Return to letters menu

Letter dated January 2005 reporting on the fourth quarter of 2004

     After abnormal weakness in 2000-02 and unusual strength in 2003, the stock market produced a more normal return in 2004.  Most of the net gains came from a rally that began in late October.  The S&P 500 Index rose 9.0% for the year, due largely to a gain of 8.7% in the fourth quarter.  Your account gained 9.0% in the fourth quarter, and 10.6% for the year.  These figures reflect the addition of all dividends and the quarterly deduction of my fees in your account, and measure simple price appreciation in the S&P 500 Index.

     There were a lot of cross-currents and fewer sustained trends this year, which made it difficult to beat the market.  It was also a year in which I was disappointed in certain stocks that seemed promising, and missed rallies in stocks which had seemed overvalued to me.  Nevertheless, we had a number of success stories and an overall result for the year which outpaced not only the S&P 500 Index, but also the Nasdaq Composite Index (+8.6%) and the Dow Industrials (+3.1%).

     Pfizer embodies the frustrating part of the year as well as any stock.  The stock had fallen from near 50 back in 2000 to the mid 30s.  It was widely viewed as the best of the pharmaceuticals, by experts ranging from Goldman Sachs to Value Line.  It was on of our larger holdings until late in the year, when the questions about Celebrex surfaced in the wake of Merck’s Vioxx disaster.  Pfizer dropped from $35.07 to $26.89 during the year.  Yuck.  We had some other disappointments in the healthcare field earlier in the year, but caught the year-end rally in the health insurance stocks well with Aetna and Wellpoint.  Our biggest position of the year was in Johnson and Johnson, which rose from $51.37 to $63.42, in part because money allocated to the pharmaceutical sector had few other safe havens (“safe” because JNJ is diversified into consumer products and medical devices).  J&J and Pfizer seemed like equally good plays at the beginning of the year, which is a good illustration of how seemingly similar risk-reward situation can produce very different results.

     Technology was another sector with widely diverging results.  Many of yesterday’s star performers disappointed.  We bought no shares of Cisco this year, and the stock fell from $24.23 to $19.32.  We liquidated our remaining shares at an average price about $24.  Intel was similarly uninspiring, dropping from $32 to $23.39 for the year.  Our superstars of last year, Nam Tai and UT Starcomm, both moved lower this year as well.  There were a few shining stars.  We bough Cree Inc, which makes light-emitting diodes for TV and computer screens and items such as digital clocks, in July at $22.03.  It ended the year 82 percent higher at $40.80; we still have a good position even though we took some profits along the way.

     Some of our best “tech” investments of the year were in cell phone companies in the developing world.  Cell phones usage has grown much more rapidly overseas.  According to Forbes, only 57 percent of the US population uses wireless phones.  By comparison, in Hong Kong there are 105.75 mobile subscribers for every 100 inhabitants.  In Taiwan, there are 110.  It is a good bet that cellular companies in emerging economies such as Russia, Turkey, Indonesia, and the Philippines will grown quickly; we have invested accordingly.  You own a diversified portfolio of stocks in those countries; my stem projects a compound annual return of over 20 percent on each of these stocks.

     We missed a few big moves in technology companies such as Ebay and Yahoo because they seemed overvalued relative to estimated earnings early in the year, but then they exceeded earnings estimates and got even richer.  Twelve months ago, analysts thought that Yahoo would earn 26 cents per share this year.  My system projected a negative return based on the 26 cent estimate and an expected earnings growth rate of 27.5 % over the next five years.  In fact, Yahoo came in 30 percent above the early estimates in 2004, and has grown earning s at an astounding compound annual rate of 95 percent over the past two years.  I simply didn’t expect such growth, and was unwilling to bear the risk of getting clobbered if already aggressive growth targets were missed, so we did not participate in Yahoo’s 67% rally.  A very similar analysis applies to Ebay, which has grown earnings at a compound annual rate of 91% over the past four years.  Needless to say, those growth rates are at the outer boundary of historical earning s growth data.

     A main reason that the bear market was so bad in tech shares in 2000-02 is that there is a double whammy when earnings growth targets are adjusted downward; both the earnings estimates themselves drop, which in itself puts downward pressure on the stock price, and the P/E ratio usually contracts at the same time, compounding the downward pressure.  For every Yahoo and Ebay, there tend to be too many of the latter cases.  Even so, I’m disappointed that I missed two of the great success stories of the year.

     When you hear people say “it is a market of stocks”, it is because earnings growth rates vary much more than one would intuitively imagine.  Over time, earnings growth rates vary much more than one would intuitively imagine.  Over time, earnings growth tend to mirror GDP growth in the aggregate, but there are many outliers.  Consider these compound annual earnings growth rates over a three year horizon for the following companies:

All of these stocks have exceeded estimates of earnings and longer-term earnings growth, and most of them have appreciated substantially.  All of these stocks were in your portfolio in 2004.

     These stocks beg the question: why didn’t we have another 50% year?  Well, Pfizer wasn’t the worst news.  We got beat up in AIG when Eliot Spitzer went after the insurers.  We had losses in some smaller tech stocks such as Lexar, Credence Systems and Brooks Automation that had looked promising.  Eresearch had great earnings growth, which halted abruptly and sent the stock into a tailspin due to the double whammy of earnings reductions and PE compression described above.  And then, there were certain shares that I sold too soon, such as Thor Industries or US Steel.  I was also a little too defensively postured coming into the election, which caused us to lag the market in early November.

     Strong historical earnings growth does not guarantee a rise in the stock price.  Boston Scientific has grown earnings at 59.4% compounded annually over the past here years, but the stock traded down from $46 to as low as $32 earlier in 2004 on fears of increased competition in the stent market.   We recently bought it at $35.51, right near the year’s closing price.  Our other recent addition is Reebok, which has earnings growth of 21.1% annually over the past three years.  We paid $37.67 in early November and it ended the year at $44.

     Not all stock prices correlate so closely to such compound growth rates.  An industry such as steel is much more cyclical, so the compound growth rates are erratic.  But we caught the up cycle in steel pretty well this year.  The same is true in energy.  While I wish I had been more aggressive in the exploration and production stocks earlier, we are also cyclical; the compound growth rate above for Yellow Roadway looks particularly good because it is off an unusually low base year.  But we were aware of the good earnings momentum (and cheap valuations relative to that momentum) in the trucking sector, and have been rewarded accordingly in both Yellow Roadway and Arkansas Best.

     Good stocks can go down in bad markets, and bad stocks can go up in good markets.  So as usual, I’d like to turn some attention to the state of the overall market.

     I look at valuation, earning momentum, and the primary trend as measured by the 200 day moving average to assess the condition of the overall market.  Valuation must be considered in light of potential alternate investments, which means considering the relative value of stocks and fixed income securities.  As the current bull market began in March/April of 2003, I wrote that the earnings yield on stocks (ie, the reciprocal of the P/E ratio) was 2.24 times the yield on the five year Treasury note.  That meant that if you were looking at a five year holding period, stocks offered a lot more potential return than bonds.  We invested accordingly, and were well rewarded.

     But recently that ratio has slipped to where the earnings yield (now about 5.2%) is slightly below 1.5 times the yield on the five year Treasury note, which ended the year at a 3.6% yield.  On occasion, that ratio has even dipped below 1.0.  That may be explainable for short periods but is unsustainable over the long term because the increased risk of stocks versus bonds suggests that the potential “yield” on stock investments should higher.  History shows no clear breakpoints; if it did, investing would be too easy.  All we have are guideposts, and they are hardly precise.  But consider the following table using weekly Value Line data from 1994 on:

     There is clearly a downward bias in anticipated six month returns from stocks as the earnings yield/Treasury yield ratio becomes less favorable to stocks.  But there is no clear or consistent demarcation, and some of the best stock returns have come in periods when the ratio would have suggested otherwise.  (Remember, these are six month returns rather than annualized figures).  Nevertheless, based on this simple analysis, one would tend toward greater caution as this ratio slips under 1.50.

     These data should not be considered in a vacuum.  Clearly, the market is anticipating higher rates.  Let’s postulate that by March, the market has a PE ratio of 20 and the five year Treasury has moved up from its year end level of 3.60% to 4%.  That would put our ratio at 1.20, or at a level that is not terribly attractive for stocks.  So we will monitor the relationship between stocks and rates closely.

     Speaking of not looking at things in a vacuum, that is why I look closely at the rate of change of earnings growth as well as expected changes in interest rates.  If we were highly confident that earnings would double in 2005, the ratio at a given point in time wouldn’t concern us too much.  But the growth rate for corporate earnings is expected to decelerate in 2005.  This could put further pressure on stocks at some point; since the market is constantly reflecting revised expectations, it is very sensitive to rates of change.  Corporate earnings grew at about 19% in 2004, but are expected to rise only about half that much in 2005.

     You have to put your money somewhere.  As short-term interest rates rise, bond prices are likely to decline and so total returns in fixed income could be negative or nearly so.  Cash is safe, but the interest on cash may not be enough to keep up with inflation.  So even if the return from stocks is only in the single digits, that may still outpace other major asset classes in the coming year.

     Enough people must feel that way, because the market’s primary trend is bullish as measured by the 200 day moving average.  The S&P 500 Index has been above its 200 day moving average since shortly before Election Day.  I don’t see this as having predictive value, but I do view the long term moving average as a decent perspective on a market’s primary trend.  For instance during the bear market of 2000-02, there were many times where one could have reasonably thought that things would get no worse.  Then they did.  Being mindful of the fact that the primary trend was till bearish was helpful in remaining appropriately defensive.

     My job is to be mindful of the macro environment, but to focus on finding the best possible investments at a giving point in time.  I hope to produce good returns once again in 2005.  I appreciate your confidence, and wish you all the best in the New Year.