SPY Rebounds From a 1%+ Intraday Decline but Fails to Close Above Open

The S&P500 tracking stock (SPY) traded down over 1% intraday yet managed to rebound in afternoon trading to finish slightly positive Tuesday. That in itself isn’t that unusual – it’s happened 92 times since 1999. What’s unusual is that the SPY didn’t manage to rally back above the opening, creating a black candlestick in the process. This is highly unusual. Out of the 92 times the SPY rebounded from a 1%+ intraday deficit to close higher, 86 instances coincided with a white candle, meaning the SPY closed above the opening print. In only six cases was the close below the open…

SPY -1% intraday, closes up but below open
09/16/99… S&P500 -2.8% one week later
10/26/00… S&P500 +4.7% one week later
02/22/01… S&P500 -0.9% one week later
09/05/01… S&P500 -8.7% one week later
11/02/07… S&P500 -3.7% one week later
09/19/08… S&P500 -3.4% one week later
11/25/08… S&P500 ??? one week later

NYSE volume declined on Tuesday in choppy trade. When the S&P trades down over 1% intraday and then reverses to close higher on the session, it typically coincides with stepped-up volume. Tuesday’s session lacked that kind of institutional participation. The table below notes the last thirty instances in which the S&P traded in a similar manner – down 1% intraday, up at the close, all on lower NYSE volume. Note the definitive tendency for the S&P to post a subsequently lower close over the short-term. In 24 out of 30 occurrences stretching back to 1999, the S&P closed at a lower level 2-3 sessions later. That 80% win rate is significantly better than the 59% at-any-time odds of a lower S&P close 2-3 days later…

S&P -1% intraday, closes positive, volume down
11/25/08… S&P500 ???
09/26/08… S&P500 -3.9% two sessions later
09/11/08… S&P500 -4.5% two sessions later
09/05/08… S&P500 -1.4% two sessions later
02/22/08… S&P500 +2.0% three sessions later
12/13/07… S&P500 -2.9% two sessions later
11/02/07… S&P500 -2.3% three sessions later
08/10/07… S&P500 -1.9% two sessions later
07/28/04… S&P500 +1.0% three sessions later
08/04/03… S&P500 -1.6% two sessions later
07/01/03… S&P500 +2.3% three sessions later
10/18/02… S&P500 +1.3% three sessions later
09/16/02… S&P500 -2.4% two sessions later
09/09/02… S&P500 -1.8% three sessions later
08/26/02… S&P500 -3.2% two sessions later
08/09/02… S&P500 -2.7% two sessions later
07/03/02… S&P500 -0.1% three sessions later
06/27/02… S&P500 -2.2% two sessions later
06/24/02… S&P500 -1.9% two sessions later
05/01/02… S&P500 -1.2% two sessions later
10/05/01… S&P500 -1.4% two sessions later
09/10/01… S&P500 -5.5% two sessions later
08/16/01… S&P500 -0.9% two sessions later
08/10/01… S&P500 -0.3% two sessions later
05/04/01… S&P500 -0.4% two sessions later
02/22/00… S&P500 -1.4% three sessions later
10/18/99… S&P500 +2.4% three sessions later
10/01/99… S&P500 +3.3% three sessions later
09/16/99… S&P500 -0.8% three sessions later
05/17/99… S&P500 -0.1% three sessions later
05/12/99… S&P500 -1.9% two sessions later

On Monday, S&P futures posted an unfilled upside gap on lower volume than the previous session. Historically, low-volume upside gaps are short-term sells. Consider that since 1990, there have been 187 sessions in which the front-month S&P futures contract posted an unfilled upside gap. Only 49 of these occurrences, or 26%, coincided with a decrease in NYSE volume from the previous session. Those occurrences share a common theme – they tend to relate to short-term tops in the S&P, either immediately or after another day or so of gains. The last thirty instances of this pattern are noted in the table below. Note that in 26 out of 30 cases, or 87% of the time, the S&P posted a subsequently lower close within the next three sessions. That’s significantly better than the 69% at-any-time odds of a lower S&P close within three days…

Unfilled Upside Gap for S&Ps & Dropoff in NYSE Volume
11/24/08… ???
08/22/08… Lower S&P close one session later
03/24/08… Lower S&P close two sessions later
03/18/08… Lower S&P close one session later
12/24/07… Lower S&P close two sessions later
08/17/07… Lower S&P close one session later
07/03/07… No lower close within three sessions
06/15/06… Lower S&P close one session later
05/05/06… Lower S&P close one session later
05/26/05… Lower S&P close two sessions later
10/04/04… Lower S&P close three sessions later
02/18/03… Lower S&P close one session later
10/11/02… No lower close within three sessions
07/29/02… Lower S&P close three sessions later
07/05/02… Lower S&P close one session later
06/17/02… Lower S&P close two sessions later
09/24/01… No lower close within three sessions
04/05/01… Lower S&P close one session later
03/26/01… Lower S&P close two sessions later
11/14/00… Lower S&P close two sessions later
10/19/00… No lower close within three sessions
03/03/00… Lower S&P close one session later
01/10/00… Lower S&P close one session later
12/23/99… Lower S&P close one session later
10/29/99… Lower S&P close one session later
10/22/99… Lower S&P close one session later
10/04/99… Lower S&P close one session later
09/27/99… Lower S&P close one session later
09/03/99… Lower S&P close one session later
06/04/99… Lower S&P close two sessions later
04/05/99… Lower S&P close one session later

While we’re in a positive seasonal time frame until Friday’s close (see yesterday’s column), the setups above suggest overall conditions are likely to be choppy (perhaps down-up-down in the Wednesday-Monday time frame). In the last few years the S&P has taken a hit on the Monday following Thanksgiving, and unless evidence surfaces to the contrary, I’d expect a similar outcome this time around.

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

 Print This Post