>>>You posted earlier on the summation index being at a very low level. I keep the 25-day moving sum of advance/declines on the NYSE and it exceeded -17,000 in September. That negative amount has been worked off. That bested the -12,000 numbers from late 1973 and about 11/87. Have you figured out why the summation index and this sum could go so low yet?<<<
Looking at the summation without adjusting for issues traded, you could make the case that, so far, it has not done anything all that extraordinary, compared to other extremely negative readings, except for making a new, all time low, in nominal terms. The reason this nominal reading is so low is due in great part to the fact that the starting point for this decline was from a low reading to begin with.
9/10/98, -3562, started at 645 on 7/20/98; net change 4207
10/29/87, -1976, started at 1695 on 8/18/97; net change 3671
3/28/80, -2157, started at 1832 on 1/28/80; net change 3989
12/7/73, -645, started at 2427 on 10/12/73; net change 4072
5/27/70, -2247, started at 710 on 3/9/70; net change 2957
Since the summation is just a cumulative sum of the daily readings of the McClellan oscillator itself, you need to compare the oscillator to the 25-day moving sum of advance declines, and they both have indeed turned positive late in September.
When the summation hit 4000 last year, the question was whether this reading was a valid comparison to previous high readings, given the fact that more issues are traded now. One answer was that since, over time, there are both positive and negative readings in the a-d figures over the course of the calculation of the formula, that, for the most part, they even out the readings; in other words, close enough.
However, the expert in this area, Tom McClellan, says as follows:
"The greater number of issues traded on the NYSE definitely has an effect on the amplitude of the Oscillator and Summation Index. By the way, our calculations showed that the all-time low was at -2247 on May 27, 1970. On an adjusted basis, this equates to around -6200 given the current number of issues traded. Wouldn't that be scary?!"
Message 5587883
On a ratio basis, on 8/31/98 we hit -6.69 This compares to:
-7.8 on 10/26/87
-6.98 on 8/23/90
-8.24 on 3/27/80
-7.66 on 12/5/73
-9.41 on 5/26/70 (the most extreme reading looking at data from 1956)
Conclusion: We are not yet oversold enough for "the" bottom, although we do not necessarily have to get to those extremes.
Vitas
|
