I tinker a lot with different market indicators and usually end up at a dead end. That’s no fun but I don’t consider the time spent on such fruitless endeavors as wasted because they are really mental calisthenics. Sometimes though, quite rarely, the result is interesting enough or useful enough that it is worth sharing with others.
Back when I was scribbling into the void, I started this blog with a wacky idea that combined two measures of the percentages of stocks above their moving averages: Timing the Market with Percent Above Moving Average Ratios. Yes, that is a mouthful. But it is also an interesting twist on the breadth indicator.
Most recently I’ve been fixated on the number of new 52-week highs in the market. Their significance is derived from their ability to signal a market turn well ahead of the actual top. Of course, this makes it both intriguing and infuriating because even if the number of new highs tops and declines, you don’t really know when exactly the market is going to crest. It could be right away or months in the future.
This week I mentioned the divergence in the number of new highs. As I continued to play with this concept I decided to look at the cumulative NASDAQ new 52 week highs. By itself, that turned out to be rather boring because it just looks more less looks like a straight line going up across the chart.
To make it more interesting and bring out the “bumps” in the chart I took a 10 day rate of change (ROC) of the cumulative new high data. Why 10 days? There is nothing magical about that number but it accounts for 2 weeks of trading and is long enough to be smooth but short enough to be agile. You can easily choose another and see if it works better.
I have no idea if anyone has ever used the new 52-week high data like this. There is a distinct possibility but I’ve never seen it. If you have, then please let me know that I’m just re-inventing the wheel here and should go for a brisk walk instead. Until then, I’ll continue with your indulgence.
The first thing you notice is that ‘peaks’ in the ROC (bottom chart) correspond with market peaks in the S&P 500 index. There were several in 2007: April, July and of course, October. By the way, I realize that we are looking at a very short period of time and that it may not be enough to glean anything useful. But this is a starting point, not a conclusion.
Another point is that the higher the ‘peak’ the more significant the top seems to be. That would be natural since markets can not push forward without a significant number of individual stocks pushing higher into new highs. During most of the bear market, there is a lull in the ROC as new highs become as rare as Dodo bird sightings. Then more recently, we come to see several ‘peaks’ as the ROC has increased in volatility.
The other observation is that the 3 peaks that we saw in the ROC in 2007 were each lower than the previous. So as the market was heading higher, the momentum was waning. We have the opposite right now with the peaks overtaking the previous one.
Finally, the slope of the ROC itself may be a useful indicator. It seems that when it is rising, the stock market is also rising. When it curves down and heads lower, the market also falls. Or am I just seeing things?
I’ll continue to play around with this a bit more but wanted to share it with you in its raw form. I’d love to hear your ideas, whether you agree with this line of reasoning or not. Drop me a comment below with your thoughts.
Enjoyed this? Don't miss the next one, grab the feed or