S&P Extrema Distribution by Year, from Tim Hesselsweet
Looking at the distribution of extrema (20-day min and 20-day max) on a closing basis in the S&P futures by year:
2003 69
2004 67
2005 68
2006 65
2007 63
2008 60
2009 81
2010 96 (annualized)
The process generating S&P prices seems to have shifted somewhat on this metric. What are other ways markets are trading differently? On what metrics are markets trading in line with prior years?
Kim Zussman responds:
A variation on this is count of 20D highs and lows for non-overlapping 200D periods; this time using SPY counting back from 9/14/10. Here are the counts for 20D high, 20D low, H+L, and respective 200D returns:
Date 20D H 20D L H+L 200D ret
09/14/10 49 22 71 0.04
11/25/09 50 14 64 0.35
02/11/09 13 32 45 -0.39
04/28/08 19 24 43 -0.08
07/12/07 57 6 63 0.18
09/22/06 39 12 51 0.06
12/06/05 41 18 59 0.07
02/22/05 39 21 60 0.10
05/06/04 44 10 54 0.14
07/22/03 36 13 49 0.24
10/03/02 12 42 54 -0.28
12/17/01 25 29 54 -0.07
02/27/01 21 24 45 -0.11
05/11/00 27 23 50 0.06
07/28/99 44 4 48 0.37
10/09/98 42 19 61 0.07
12/23/97 41 11 52 0.18
03/11/97 43 10 53 0.22
05/24/96 47 6 53 0.24
08/10/95 61 3 64 0.23
10/25/94 25 17 42 0.00
01/10/94 26 9 35 0.08
The current H+L is the max since 1994.
Correlation between 20DH and 200D return = 0.81
Correlation between 20DL and 200D return =-0.85 (no surprise)
Correlationbetween H+L and 200D return = 0.3