Trader’s Narrative

As I alluded to in the weekly sentiment review, the bears - or rather, the permabears - have been saying that this rally has been built on a shaky foundation because of the lack of volume.

This in spite of record shattering daily volumes during the deepest and darkest days of the sell-off. And in spite of the calendar being in the thick of the “summer doldrums”. And in spite of the total volume averages reaching unheard of extremes.

Here is the 15 day moving average of the NYSE total volume:

And here is the 15 day moving average of the Nasdaq total volume:

As you can clearly see, ever since 1999, we’ve followed a predictable script on Wall Street: the summer sees the lowest volume period of the year.

I’ve marked the month of August in yellow on the charts so you can compare what has happened previously on 8 previous occasions.

I’m not really sure what to call this indicator. The new highs/new lows indicator sounds awkward. I wish some technical analyst with a cool sounding name had created it so we could be saved from such banality.

Whatever you wish to call it, this indicator looks at the 52 week new highs and the 52 week new lows. It is calculated by taking the new highs and dividing by the sum of the new highs and the new lows.

So when it returns zero, there were no new highs and when it returns the other extreme (100) we know that were no new lows. And when it is at 50, this means we had the shocking circumstance of exactly the same number of new highs and new lows. Simple, right?

On with the analysis. Let see what this indicator can add to our understanding of the market. Because it can be a bit jittery, I like to smooth it out with a very short term moving average (5 day). So when I mention the indicator below, I’m really referring to this moving average.

Bottoms Up!
For one, new highs new lows (NHNL) does a phenomenal job of pointing out inflection points after a downtrend. Historically, when it breaks below 20, the market has taken enough of a beating and finds a bottom either immediately or within a short period of time. Of course, there is no reason why it can’t go right down to its minimum of zero.

Last time this happened was in July and September 2002 - the darkest (and final) days of the bear market. But it reaches such extremes rarely. Usually it reverses at or around 20. The last time it did that was last year (July 2006) which marked the last great buying opportunity of the summer doldrums.

Shallow Retracements
But inflection points don’t necessarily even need a reading of 20 on this indicator. For example, this year’s early March bottom was formed at a reading of only 40. Shallow retracements are a hallmark of strong bull markets. Take for example what happened in 2003. After bottoming in just below 10 in the early months of the year, both the market and this indicator shot up to much higher ground.

From the summer onwards however, the indicator saw almost no more real retracements. Eventhough the S&P 500 didn’t go straight up, each time it faltered, the new highs new lows indicator only dipped minimally. And then recovered right away. It spent most of the time hugging the 100 line. And its largest retracement only reached only 80. On it went like that for the rest of 2003 and into 2004.

Now that’s a strong bull market!

Another characteristic of this indicator that I noticed is that shallow retracements have different meanings. It depends if they happen after an extreme low has already been reached prior to it or whether the shallow retracement occurs after a maximum has been reached.

Caution
Let me explain. If we keep walking forward with the example I last mentioned, in early 2004 we had our first correction of the new bull market. But the NHNL indicator dropped to almost 70. The next correction was in May 2004 with the market dropping a little bit lower than the March 2004 lows. NHNL fell again but this time to almost reach 20. Then again in a few months, the market fell again (August 2004) and this time the NHNL got a real extreme low reading (below 10).

My point is that the first time the NHNL fell, it wasn’t really extreme. And since it was after it had already been at the extreme high, it was telling us that this wasn’t a real market bottom. Only after reaching a real extreme (below 10) did the market finally find its footing again.

Smooth Sailing
So shallow retracements of NHNL while the market is dropping are not to be trusted. But if they occur after it has already reached an extreme they signal a continuation of the bull move already underway.

Let me explain by continuing the 2004 story. In October 2004, the market again dipped. This time however the NHNL indicator only dipped below 50 - a shallow retracement. But since this was immediately after a very extreme retracement (below 10) in August 2004, it was a signal that the bull move was still going strong.

So what does this indicator tell us about the current market?

For starters, the correction in March 2007 corresponded to a shallow NHNL retracement (40). And since it ocurred after it had already been high for some time this meant that the decline wasn’t finished. But of course, it was - atleast as far as we’re concerned. This was the first time since I have data for the NHNL indicator that such a shallow pullback following a time at the maximums was the real deal (significant bottom).

And the latest correction which we had hardly registered on this indicator. In early June 2007 it slipped just below 80. So are we seeing just a correction within a strong bull market?

All the NHNL indicator can really tell us is that the latest correction was very muted. Even more so than the March 2007 one. But as long as we are in a strong bull market, that isn’t a problem. Short and shallow retracements are par for the course. The best way to be more certain that we are seeing that sort of scenario is to reach new highs.

Click to Enlarge Graph
Notice how the shallow pullbacks (orange arrows) that follow immediately after a climax low are good buying opportunities (October 2004, May 2005 and August 2006) but those that happen some time after treading maximums are not (August 2005 and March & April 2006)

It is notoriously difficult to pinpoint tops because unlike bottoms, they tend to form over a longer period of time. Still, there are several factors which are making me cautious here - just as I was bullish in mid-March.

Pedal to the Metal
Last week marked a historic occasion for the market. Albeit one that went unnoticed: the Dow closed positive in 19 out of 21 trading days. This was only the third time this has happened! The other two times? I’m not sure you want to know…

The first time was in August 1927 when the Dow corrected 10% (only a pause in its rocket ride to the 1929 top). The second was in July 1929 just before the last surge of the great bull market of the 1920’s. I don’t think this means necessarily that we’re going to crash. Just something to tuck into your hat.

Sentiment
The usual sentiment measures aren’t actually showing a lot of froth. Investor’s Intelligence is showing a tad too much bullishness from newsletter writers. But other sentiment measures are lukewarm at best. Put/Call ratios are, as well, in the middle ranges.

Annual Stock Market Cycle
There’s a saying on Wall Street: “Sell in May and go away!” for good reason. We have just entered one of the weakest portions of the 12 month stock market cycle. Many studies have shown a persistent bias towards the later months in the year (Sept - March) for relative outperformance. Obviously this is a pattern that emerges over many, many years and no one specific year has to follow it. Also, there are some rumblings that the pattern may be pushed out to late May or even June. According to Hirsch, from 1985 to 1997 May has been a stronger month than before. But summer doldrums are still summer doldrums.

Technical Signposts
Mark Hulbert mentioned today that Investors Intelligence is raising a red flag citing their “buying climaxes” proprietary measure. A buying climax is defined by II to have taken place when when a stock makes a 52 week high and then closes the week down. This is seen as a sign of distribution and as such, marks tops.

Last week there were 206 buying climaxes. But readings as high as 500 or 600 are not unheard of.

Conclusion
If you’ve been intelligent (aka lucky) enough to participate in this short burst rally, consider paring your longs and taking a defensive posture to allow the market the respite it needs.