Stock-Market-Observations

Short Range T Update for Friday February 6 2004

The Short Range T picture is unfolding normally in the chart below with the blue volume oscillator rebounding on Friday from a very oversold condition. Click on the image for a larger view.

It is too early to think about a new T because the market correction is too short time-wise and as a minimum a rally that fails is needed to set up the next T. In other words the current rally has limited upside potential and will likely fail in due course. The market would normally be expected to fall to new lows in a few weeks and at that time conditions for the formation of a new T would be much more likely. I think we have plenty of time to work out the details out this upcoming T, so I will delay some of the questions until another week or two has passed.

I was also interested in the Mutual Fund Cash Percent Assets chart above which was published at Jim Sinclair’s Web site over this weekend and shown here courtesy of its original source http://www.contraryinvestor.com also listed in my web sites below. I believe both these site are worth viewing and would check them out for “forward looking” views that I think have stood the test of time, but are often overlooked by trend following investors.

With respect to the new T that may have its center date low later this month there appears to be a problem with its prospective rally in the right side. I believe this very long time spanning T will require that we include the total (?-?-?) Cash Build-Up period as part of the time span within the left side of the next T. This will result in a very long spanning T that history suggests may not be able to sustain a market rally throughout its very long right side.

In T Theory any T that cannot support a price trend rally within its right side is term a “bearish type of T” and under adverse circumstances can allow a serious market decline to persist through most of its right side. The Mutual Fund Cash Percent Assets chart conclusions are particularly significant right now because it suggests that cash levels by this measure are at historic lows and cannot provide the fuel to maintain the long rally called for this T’s time nature.

Yet in a few weeks, an new, “yet to be seen”, over-sold condition should technically lead to a 2 to 3 week rebound rally that confirms the time span of a new T by breaking above the green down trend line in the blue volume oscillator. This next rally may last some 2-3 weeks and look strong for its short duration, however this rally should then fail quickly for lack of “real mutual fund cash”. In T Theory this new T should then collapse in the form of a bear type T since the long graphical representation of the Cash Build Up period has no basis in the real world of mutual fund cash (as Husk has commented). Later this year a new oversold condition should result.

A Summary of Basic Valuation Concepts

In the past I have presented basic historical Price to Earnings Ratio Studies from the 1880 to present time in order to suggest that common stocks are over-priced. In this initial “typepad” discussion of this fundamentally important measure of earnings and equity value, I want to expand my discussion the three approaches that look worthwhile to me looking ahead for the next decade. I also want to comment on a new technical trend approach that may be promising and new Advance-Decline T research projects.

The three interesting fundamental approaches to equity valuations are:

A. The simple standard S&P Ratio definition which equals the current S&P 500 price divided by the last 12 months actual net earnings using standard accounting procedures.

B. The more sophisticated Dr. Robert Shiller P/E Ratio definition, which equals the current S&P 500 Price, divided by the 10-year average of these same standard earnings.
C. A more recent, and I think very innovative, Dr. John Hussman conceived P/E Ratio measure. Please read the web article provided in the web links at the lower left for details. Dr. Hussman defined the P/E Ratio as the current price of the S&P divided by the peak annual earnings seen over the last 10 years. This represents an optimistic measure of earnings and circumvents Wall Street’s arguments that P/E Ratios don’t matter when earning are depressed.

The whole point of these three defined measures is to help define the historical “fairness” at which an investor is buying earnings. When P/E ratios are high, as they are now, investors are paying a historic high price for a dollar of earnings. Buying high may work for the near term,
But the historic outlook for the longer term is for a proportionally poor rate of return. Historically buying when P/E Ratios are low has the opposite outcome; that is an above average rate of return is the more likely outcome if one buy low as defined by any of these P/E Ratios.

The problem with all these P/E Ratio concepts is that the current environment is being over ridden by the very low rates of interest as the Fed tries to halt a deflationary downward economic trend. Others have assumed that if the Fed prints money fast enough and in great enough quantity, then the high P/E Ratios normal historical negative connotations can be negated. I am very anxious to test this dubious claim. To do so, I am setting up an extensive database of monthly data.

It includes interest rates along with the usual earning, price and dividend info to get my T Theory interpretation working on “big picture” stuff. These amounts to a bit of work, so along the way, it pays to pick up a few other important technical ideas that could be tested very easily once the basic database has been set up. This will probably take a few months.

Michael Belkin has proposed an alternate technical criterion of the relative valuation of stocks. His moving average technique could make a valuable addition to the basic fundamental P/E ratio measure. It comes out of Michael Belkin’s very long-term technical work covering the last century so it overlaps the more fundamental P/E Ratio studies. He has proposed long-term technical trend measures that might represent an alternate view of over-valuations verses under-valuations. I have not had a chance to verify his claim that 200 week and 200 month moving averages can define key points over the last century but I can easily test it.

His trend technique makes use of unusually long term moving averages (MA), the 200-week MA and 200-month MA of stock indexes like the S&P 500 and NASDAQ index. Since I have the basic blue chip data needed to reproduce these moving averages in my database as part of the P/E ratio Studies, I can study Belkin’s historical conclusions and get a better idea of its validity relative to the conventional P/E Ratio studies that cover the same period.

In a October 2003 published review, Belkin contends that speculative bubbles can be identified by noting extreme upside deviations from the very slow moving 200 week MA, a 4 year MA. Like the Shiller interpretation, he reached the correct conclusion in 2000 for a NASDAQ bubble. Naturally I am anxious to see the 100-year history of his criteria and how it compares to Dr. Shiller’s interpretation of his high Price to 10-year Average Earnings valuation criteria. I do have monthly data for the Dow Industrial Average equivalents from 1800 and I can easily plot the 200-week MA as a monthly equivalent to his basic contention, then vary it to see how his criteria pans out over my 200 year database history.
On the downside, Belkin saw the NADAQ collapse after the 2000 bubble break as finding support at the 200 month MA (a 16.5 year MA). This level proved correct as support at the recent 2002 low. These criteria, if verified historically, would be interesting because none of the three P/E studies concluded an under-valuation low was seen a year ago.

Additionally, Belkin concluded in his October 2003 comments that a bear market recover would peak the NASDAQ near its 200-week MA (2280 at that time). If true, this still gives a little more room on the upside. This might depend on interest rates since there is some reason to believe that any increase in rates would be negative, but he was basically looking for an upside end to the current rally on any further strength.

More work needs to done in this area on my part before any definitive conclusions are justified. However the studies that I can make in my database, which include interest rate data, and P/E ratio data might shed additional light on this technical refinement. Since I already have the basic trend data for the blue chip averages, his interpretation can be verified by a fairly simple extension to my studies.

For this posting I just want to outline these key ideas to keep in mind while I set up the new research projects. As far as I am concerned, the fundamental measures of stock valuation presented by Shiller and Hussman are fine from a static view. They all conclude equities are grossly overvalued. But am also looking to include the more dynamic trend criteria that Belkin has noted for additional refinements.

To further complicate these long-term valuation issues I have also obtained, from a very personal and private friend, the daily Advance-Decline data from 1926 to the present. This is a difficult database to acquire but it is very important to the basic T Theory forecast, which can independently be derived from this Advance-Decline data. I am presently setting up this as a companion study to test A-D Ts from 1929 to the present. This will also become a topic here in a month or so.

A-D Ts have a long and reliable history as noted at my old web site and I am anxious to understand this much long history of the breadth index and how it interprets the current strength in the Advance-Decline line. My preliminary condition is the recent A-D Line strength is significant in terms of the long-range picture and it may provide downside support to the overvaluation conclusions expressed by Belkin and the three conventional P/E Ratio measures.

Nothing is certain as yet, but a long term trading range may be the logical outcome. …Terry Laundry