THE POINT AND FIGURE METHODOLOGY —A COMPLETE ANALYSIS TOOL

THE POINT AND FIGURE METHODOLOGY —A COMPLETE ANALYSIS TOOL

Chapter 9

FIXED INCOME INDICATORS

One of the most important themes in the market to keep close track of is the fixed income area. Fixed income is always battling the stock market for investor’s dollars. Fixed income is consid- ered the stable part of an investor’s portfolio while the stock side is considered the volatile side. By combining them, the risk is mitigated considerably. In fact, in our money management di- vision one account we manage is a model that combines stocks with strong Relative Strength (RS) and a bond side of the portfo- lio that makes up about 40 percent of the total. At this writing, the model has not had a losing year going back to 1997 in our back test. The one thing we endeavor to do with respect to the fixed income indicators is to determine if the probability is that interest rates are likely to rise or fall. This is one part of the equation that if we could be dead right on our expectation, we would own the world in very short order. Since we still come to work each day, this is not a perfect science. However, I have never seen any other methods that can bring you closer to the right answer than what we do with our charts. In 1994, the indi- cator we used back then was the Dow Jones 20 Bond Average. Dow Jones maintained the average and we then plotted a Point and Figure chart of it. Eventually, Dow Jones gave up keeping the index and we began to be the keeper of it. As time went on, the data we got became less reliable and we eventually discon- tinued maintaining the chart and picked up another bond index, the Dow Jones Corporate Bond Index, which has been just as ro- bust. Looking at signals in history can give us a grasp of how

279

280 The Point and Figure Methodology—A Complete Analysis Tool

these indicators operate and when the Dow Jones 20 Bond Aver- age was in existence, it was fantastic. When it was on a buy sig- nal. You could darn near take it to the bank that rates would fall and bonds would rise.

Let’s look at just one example of a signal given by the Dow Jones 20 Bond Average. In January 1995, the Dow Jones 20 Bond Average gave a buy signal saying that we would see higher bonds and lower rates. By June, bonds had taken off, following this indi- cator had you on the right side of the equation. Take a look at a story by Suzanne Wooley from BusinessWeek (June 19, 1995):

Hang on: More Surprises on the Way

Bond funds did a dramatic about-face from 1994, when virtually every category of bond fund was in negative ter- ritory. But in the first half of 1995, every category sported positive returns. Only two fund groups—short-term world income and government adjustable-rate mortgage— failed to generate a return of more than 8 percent. Gov- ernment bond funds investing in Treasuries took the top slot, with an 11.5 percent average return, boosted by the powerful rally in the bond market. But many government bond fund managers, having built up more conservative positions in short and intermediate maturities in the wake of last year’s market rout, didn’t participate in much of the rally. World bond funds came in second, gaining 10.2 percent.

I thought for a second how we had been long bonds since Jan- uary 1995 because our bond indicators had given a buy signal. Then I thought about all those high paid, and I mean high paid, bond analysts and traders who missed the whole move. What were they looking at or evaluating that caused them to be totally opposite the trend in bonds? As I have said, we subscribe to the Yogi Berra School of Investing: “You can observe a lot just by watching.” The charts are silent, yet all knowing. All you have to do is watch and have faith. There is an old saying, “Those who know, don’t speak, those who speak don’t know.” The charts don’t speak. They are there for your observation and typically the signals are long term in nature.

Fixed Income Indicators 281

So let’s get into the most important chart we keep on inter- est rates, the Dow Jones Corporate Bond Index, which is what I consider as the modern day version of the Dow Jones 20 bond Average we use to follow years ago. This index is comprised of 96 investment grade issues that are divided into the industrial, financial, and utility/telecom sectors. As well, they are further divided by maturity with each of the sectors represented in the 2, 5, 10, and 30-year maturities. All issues are equally weighted and strict rules for liquidity, rating, and issuers are applied. The index is reviewed monthly and maintained by Ryan Labs. We use the price return value for the index, instead of the total re- turn index, because a total return index will have a positive bias and we feel that evaluating the index itself is more appropriate for our needs of determining strength or weakness in the corpo- rate bond area. Recently, I was called by a broker, a client of ours. He asked me what I thought about interest rates. The first thing I looked to was the Dow Jones Corporate Bond Index to get a good feel for the current interest rate environment. All you have to do is look at Figure 9.1 and you will see that in October 2005 the index not only gave a sell signal but also penetrated the long-term Bullish Support Line. That was the moment it was clear that interest rates were going to rise and bonds decline on a long-term basis. One chart spoke volumes to us months ago. Here we are in June of 2006 and rates are still rising. When will they stop rising? Once this index bottoms and gives a Double Top buy signal.

As you can see in Figure 9.1, the Dow Jones Corporate Bond Index (DJCORP) is longer term in nature—this chart goes back to 1996. If we want to get a nearer term outlook, we can adjust the box size down. Recall that the smaller the box size, the more movement and thus the shorter term in nature. When evaluating the DJCORP chart, we look at buy and sell signals as well as High Pole and Low Pole Warnings. That is, when the DJCORP is on a buy signal, it suggests higher bond prices and thus lower rates. Conversely, when this chart is on a sell signal it suggests lower bond prices and higher rates. A High Pole warning is defined as a “pole” of X’s up more than 3 boxes than the previous X column. The warning comes when the index falls more than 50 percent of the number of boxes up in the pole.

282 The Point and Figure Methodology—A Complete Analysis Tool

Figure 9.1 Dow Jones Corporate Bond Index.

Let’s say for instance that the pole of X’s up was 10. If the DJCORP were to fall 6 O’s, that would cast a shadow on the last buy signal. The Low Pole Warning is the exact opposite with the warning, of potential improvement, would be when the re- versal up to X’s exceeds more than 50 percent of the O’s down in the pole. The DJCORP index is robust in its design and diver- sification and the ability to follow its components gives us a heads up with regard to its underlying strength or weakness. The 2-year, 5-year, 10-year, and 30-year charts that comprise

Fixed Income Indicators 283

this index are all available on the DWA web site using the fol- lowing symbols: DJCORP02YR, DJCORP05YR, DJCORP10YR, DJCORP30YR.

In addition, we also follow its sector components the indus- trial, financial, and utility sectors. These sectors are also broken down into 2-year, 5-year, 10-year, and 30-year maturities and we follow these charts as well. These maturities for each of the sec- tors are combined to create a total index. Evaluating these com- ponents helps to shed some light on the status of the DJCORP as well as these broad industries. For example, in early February 2002, the utility/telecom component went to a sell signal across the board and stayed this way until mid-August 2002. I like to go back and evaluate this time in the market because it was a water- shed of stock declines. There have actually been several of these periods since the first edition of this book including a four-week period of May—June 2006, when two trillion dollars was lost in the market on a global basis. Let’s go back to 2002 for a moment. The weakness in the utility component in July 2002 was sub- stantial, especially when compared to the industrial and finan- cial components, and it was this weakness that lead to the sell signal in the DJCORP in July 2002. It also was an indication of the marketplace’s view of the creditworthiness of this sector. It is interesting now we look back to 2002 that the long-term trend line did not break on this reversal and sell signal in 2002. It con- tinued its rise to its top of 116.25 where it began to produce a long-term series of lower tops. Each time the bond index rose, it somehow was unable to get as high as it was previously. It shows that interest rates were in a battle between higher rates and lower rates. It’s a time where the economy is in transition, in this case to higher rates. Notice how the index produced lower tops right down to the trend line break and sell signal in October of 2005. This was the point it was clear that we could expect a long-term series of rate hikes by both the fed and the market it- self. It was not until May of 2006 that investors began to believe this change was happening. In one month, global investors lost $2 trillion in equity. This bond index clearly showed that rates were on the long-term rise with the trend line break. The next piece of the puzzle was to figure out how this would impact equities. This is why we have the Bullish Percent Indexes you have already

284 The Point and Figure Methodology—A Complete Analysis Tool

read about to signal when the equity play is either on or over. It is interesting that in November 2005, we put the offensive team on the field for equities while rates were rising. It is not unusual to see rates and equities rise at the same time. This remained the case until May 2006. The run in equities was fantastic dur- ing that period but when the NYSE Bullish Percent went to de- fense the play was over and the down move happened so fast that most investors were caught unaware. Our clients were not, however. We rang the bell in time to play defense before the car- nage set in.

In addition to the Dow Jones Corporate Bond Index (DJCORP), we have a number of indicators to measure the treasury markets. Sometimes the best thing to do is just go to the source. For us, that means looking at the supply-and-demand picture. Just as we can chart the supply-and-demand relationship for equities, we can also do it for the fixed income market. We can chart bond futures, fixed income Exchange Traded Funds (ETFs), yield indices, and in- verse bond funds. When bonds are rising, we will see positive pat- terns for both the bond futures and fixed income ETFs. If bonds are rising, that means that rates are falling so conversely we will see negative patterns from the yield indices, which measure rates themselves, and from the inverse bond funds. The exact opposite picture will develop when rates are rising. That is, we will see the yield indices showing positive patterns and a positive trend along with the inverse bond funds. If these charts are positive, by defini- tion, the chart patterns of the bond futures and the fixed income ETFs will be giving negative Point and Figure patterns and gener- ally trading below the Bearish Resistance Lines. Figure 9.2 is a summary of the most frequently used Point and Figure chart by the DWA analysts for you to keep and periodically review to get a good grasp on the interest rate environment we are currently in.

As you can see from Figure 9.2, we look at bond futures, yield indices, fixed income ETFs, and inverse bond funds to get a compos- ite picture of the trend of rates—not the financial media. Just as we presented in seminars and training classes, magazine covers can be great contrarian indicators. For example, in June 2004, the Econo- mist cover said the following; “Back to the 1970s: Inflation Re- turns, Worldwide.” The cover also had a very dramatic picture— a large platform, high-heeled shoe, reminiscent of the 1970s. But

Fixed Income Indicators 285

Figure 9.2 Interest Rate Sensitive charts.

the Point and Figure chart at the time shows that the 10 Year Yield Index (TNX) had given a Double Bottom sell signal at 45.50 in July 2004, and this began a series of lower tops and lower lows for the TNX over the next year. In fact, during the next year, the TNX fell 13 percent, not rose. The Point and Figure chart is an excellent re- source on the supply and demand nature of rates just as it is for eq- uity prices. Let’s look at another example.

In Figure 9.3, we compare the 30 Year Index (TYX) beside the iShares Lehman 20+ Year Treasury Bond (TLT). It shows in Sep- tember 2004 that long-term rates had just violated a Bullish Sup- port Line and turned to a negative trend. This tells us that rates are on I-95 South. Not surprisingly, a look at the TLT chart shows that it had just violated the Bearish Resistance Line and moved to I-95 North. Understanding the interest rate picture can have im- plications across everything you do. In your 401k, it might prompt you to take a closer look at the bond funds available for positive pattern.

This trend stayed in effect essentially a year, until October 2005 (see Figure 9.4). In October 2005, the 30 Year Yield Index moved from being on I-95 South to I-95 North. This means that long-term rates are in an intermediate term uptrend and consequently, long- term bonds are in an intermediate term negative trend. When eval- uating fixed income ETFs and yield indices, we like to follow the trend primarily but there are certainly times in which the support

286 The Point and Figure Methodology—A Complete Analysis Tool

Figure 9.3 Interest Rates versus Fixed Income ETFs: September 2004.

or resistance line is too far away. In this instance, either a subse- quent line should be drawn, or more emphasis is placed on the sig- nal than the trend. We should note here that the iShares are ETFs and can be bought and sold just like stocks. The only way to play the Yield Indices is through the options market with calls and puts. Again, being aware of the interest rate environment can be helpful for a variety of reasons. Let’s say you were purchasing a house. Knowing that interest rates were rising might suggest you get a fixed rate mortgage instead of an adjustable rate.

In investing, there are numerous truisms that don’t in fact hold a lot of water. They are truisms that are sometimes true and sometimes not. For instance, it has been said that falling rates are good for the stock market. However, during the year 2000, the 30 Year Yield Index fell 15 percent and the S&P 500

287

Figure 9.4 Interest Rates versus Fixed Income ETFs: October 2005.

288 The Point and Figure Methodology—A Complete Analysis Tool

fell 10 percent. In 2002, the 30 Year Yield Index fell 12 percent and the S&P 500 fell 23 percent. Conventional wisdom would say that in a rising interest rate environment, the long end of the curve would be the most volatile and thus the most risky. This isn’t always the case. The 5 Year Yield Index moved to a buy signal in 2004, well before the 30 Year Yield Index did in October 2005. The 5 Year Yield Index is up substantially more than the 30 Year Yield Index since April 2004, and thus the short end bond ETF, the iShares Lehman 1 to 3 Year Bond Fund, has underperformed the long end, the iShares Lehman 20+ Year Bond Fund.

It reminds me of the following Zen story (source: http://www .rider.edu/∼suler/zenstory/zenstory.html):

After 10 years of apprenticeship, Tenno achieved the rank of Zen teacher. One rainy day, he went to visit the famous master Nan-in. When he walked in, the master greeted him with a question, “Did you leave your wooden clogs and umbrella on the porch?” “Yes,” Tenno replied. “Tell me,” the master continued, “did you place your umbrella to the left of your shoes, or to the right?” Tenno did not know the answer, and realized that he had not yet attained full awareness. So he became Nan-in’s apprentice and studied under him for 10 more years.

In the end, the market will do what it wants to do. Our job is to listen and be aware of the market’s movements by examining the charts and then taking the appropriate actions.

At this point in the book, you have really learned all of the concepts there are to know. It’s just a matter of applying those concepts to different areas of the market. For instance, let’s take the RS tool that is applied to individual equities and apply it to the fixed income markets. Using this dynamic tool, we can compare the fixed income markets to the equity markets and within the fixed income markets, we can find out those areas with the weakest and the strongest relative strength. For this comparison we will use the iShares Lehman Aggregate Bond Index (AGG) to the S&P 500 (SPX). The iShares Lehman AGG is

Fixed Income Indicators 289

a measure of a broad spectrum of different types of bonds as well as maturities and quality so it is a good overall picture of the bond market. In the RS chart we will use a smaller scale than the traditional equity chart so we can still look at the buy and sell signals for guidance as to when we want to overweight fixed income in the portfolio. The gray shaded areas are when this RS chart is on a sell signal and the denominator, the S&P 500, should outperform. The lightly highlighted area is when the RS chart is on a buy signal and suggests the numerator, the AGG, should outperform. As with most RS charts, once a trend is established, it tends to last several years. And you see that the RS chart has aptly guided us with respect to the weighting of fixed income versus equities in the portfolio. From Novem- ber 2000 to July 2003, this RS chart was on a buy signal and the AGG was up 23 percent, while the S&P 500 was down 25 per- cent. Once the chart moved to a sell signal, one would have lightened up their fixed income positions in the portfolio and added to their equity positions. Since moving to a sell signal, the AGG is down about a percent while the SPX is up almost 30 percent (see Figure 9.5).

Another concept that you’ve learned about that can also be applied to our fixed income fund analysis in the same manner as analysis on the equities funds is the Bullish Percent. We chart the bond fund prices using the Point and Figure method and then use the basic Bullish Percent method where we measure bond mutual fund charts for buy signals just as we do for stock and eq- uity funds. Figure 9.6 is an example Bullish Percent for the group All Fixed Income funds (BPMU99), which covers over 3,000 dis- tinct bond portfolios. This indicator tells you the percent of the entire fixed income fund universe that is on a buy signal on their default trend chart. (We also have Bullish Percents on subcate- gories of the fixed income universe.) The back and forth action from X’s to O’s is interpreted the same way that you would inter- pret a stock bullish percent. In Figure 9.6, you will notice the BPMU99 chart reversed into O’s in October 2005. That was also when the trend of the 30 Year Yield Index we discussed earlier moved to I-95 South. Do you see how these pieces of the puzzle are coming together?

Figure 9.5 Fixed Income versus Equities. 290

Figure 9.6 Bullish Percent for All Fixed Income Mutual Funds. 291

292 The Point and Figure Methodology—A Complete Analysis Tool POINTS AND FIGURES BY DORSEY, WRIGHT MONEY MANAGEMENT

Anything you can do, I can do better. Those words from an old Irving Berlin song encapsulate the current status of the battle of man and ma- chine. According to a recent article in the New York Times (“Maybe We Should Leave That Up to the Computer,” July 18, 2006), professor Chris Snijders of the Eindhoven Institute of Technology is convinced that com- puter models can do a better job making decisions than humans. He even issued a challenge to any company willing to have its humans compete against his computer models.

Scientists have known for a long time that mathematical models gen- erally perform better than humans on a variety of complicated tasks. Ac- cording to the article, studies have shown that models can better predict the success or failure of a business start-up, the likelihood of recidivism and parole violation, future performance in graduate school, various med- ical diagnoses, picking the winning dogs at the racetrack, and credit scor- ing. You can add to that list playing chess, as no human player, including a couple of World Champions, has been able to defeat one of the silicon monsters in a match for several years.

Yet, each time I see one of these articles, I always think to myself that they haven’t got it quite right. The articles always somehow imply that computers make better decisions than people, without fully incorporating the notion that humans simply wrote a computer program to reflect the knowledge that they already possessed. What’s really happening is that computers are much more efficient at executing the decisions that knowl- edgeable humans might make under ideal circumstances.

In the case of the early chess programs, continuous improvement from feedback was critical. The first generation of programs played horri- bly and was beatable by low-level players. As programmers got more guid- ance from chess masters and learned better ways to reflect the knowledge that expert human players already had, the programs got stronger and stronger. Humans, as many of us know from experience in the office, often don’t take feedback well, and/or don’t incorporate it going forward! But chess programmers used the feedback and now a $50 CD that will run on any desktop can wipe the floor with a regular player.

The second big advantage computers have is their consistency in making decisions. Really, it’s just that they are consistent and humans tend to be inconsistent. Computer models do not have psychological hang-ups stemming from childhood trauma. Models do not get emotional, tired, hungry, or cranky, but people do. This is very apparent in a chess pro- gram. You can set up an inferior position for it to play from, but rather than panicking or giving up, it will continue to grind away and often will

Fixed Income Indicators 293

be able to draw or even win against a human. When money is involved, people definitely get emotional, so there is a big advantage for computers in financial modeling.

The traditional knock on computer modeling is that it can’t recognize when conditions have changed. (Actually, humans aren’t very good at that either!) If the model doesn’t adapt as conditions change, it can be thrown for a loop when the underlying data distribution changes. Developing a model that is adaptive is important to overcoming that hurdle.

We tried to utilize the advantages of computerized modeling and to minimize the disadvantages when we developed the Systematic RS portfo- lios. We feel the consistency of executing the strategy improves the perfor- mance. To minimize the problems, the stock selection is based on relative strength, the best adaptive tool we know. The portfolios are designed to change as market themes change. It might just be the tool you need to move your business to the next level.