Letter – 2003 Q4

The stock market had a good year, and our year was [almost; nearly; more than] twice as good.  For the year, the S&P 500 Index rose 26.4%.  Your account gained xx% (using the standard geometric average measure) or xx% using the internal rate of return measure (which adjusts for the timing of contributions).  We achieved this without being too far out over our skis; our gains were in many industry groups and not a result of being too heavily weighted in a lot of risky tech stocks and the like.  Your return includes all dividends received, and reflects the deduction of my 0.x% annual fee.  The index returns do not reflect dividends, because I don’t have total return figures for the indexes readily available.

For additional reference points, the Dow Jones Industrial Average gained 25.3% and the Nasdaq Composite Index rose by an incredible 50% (which compares to +39.6% in 1998 and +85.6% in 1999).  I don’t use either of these indexes as a benchmark because the Dow is too narrow, and the Nasdaq is much more volatile than I would like for a portfolio.

As indicated in my letter of November 4, I pulled in the reins a bit late in the year.  I thought prices were running too far ahead of earnings – in general and in some stocks that we had owned for a relatively long time.  While the market remained strong, certain stocks which had run well ahead of the market all year did indeed cool off in December (leading to our first month of underperformance since February) while other stocks that had been laggards all year (eg, Ford, GM, oil stocks) had their day in the sun.  It’s unusual to have autos and oils moving in tandem, but December was an aberrational month in several respects, which made it tough to beat the indexes.  While many laggards rallied, certain of our biggest winners for the year sold off in December.  For example, at its best levels we had a four-fold gain in Nam Tai Electronics.  But as earnings momentum has slowed from over 100% to a mere 35% year-over-year, the stock has slowed too.  It fell from $41 on December 1 to $28 by year-end.  So that was one thing that put a dent in our December and in our fourth quarter performance.

We still managed to beat the market in the fourth quarter; your account was up xx% while the S&P 500 rose 11.6%.  I think I handled Nam Tai reasonably well; I sold about half the position when we had more than tripled our money.  But I left the other half on, in the hope of “letting profits run”.  That’s a reasonable expectation here; Nam Tai’s PE ratio is 32 versus 2003 earnings and about 21 versus estimated 2004 earnings.  Not bad if earnings growth continues at 35% or better.  My model projected a 44% compound annual return for Nam Tai when we first bought it early in the year; the model now projects a more modest 14% — still decent relative to most other tech shares.  For instance, we’ve owned Foundry Networks from a price of $5.20 a year ago to $27.33 at year-end.  Although we’ve sold part of the position, it’s a great example of the reward of holding on tight and using a trailing stop-loss order, which we ratchet up every few weeks.  [We’ve done the same with Cisco Systems, which we bought below $10 a little over a year ago.]  We still own both Cisco and Foundry, even though my model now sees both stocks as quite overvalued.  But sometimes price movements accurately foreshadow higher earnings estimates.

Nam Tai wasn’t the only stock to disappoint late in the year.  Many retailers sold off in December.  Pacific Sunwear doubled this year, but sold off 7.5% in December.  Barnes & Noble, another of our largest holdings, has gained over 50% from our entry point last April, but still sold off in December.  Abercrombie posted disappointing numbers and lost over 15% in December.  Michaels Stores peaked in mid-November and fell about 15% from then to year-end.  And our givebacks on big winners extended to other sectors.  For instance, Inamed (obesity treatment) has been a big winner this year but fell 16% in the last two months of the year.

Not all of our late year losses were due to pullbacks in our big winners.  I put on some positions that have been lousy.  Most notably, I added a position in Fresh DelMonte Produce, a company with a PE ratio of 7 in a nice defensive sector like food.  Well, the the PE of 7 became a PE of 6 when Merrill downgraded the stock on a forecast of lower fruit prices.  Although I am wary of the thin margins in the food business  (Merrill’s analyst knows more about bananas and pineapples than I do), I still think the company is a good value play.

Speaking of Merrill analysts, one of our longest term holdings has been Sunrise Assisted Living.  I bought it a long time ago at about $20.  It would get to $25 or $30, and then a Merrill analyst would trash it based upon what I thought was a faulty analysis of the contribution of real estate transactions to their earnings.  Such transactions have become less important to Sunrise, and the Merrill analyst seems to have recognized this, or in any event has grown quiet – and the stock has finally started to move, ending the year at $38.74.  It remains a core holding.

Other major holdings have continued to chug right along during December.  Nextel has doubled in the past eight months.  I sold about half the position as the stock got rich relative to earnings estimates.  Then the earnings estimates went up.  Back at the end of September, Nextel was expected to earn $1.44 per share in FY ’04.  By late December, that estimate had increased 22% to $1.76.  And the stock continued its climb, from about $20 to $28.  So the market led analysts here.  There is a lesson there, and that has been all the more reason I’ve tried to let our profits run this year, even if we endure some painful pullbacks.  We’ve had such pullbacks in the housing stocks as they’ve marched higher in the last two years, in Nam Tai in December, and in plenty of other situations.  Sometimes the “pullbacks” turn out to have marked major tops in a stock, and in retrospect, we let too much of a profit melt.  But many primary trends last a long time.  Peter Lynch talks about “ten baggers”.  It is worth enduring some tough pullbacks when there is a reasonable expectation that a primary trend in a stock will continue.

Steel was another sector that finally benefited from increased earnings expectations.  We had to be patient, as these stocks took longer to turn than I thought they would.  But US Steel just about doubled in the quarter, rallying from $18.38 to $35.  We own it at $xx, having bought it when it was trading at a mere 3 times peak earnings.  We’ve also had good appreciation in Steel Technologies.  The steel holding that screens the best has done the worst so far – Pohang Steel of Korea.  My model projects a compound annual return of over 20% on Posco, versus a slightly negative projected return now on US Steel (but again, I’m doing my best to let profits run, knowing that earnings estimates are still only half of peak earnings and could trend higher).

The other Korean laggard (and other disappointing value stock) is Korean Electric Power (KEP).  Patience has paid off on stocks ranging from US Steel to Sunrise.  KEP could be another such case.  It is trading at a mere 5 times 2004 projected earnings.  The depressed price is probably due in part to the tensions on the Korean peninsula.  But Merrill Lynch points out that its valuation is lower than during the Asian financial crisis of 1998; this time I’m rooting for Merrill to be correct!  We bought the stock three months ago; it hasn’t moved much.

There are a lot more good stories from 2003.  We bought Tesoro Petroleum at $xx in early ’03, when it seemed that refining margins simply couldn’t get much worse.  I’ve held on to about two-thirds of the position, and the stock has about tripled, ending the year at $14.57.  More recently, we’ve profited from higher tanker rates by owning OMI Corp (a shipping company).  We bought Dicks Sporting Goods at $14.35 in October 2002; it finished at $48.66 in 2003.  [enough accts?] We picked niche sectors of auto manufacturing, buying Noble (laser-assisted assembly lines) and Autoliv (airbags and other safety devices); they have both appreciated substantially in the last quarter.  We did well in boats too; West Marine rallied from $19.05 to $27.50 during the quarter.  We’ve done very well in the materials sector with Universal Forest Products, which gained 32% in the quarter.  Helen of Troy (hair care) is a classic example of a great profit in a rather mundane business.  Although the fourth quarter was pretty flat, we rode the stock from $11.64 [yr end] to $23.14 last year.   [cfc, ]

Even though most healthcare stocks lagged the market this year, we did well in selected areas.  For instance, prescription benefit managers such as Advance PCS about doubled last year.  [We still own Humana from $13.25; it ended the year at $22.85.]  [Sicor (injectable pharmaceuticals) moved sharply higher especially after the announcement of a proposed merger with Teva Pharmaceuticals.]  And while I rarely use mutual funds, I did so in March to provide a one percent exposure to the biotech sector just as a sharp climb was beginning.  And I recently added a position in Martek Biosciences.  I stayed very underweighted in the major pharmaceuticals, whose problems are by now familiar.  But I’ve recently taken a large position in Pfizer.  Revenue growth and earnings growth have started to accelerate (in part due to the Pharmacia acquisition), and its pipeline has promising new drugs including one for congestive heart failure.  We bought the stock at 40% below its peak; Value Line opines that “the stock could at least double in price out to 2006-08”.

Overall Market.  There is more I could say about individual stocks, but let me turn to the market as a whole.  For years, I’ve focused on three major indicators – valuation, earnings momentum, and long-term trend as indicated by the 200 day moving average.  The latter two indicators are clearly bullish, and require minimal discussion.  Valuation is trickier to assess here.

The market’s PE ratio here is fairly high.  Assuming S&P 500 earnings of $55.25 in 2003 and $61.00 in 2004, we have a trailing PE of 20.1 and a forward PE of 18.2 based on the year-end index level of 1111.92.  But when you take the reciprocal of those PE ratios to find the earnings yield, valuations are reasonable in relation to interest rates.   Here is a look at that relationship at some previous market bottoms and tops:

In the last 15 years, the earnings yield on stocks got to be over 200 basis points less than the yield on bonds before a market top.  But two of those three periods marked bubble tops, and it’s not prudent to play for that kind of peak again.  As 1961 and 1968 make clear, valuations did not reach such extremes at earlier market tops.  Indeed, common sense suggests that when the same return is available in a risk-free security, the odds of superior returns from riskier assets grow longer.  So it logically follows that the earnings yield should always be higher than a risk-free security with a comparable holding period.  But the market is not always logical.  Also, a reminder that the spread relationship can change rapidly if rates start to rise – as seems likely later in 2004.  Given the absolute level of the earnings yield, the market is quite vulnerable to any sustained rise in rates.

The bottom line is that although the market is not cheap by this measure, history suggests that there is still room for considerable upside.  Some market historians are sanguine about this, noting that a bull market can become overvalued in its early stages as prices advance and earnings “catch up”.   If the spread matches the 1961 top, the earnings yield could fall to 4.45%, which translates to a PE ratio of 22.5.  That in turn suggests that the market could gain over 20% on PE expansion in 2004.  The downside risk is tougher to quantify, but the market is clearly quite vulnerable to earnings disappointments and/or external shocks.

It is worth noting that Value Line takes a dimmer view of the market’s valuation.  In previous letters, I have cited its measure of “Estimated Median Price Appreciation 3 to 5 years hence”.  That reading has been at 115-125% at the market lows in 1987, 1990, and 2002 (and even higher in 1974 and 1982).  It is now at 40%.  The reading got that low often in 1997 and 1998, a year or more “too soon”.  The previous instance of this measure being so bearish was in much of February through August of 1987.  I treat it as a useful barometer; it doesn’t necessarily predict tomorrow’s weather, but if it gets too low, you know there could be a hurricane.

Earnings momentum has been another useful indicator of the market’s overall condition.  In the last two quarters, S&P 500 quarterly earnings have been about 20% better than the same quarter a year earlier.  That’s about as good as it’s gotten in the past decade.  But history suggests that this rate of improvement is sustainable for several quarters in a row.  Even if the earnings acceleration has peaked, continued earnings improvements are a bullish factor for now.

The 200 day moving average is also bullish.  When the S&P 500 Index moved above its 200 day moving average early last spring, it never looked back.   The index is comfortably above this long-term moving average, but is not so high as to be above it by an unsustainable amount.  The 200 day moving average is now at 994.  Based on historical data, the market is usually ahead of itself if it gets 15% above this moving average (~ 1145 now) and almost always unsustainably high at 19% over (~ 1182).  Obviously, this moving average will rise at least through early 2004.  It is fair to presume that the primary trend is bullish as long as we stay above the 200 day moving average.

In sum, 2003 was a good year – albeit a long time in coming.  We did well on an absolute and relative basis.  It can’t always be this good; markets go down and I make mistakes.  That’s why the regulators make us say that “past performance is no guarantee of future results”.  True.  But we have a logical methodology for stock picking, and reason to believe that there will be good investment opportunities in 2004.  I will do my best to capitalize on them, and I appreciate your continued confidence.

With all good wishes for the New Year,

* Past performance is not necessarily indicative of future performance. Results for individual clients may vary. Results are not audited. Byrne Asset numbers reflect the addition of certain dividends and deduction of all fees. S&P numbers are based on the total return of Vanguard’s S&P 500 Index Fund.