My first observation would be that only three of those were at a short term high, though another five topped out within 2% of the punch level, before making a retracement that went at least back below the punch level and ranged in size from 3% to 21%. The exception was in June 1997 which went into an eight day upper band ride that ran up 80 points before retracing 70 points. This is a fairly bearish history, with one strong exception.
My second observation would be that though the following day was mixed, there was only one instance where the next day close was more than 4 points higher. The following week also averaged flat to down.
To see anything else really useful from the results here though I would divide the punches into two group, the first in the relatively high-friendly March through August period, and the second in the relatively high unfriendly September through February period. This leaves four previous instances in the first group which were:
- 1998 July - At the 1998 summer high, next decline 21%
- 1997 June - Rose 10% into July first high double top
- 2013 May - Rose 1.5% into spring high made following week, then decline 7.5%
- 2010 April - Rose 2% into spring high made next three weeks, then decline 17.1%
That last group gives us a reasonable feel for the odds here in my view. 75% we make a significant high shortly followed by a substantial retracement that should at least deliver a retest of 1850 and might go a lot lower. The other 25% is that we see an unexpected breakout here that would at least break 2000 SPX and might go considerably higher. Even that outlier June 1997 high however moved up to make the first high of a double top, leading to a 13.5% retrace in October 1997 that retraced 4% below the weekly punch level, and so regardless of what happens here history suggests strongly that we are likely to see a decent retrace (average 15%) decline starting sometime in the next four months.
So what is the immediate setup here? Well I'm favoring an imminent high followed by a strong retracement, and my main reason for favoring that here is the primary rising channel from the October 2011 low. SPX tested channel resistance again on Friday, and break up through resistance on such a channel would be rare here, though not unknown. Of these nine previous instances only two hit a strong established primary trendline in the next few days, and both of those reversed there. The first hit of channel resistance at 1850 was firmly rejected into the 1737 low. We'll see now whether this hit at 1950 will do better. SPX weekly chart:
If this goes the other way, and we see a break up across multiple indices through these and other impressive resistance levels, then the chances are that would run up a while further. I'd be surprised in that case not to see a test of 2000 SPX and would have my eye on that 2030 rising wedge target. Even in this case however, the odds would favor that break up being part of the topping process for a significant high not much higher than that.