Jim Nickerson
10-19-07, 11:47 PM
I don't believe in magic, but I do believe in market timing in some degree as opposed to a buy-and-hold strategy. That I believe in timing and attempt to get in and out of market moves doesn't mean I necessarily always succeed, because I don't. My simplest opinion is there are good times to be long and good times to be short, the only problem is to figure which is the right time.
In one or both of the bullish, bearish information threads from time to time I put up articles by Sy Harding, who posts every Friday on decisionpoint.com. His article today seems pertinent with regard to "seasonal" timing, and he has done some research on this notion--which I haven't read and won't--and he explains his timing model explicitly in this article, which I think is worth reading--it is short.
http://www.decisionpoint.com/TAC/HARDING.html
My 1998 research discovered that on average the best date to enter in the fall is October 16. However, it also revealed that the length of the favorable and unfavorable seasons vary significantly from year to year. So we incorporated the use of MACD, to be used in conjunction with the calendar, to better pinpoint the entries and exits.
The rule is that when October 16 arrives, if the MACD indicator is on a 'buy signal', that is showing that short-term market momentum is to the upside, the entry is to be made at that time. However, and this is the biggie, if the MACD indicator is on a 'sell signal', indicating that short-term market momentum is to the downside, the entry is delayed until MACD triggers its next 'buy signal'. That has delayed the entry in some years to as late as late November.
In the spring, the best exit date on average is April 20. However the same MACD indicator is used to sometimes delay the exit. The rule is that if MACD in on a buy signal when April 20 arrives, the exit will be delayed until MACD triggers its next sell signal. In some years that has extended the 'favorable season' by as long as two months.
This year has been very interesting. The exit signal was triggered on May 15, which extended last year's favorable season by about a month, and allowed for additional gains. During the unfavorable season this summer the S&P 500 experienced a 10% correction in July and August, while seasonal investors were safely on the sidelines adding to profits from interest on cash. It looked like the Seasonal Timing Strategy would surely re-enter in the fall with the market at a lower level than the May exit, and so would easily outperform the market for the year.
However, as you know, the market rallied back off the August low, reaching new highs, and the seasonal investor was behind by two or three percent even counting the interest being earned on cash.
More worrisome, MACD was recently still on a buy signal as October 16 approached, making it look like the seasonal investor might be re-entering near those high prices. That had me worried, given the many negatives that have been piling up for the economy.
Fortunately, or so it seems at this point anyway, the market lost enough momentum that as of the close a week ago Friday, MACD triggered a 'sell signal'. And since it remained on that sell signal on Tuesday, October 16, the seasonal entry will not take place until it triggers its next buy signal.
He doesn't reference which index he uses from which the MACD calculations are made, but I assume he uses the SPX. Because he does not explain the mechanics of exactly what the MACD must do in order to generate "buy" and "sell" signals, one must develop familiarity with that indicator. I am sure it is explained somewhere on the web. Personally, I have a book Trading for a Living by Alexander Elder, John Wiley & Sons, 1993 which explains more about indicators than I care to know, but the book has been useful to me in using the technical indicators of Wilder's RSI and the MACD. These two indicators are available on stockcharts.com, online.wsj.com, and bigcharts.com all of which allow interactive charting.
As far as I am concerned with the SPX when it hit its most recent high on close on 10/9 or intraday on 10/11, either day was a non-confirmation using the MACD histograms and the RSI (which Harding doesn't reference). I don't know why Harding's statement that the MACD was "still on a buy singal as of October 16" is correct. At any rate, now it is clear that the MACD is not on a "buy signal" and if one is inclined to participate in being long during the period of "favorable seasonality," it seems reasonable to me to wait for a buy signal to appear in the MACD and for me the RSI too.
In one or both of the bullish, bearish information threads from time to time I put up articles by Sy Harding, who posts every Friday on decisionpoint.com. His article today seems pertinent with regard to "seasonal" timing, and he has done some research on this notion--which I haven't read and won't--and he explains his timing model explicitly in this article, which I think is worth reading--it is short.
http://www.decisionpoint.com/TAC/HARDING.html
My 1998 research discovered that on average the best date to enter in the fall is October 16. However, it also revealed that the length of the favorable and unfavorable seasons vary significantly from year to year. So we incorporated the use of MACD, to be used in conjunction with the calendar, to better pinpoint the entries and exits.
The rule is that when October 16 arrives, if the MACD indicator is on a 'buy signal', that is showing that short-term market momentum is to the upside, the entry is to be made at that time. However, and this is the biggie, if the MACD indicator is on a 'sell signal', indicating that short-term market momentum is to the downside, the entry is delayed until MACD triggers its next 'buy signal'. That has delayed the entry in some years to as late as late November.
In the spring, the best exit date on average is April 20. However the same MACD indicator is used to sometimes delay the exit. The rule is that if MACD in on a buy signal when April 20 arrives, the exit will be delayed until MACD triggers its next sell signal. In some years that has extended the 'favorable season' by as long as two months.
This year has been very interesting. The exit signal was triggered on May 15, which extended last year's favorable season by about a month, and allowed for additional gains. During the unfavorable season this summer the S&P 500 experienced a 10% correction in July and August, while seasonal investors were safely on the sidelines adding to profits from interest on cash. It looked like the Seasonal Timing Strategy would surely re-enter in the fall with the market at a lower level than the May exit, and so would easily outperform the market for the year.
However, as you know, the market rallied back off the August low, reaching new highs, and the seasonal investor was behind by two or three percent even counting the interest being earned on cash.
More worrisome, MACD was recently still on a buy signal as October 16 approached, making it look like the seasonal investor might be re-entering near those high prices. That had me worried, given the many negatives that have been piling up for the economy.
Fortunately, or so it seems at this point anyway, the market lost enough momentum that as of the close a week ago Friday, MACD triggered a 'sell signal'. And since it remained on that sell signal on Tuesday, October 16, the seasonal entry will not take place until it triggers its next buy signal.
He doesn't reference which index he uses from which the MACD calculations are made, but I assume he uses the SPX. Because he does not explain the mechanics of exactly what the MACD must do in order to generate "buy" and "sell" signals, one must develop familiarity with that indicator. I am sure it is explained somewhere on the web. Personally, I have a book Trading for a Living by Alexander Elder, John Wiley & Sons, 1993 which explains more about indicators than I care to know, but the book has been useful to me in using the technical indicators of Wilder's RSI and the MACD. These two indicators are available on stockcharts.com, online.wsj.com, and bigcharts.com all of which allow interactive charting.
As far as I am concerned with the SPX when it hit its most recent high on close on 10/9 or intraday on 10/11, either day was a non-confirmation using the MACD histograms and the RSI (which Harding doesn't reference). I don't know why Harding's statement that the MACD was "still on a buy singal as of October 16" is correct. At any rate, now it is clear that the MACD is not on a "buy signal" and if one is inclined to participate in being long during the period of "favorable seasonality," it seems reasonable to me to wait for a buy signal to appear in the MACD and for me the RSI too.