When the S&P500 Closes at a New 52-week Low

Monday marked the fifth consecutive session that the OEX has closed under its lower bollinger band, reflecting one of the most extreme oversold conditions in the 25-year history of the index. Similar price action was seen in the days following the 9/11 terrorist attacks and the crash of ’87. There have only been ten instances since the inception of the OEX in which it closed under its lower band four consecutive sessions (much less five). In all previous occurrences, the OEX was trading at a higher level one month later…

OEX Closes Under its Lower Band Four Days
02/20/09… OEX ??? one month later
10/09/08… OEX +0.2% one month later
11/12/07… OEX +3.1% one month later
03/02/07… OEX +2.3% one month later
09/20/01… OEX +10.0% one month later
04/02/97… OEX +7.7% one month later
03/30/94… OEX +0.7% one month later
10/07/92… OEX +3.0% one month later
01/07/91… OEX +10.9% one month later
10/19/87… OEX +11.8% one month later
09/17/85… OEX +4.0% one month later

The sample size is too small to be considered statistically significant, but it does suggest the market could bounce back from intermediate-term selloffs over the next couple of weeks and end up near current levels or higher this time next month.

Short-term, conditions are bound to remain volatile. A bounce still looks likely, but the sustainability of the bounce beyond midweek is in question. As I noted in today’s intraday update, the market typically struggles over the next week when it fails to bounce out of extreme oversold territory. Monday marked only the tenth time since 1970 that the S&P has failed to bounce (or at least hold its own) in the three days following a -200 McClellan reading, with 9 out of those 10 occurrences leading to a lower S&P one week later. We have a large number of bullish short-term setups on the board, but if a decent bounce manifests, every one will fall off by midweek.

Institutions were aggressive sellers once again Monday. TICKscore closed at -99, the fifth consecutive reading below the -50 level. That sent the Cumulative TICKscore line below its November low, continuing the long-term downtrend that’s been in effect since mid-2007. Cumulative TICK settled at -134,800, keeping the 20-day moving average in a solid downtrend. That trend will most likely remain in effect for the next couple of weeks. Momentum leads price, and I think the raw Cumulative TICK readings are a good example of the market’s underlying momentum. The 20-day average of the Cumulative TICK is more like price – note that there’s virtually no lag between the moving average and the S&P itself despite the average being based on the last twenty sessions. Each of those individual sessions can be looked upon as a lead indicator of sorts. And with so many extreme negative readings over the last few sessions, it appears we’re making a momentum low, but not necessarily a (sustainable) price low.

One thing the market has working in its favor is the fact that the S&P500 closed in new lows for the year Monday. You might be surprised to discover that there’s a bias towards higher prices after the S&P closes at a new 52-week low. Over the last thirty years, there have been 101 sessions in which the S&P settled at its lowest level of the past 255 trading days. In 73 cases, or 72% of the time, the S&P was trading higher ten trading days later, better that the 58% at-any-time odds for a higher S&P two weeks later. While that’s not a significant edge, it does suggest viewing an S&P close in new lows for the year from a contrarian standpoint.

On a short-term basis, it was noteworthy that new 52-week lows declined on both the NYSE and NASDAQ despite the fact that the SPX and NDX both closed at a fresh two-month low Monday. That triggers a 2-4 day buy setup. Since the inception of the Nasdaq100 in 1985, we’ve seen this pattern of a new 40-day low for the SPX and NDX that coincided with decreasing new 52-week lows (on both exchanges) a total of 34 times. Note that in 32 out of 34 cases, or 94% of the time, the S&P closed at a subsequently higher level 2-4 trading days later, significantly greater than the 70% at-any-time odds of a higher S&P 2-4 days later…

NDX & SPX Hit One-Month Lows, New 52-week Lows Decline
02/23/09… ???
11/21/08… Higher SPX close two sessions later
10/09/08… Higher SPX close two sessions later
10/07/08… Higher SPX close four sessions later
09/17/08… Higher SPX close two sessions later
06/12/08… Higher SPX close two sessions later
01/23/08… Higher SPX close three sessions later
01/07/08… Higher SPX close three sessions later
11/12/07… Higher SPX close two sessions later
11/09/07… Higher SPX close two sessions later
07/31/07… Higher SPX close two sessions later
05/18/06… Higher SPX close two sessions later
01/03/06… Higher SPX close two sessions later
08/24/05… Higher SPX close three sessions later
04/18/05… Higher SPX close three sessions later
08/13/04… Higher SPX close two sessions later
07/23/04… Higher SPX close two sessions later
09/29/03… Higher SPX close two sessions later
08/05/03… Higher SPX close two sessions later
12/31/02… Higher SPX close two sessions later
10/08/02… Higher SPX close two sessions later
09/20/02… Higher SPX close four sessions later
09/18/01… No higher SPX close 2-4 sessions later
03/20/01… Higher SPX close four sessions later
12/21/00… Higher SPX close two sessions later
10/13/00… Higher SPX close four sessions later
10/18/99… Higher SPX close two sessions later
09/01/98… Higher SPX close four sessions later
08/28/98… No higher SPX close 2-4 sessions later
06/24/93… Higher SPX close two sessions later
02/17/93… Higher SPX close two sessions later
06/18/92… Higher SPX close two sessions later
06/26/90… Higher SPX close two sessions later
11/11/88… Higher SPX close two sessions later
01/21/88… Higher SPX close two sessions later

Categories : Archived EOD Commentary, End-of-Day Commentary

 Print This Post