil and envelope can quietly study the anthills:

01/24/2003
01/30/2003
03/10/2003
03/24/2003
05/19/2003
08/05/2003
07/21/2004
10/20/2005

What is striking about Friday's decline is that it came out of the clear blue sky. It was the biggest drop since March 10, 2003. I studied such events for insight, and my studies did not stop with Laplace's methods of estimating probabilities of such "out-of-clear-blue-sky events" or the myth of Icarus. I believe this a worthy area to contemplate.

The Minister of Non-Predictive Studies is looking over my shoulder now, so all I can add is that the large declines tend to cluster. And once they cluster, as in 1907, the 1930s and 2001- 2002, the reward risk/ratio is changed. However, before his hook reaches me, I point out that everything stated above has been tested.

Jack Tierney comments:

What is striking about Friday is that we have gone over two years with only one other 2% or more downward one-day adjustment. When I read the numerous posts on DailySpec kvetching about bearish sentiment during this same period it's amazing that we have not had numerous similar days. Unless of course, there has been a disconnect between perception and reality.

We went through 2004 without the AAII marking a single week of majority bearish sentiment. This has happened on seven occasions prior to 2005; the following years were always down. Recently this same measure has been at its highest, or second highest, level ever.

The remaining seasonal bears (as opposed to permabears) are so few that, were they a species, they could apply for endangered status. The optimists continue to triumph, and try as they might to create a wall of worry, the only impediments are the occasional stumbling block.

For those who have given up on CNBC, you ought to check out their late night programs CNBC Asia and CNBC Europe. They're totally different animals with legitimate analysis and little or no cheerleading. And Louisa has the greatest widow's peak since Morticia Adams.

GM Nigel Davies adds:

'Unexpected combinations' are quite an important facet of chess. But usually they are only 'unexpected' from the point of view of conventional chess understanding, thus they offer us the possibility of improvement, if an explanation can be found.

Dr. Kim Zussman comments:

Using historical S&P 500 data, I counted weeks with returns below -2%. I then regressed each year's count versus all calendar years from 1950-2005:

Slope
Coefficients 0.0735
Standard Error 0.0326
t Stat 2.2556
P-value 0.0282

Over this period the number of -2% weeks per year generally increased and the relationship is significant. I suspect going back further to the 1920's might change this, but that may not be relevant.

The average number of 2% down weeks per year was 5.8, but the last two years were below average:

1950   3   1960   6   1970   9   1980   8   1990  14  2000   9
1951   5   1961   1   1971   4   1981   9   1991   3  2001  11
1952   2   1962   9   1972   2   1982  11   1992   2  2002  16
1953   3   1963   2   1973  13   1983   3   1993   2  2003   6
1954   1   1964   0   1974  18   1984   8   1994   5  2004   3
1955   5   1965   2   1975   8   1985   2   1995   0  2005   4 
1956   5   1966   8   1976   3   1986   8   1996   5  2006   1
1957   9   1967   2   1977   7   1987   9   1997   7
1958   0   1968   2   1978   8   1988   7   1998   7
1959   1   1969   8   1979   4   1989   4   1999  11

More by Victor Niederhoffer