A Shift into Reverse?

Fading momentum measures, combined with the intraday reversal Friday, increase the odds for some profit-taking next week

By Paul Cherney

Note: Paul Cherney will be on vacation through Monday, Sept. 1. His column will return on Tuesday, Sept. 2.

Momentum measures based on end of day data for both the Nasdaq and the S&P 500 are still technically in positive territory but have lost upside momentum and that fact, combined with the intraday reversal in Friday's markets increase the chances that there could be some profit-taking next week.

Near the close of trading on Friday, the 10-day exponential moving average of the VIX (market volatility index) was 20.18; quite often, a VIX move above its 10-day exponential moving average can coincide with stock price weakness.

Resistance: The Nasdaq resistance is 1778-1829.58.

Immediate resistance for the S&P 500 is 1003-1015.41. Its focuses of resistance are 1005-1008 and 1010-1015. The bigger picture of resistance, which was established by price action in June, 2002 is that the S&P 500 has a band of resistance at 1008-1041 with a focus at 1020-1031.

If you look at the overlap of resistances, the 1008-1015 layer is a stumbling block for S&P 500 prices.

Supports: The S&P 500 has support at 991-984, stacked at 985-974 and 976-960.84. I cannot rule out that sometime before the middle of September that 949-912 support will be tested. The S&P 500 still technically has a small chance of printing under 950. These markets have not seen a one-third or a 50% retracement for the move up since March's lows, and retracements like that are common.

The Nasdaq has a broad layer of support at 1758-1722. The index has a shelf of price traffic at 1753-1737 which has to be considered a focus of support. The Nasdaq has stacked support at 1736-1711, 1703-1695, and 1694-1681.31.

Special Study: Next week is the week before the Labor Day weekend.

Here are some historical statistical facts about the performance of the S&P 500 in the week before the Labor Day three day weekend. All numbers based on S&P 500 data from 1972 to 2002.

For the week prior to Labor Day, historically, the S&P 500 has gained ground 20 of 31 times or 64.5% of the time.

For the individual trading days in the week prior to Labor Day:

Mondays have seen closing gains (S&P 500) 17 of 31 times, 55% of the time.

Tuesdays have seen closing gains (S&P 500) 16/31, 52% of the time.

Wednesdays have seen closing gains (S&P 500) 17/31, 55% of the time.

Thursdays have seen closing gains (S&P 500) 8/31, 26% of the time.

Fridays have seen closing gains (S&P 500) 21/31, 68% of the time.

There is a real statistical anomaly in the Thursday data. In 1979 and 1992, according to my data, the S&P 500 was unchanged on the day. Unchanged on the day does not happen very often. In 2002, the Thursday before Labor day weekend came close to doing it again when the S&P 500 only managed to gain 0.07 point which represented a gain of 0.01%.

If you averaged all of the Friday performances (gainers and losers), the gain was 0.34%. If you averaged just the gaining Fridays, the average gain was 0.76%. Fridays had six trading days in the series where prices moved more than 1% as of the close: five were gainers of more than 1% and one was a loser of more than 1%.

The best chance historically for a losing session is the Thursday before the weekend, and the best chance for a gaining session is Friday. I wondered if those Friday gains might be rebounds from Thursday losses so I compared losing Thursdays to Friday's closing performance (to see how many times a loss on Thursday was followed by a gain on Friday). There were 21 losing Thursdays, 13 of them were followed by gaining Fridays, that means 62% of the time a Thursday loss was followed by a Friday gain.

I looked at each of the trade days to see which one had historically demonstrated the greatest likelihood of seeing prices close with greater than 1% movement on the day (in either direction) and curiously enough it turned out to be Thursday which had 11 closing gains or losses of greater than 1% (even though Thursdays also had two days (almost three days) which were unchanged). There were three Thursdays before Labor Day which saw gains of more than 1% and there were eight which saw losses of more than 1%. If you look at the Friday performances after a Thursday loss of more than 1%, five of them were followed by gaining Fridays -- 62.5% of the time.

Cherney is chief market analyst for Standard & Poor's

Before it's here, it's on the Bloomberg Terminal.