A Change in our DJIA Timing System
(May 9, 2007)
Dear Subscribers and Readers,
I hope you have all received our intraday email (it was sent out at around 2pm Eastern Standard Time on Tuesday) signaling our shift from a 100% long position to a completely neutral position (i.e. 0%) in our DJIA Timing System. Our 100% was exited at a Dow Jones Industrial Average print of 13,299. I have decided to send out this commentary today instead of tomorrow, given that this represents our first signal change since over seven months ago. Let us now quickly review our gains from our two most recent long positions:
1st long signal entered: 50% long position on September 7th at 11,385, giving us a final gain of 1,914 points
2nd long signal entered: Additional 50% long position on September 25th at 11,505 giving us a final gain of 1,794 points
For subscribers that have a good memory, you may recall the pessimistic atmosphere that was prevalent when we entered into those long positions back in September of last year. I remember receiving emails and getting comments as to how “wrong” I was, and needless to say, the comments at other websites were not too kind.
I am not mentioning this for bragging purposes, but merely to illustrate how sentiment can play a huge role in timing the markets, especially when one is trying to “bottom fish.” Picking tops is a skill (some would just call it luck) that is infinitely more tricky – as the emotions that are flowing when investors are greedy are nowhere near as powerful as those emotions when investors are panicking. As a result, stock market tops can take weeks or months to fall, and are usually preceded by divergences as opposed to extreme optimism. This is significantly different to a stock market that is bottoming, as significant bottoms can take mere days to form. For example, the October 1987 crash and the February 27 to March 14 “swoon” we had just experienced most recently. Such bottoms are usually accompanied by widespread pessimism and being “angry” at the stock market or to whoever is calling for the stock market to rise, such as what we had experienced during the August to September period of last year.
So where are we now in terms of sentiment? According to the various surveys out there (such as AAII, Investors Intelligence, and Market Vane Consensus), investors sentiment is not at a bullish extreme (see our weekend commentary for the chart) – we actually witnessed more bullish readings in mid November, mid January, and late February (yes, just a week before the February 27 swoon), but like I mentioned before, a stock market top does not have to be immediately preceded by extremely bullish sentiment. As a matter of fact, bullish sentiment actually peaks weeks or sometimes even months before a top in the stock market. More likely, a stock market top is usually preceded by a sustained period of time of bullish sentiment – and is usually IMMEDIATELY preceded by moderately bullish sentiment, such as a lower high in out average of the AAII, Investors Intelligence, and Market Vane's Bullish Consensus chart (e.g. the most recent May 2006 top, as well as the March 2005 top, the March 2002 top, the July 1998 top, etc.).
In other words, the necessary condition for a short-term top for the market is in place in terms of our most popular sentiment indicators. This latest “lower high” in bullish sentiment is also confirmed by the HSNSI (the Hulbert Stock Newsletter Sentiment Indicator), as well as the ISE Sentiment Index (this indicator is still somewhat oversold but has been rising quite a bit lately), and the equity put/call ratio.
Of course, as I have mentioned before, the best indicator of a stock market top (besides valuation, which is not applicable here as we are only looking for a short-term top) is always divergences, divergences, and more divergences. Prior to the January 2000 top in the Dow Industrials, the NYSE A/D line had already topped out in April 1998 and had been steadily weakening. Meanwhile, the Dow Transports had already topped out in July 1999, while many emerging markets had already topped out all the way back in mid 1997 – right before the onset of the Asian Crisis.
The January 2000 top in the Dow Industrials is a bit of an extreme, of course. Not only had these divergences been going on for months, but valuations (I am using the S&P 500 here as a proxy) were hugely overstretched as well, not only on an absolute P/E basis but also on a relative basis, such as vs. bond yields, the yields of REITs, yields of money market funds, etc. On a U.S. large cap growth basis, valuations were even more overstretched, and from a global standpoint, this was all made more extreme since the U.S. dollar was also hugely overvalued at that time. In retrospect, this is all clear to our readers – but at the time, you could have literally been thrown out of a party for espousing these views. I know – because I was trying to tell that to all my friends and acquaintances at the time.
But anyway, I digress. One of the more reliable “divergence indicators” has historically been the price behavior of the American Exchange Broker/Dealer Index (XBD). As the saying goes, a bull market cannot be sustainable without the participation of the broker dealers, since the profits of broker/dealers are usually dependent on the increase of trading volume, or in other words, a bull market. Therefore – historically, any weakness in a bull market is usually preceded by weakness in the XBD. Following is a daily chart of the XBD (courtesy of Decisionpoint.com) from January 1, 2006 to May 8, 2007:
As mentioned on the above chart, we have again been witnessing a divergence in the XBD – starting with the fact that the XBD has been extremely weak since the correction from late February to mid March. That is, while it only took a month for the Dow Industrials to recover and then to make a new all-time high, it took the XBD nearly twice as long to do so. And while the Dow Industrials had continued to make all-time highs over the last few days, the XBD has actually been exhibiting extreme weakness. While the weakness is not as pronounced as the weakness during May 2006 or February 2007, it is nonetheless a red flag – and thus this author will most probably not initiate a long position in our DJIA Timing System again until we have witnessed some kind of correction both in the XBD or the Dow Industrials or the S&P 500.
Another significant divergence that I have been mentioning over our last few commentaries has been the non-confirmation on the upside of the Dow Industrials by the Dow Transports, as shown in the following daily chart of the Dow Industrials vs. the Dow Transports from January 1, 2006 to the present:
As I have mentioned previously – while the Dow Industrials has been making all-time high after all-time high for the last month or so, the Dow Transports, on the other hand, had actually topped out on April 25th. Sure, the Dow Transports is now a mere 26 points away from that all-time high (it has risen 47 points over the last two days), but the damage has already been done. If the stock market does make a short-term at these levels – then don't say there hasn't been ample warning beforehand.
One reason why the Dow Transports has been relatively weak is the pronounced weakness in the American Exchange Airline Index, as shown by the chart below, again courtesy of Decisionpoint.com:
As one can see, the XAL actually topped out in mid January 2007, acting as a precursor to the mid February short-term top and the late February to mid March correction. More ominously, the XAL has failed to recover along with both the Dow Industrials and the S&P 500 since the March 14th bottom, and has actually continued to decline! For comparison purposes, please also note that the XAL also topped out a few weeks prior to the May 10th top. Given that the airlines are typically the most sensitive to economic and corporate profits change, my guess is that this may actually be a long summer for the bulls.
I will discuss our latest change in our DJIA Timing System in more detail in this weekend's commentary (along with featuring a brief write-up of the Berkshire Hathaway Annual Shareholders' Meeting by a close friend and colleague of mine, Richard Faw) – but given that valuation and liquidity still remains favorable for now, I would assert that the latest signal change is more of a tactical move as opposed to a strategic one. That is, depending on how the stock market performs over the next few weeks, this could very well be a short-term move – so don't be surprised if we go long again in our DJIA Timing System as soon as next month. I will definitely keep all our subscribers up-to-date in the meantime.
As for whether this latest signal could be “wrong” – sure it could be. As I have mentioned many times before, trading the stock market involves dealing with probabilities and trying to come up with a set of probabilities to deal with each possible scenario using a limited amount of information – and so this will never be a perfect science. However, given that the Dow Industrials has risen 24 out of the last 27 trading sessions (with the streak ending on May 7th and tying the summer 1927 streak) and given that the Dow Industrials is now 9.6% above its 200-day moving average (in other words, the Dow Industrials is now very overbought in the short-run), my guess is that we will at least have enough of a consolidation period for us to go back long again should I be proven wrong. Subscribers please stay tuned.
Henry To, CFA