Return to letters menu

Letter dated April 2003 reporting on the first quarter of 2003

     In my last quarterly letter, I alluded to the impact of war and/or terrorism on the market and stated that:  “Stock valuations seem reasonable if not compelling….If the market should decline, I don’t think that current valuations will be to blame.”  I think that’s what happened:  in the absence of compelling valuations, the tensions related to Iraq were enough to pull the market modestly lower.  But the huge rally when the war finally broke out may be indicative of pent-up demand. 

     The market was incredibly volatile, and though the week of St. Patrick’s Day was the best up week since 1982, the S&P 500 still lost 3.6% for the quarter.  I have maintained a fairly defensive posture, and so your account lost xx% in this period – beating the market by a fairly respectable margin.  I don’t want to become overly defensive because the mid-March rally showed what this market is capable of doing if the international situation improves. 

     The 200 day moving average has been quite an amazing barometer of the long term trend.  For three years, the S&P 500 Index has approached that level many times but has never been able to sustain a move above it.  In the huge St. Patrick’s week rally, we traded right up to that level and actually closed slightly north of it.  It felt like an unstoppable freight train.  But the next Monday, it plunged 3.5% in a day.  As I write, the market has opened above the 200 day moving average once again.  But whether this finally represents a trend change or yet another failure is not yet clear.  Who knows, it may be much clearer by the time you get this letter.    

     Valuations remain reasonable, though not compelling.  In other words, the P/E ratio is right in the middle of its range of the last 50 years.  The bearish valuation case is that major bottoms in 1987 and 1990 occurred with the P/E ratio between 10 and 11, suggesting there is ample room for the market to fall further.  The bullish valuation case is that the P/E ratio has to be looked at relative to interest rates, and when you compare the earnings yield on stocks to the rate on five year Treasury notes, the market is reasonably inexpensive.  Consider: 

     The bullish case is even stronger if you look at the spread on a ratio basis:  the earnings yield on stocks is more than twice the yield available on five year Treasuries.  That ratio reached an extreme of 2.4 at the 1974 low.  That’s a pretty strong gap in favor of stocks,  as long as those earnings don’t evaporate.  It would be far better if the comparison were just as favorable between Treasury yields and dividends, but that is not the case (due in part to how our tax code penalizes dividends). 

    There are analysts who say that the market’s P/E ratio is still about 30, but that way of measuring earnings takes into account a few major accounting write-downs that have to do with post-merger GAAP accounting for goodwill and very little to do with real operating earnings.  This is worth explaining in a little more detail.  Let’s take AOL for example.  In January, the company posted a quarterly loss of $10.04 per share by announcing a $45 billion write-down of assets.  Did AOL, whose entire market capitalization is $50 billion, really lose $45 billion in one quarter?  No!  They simply adjusted their books to reduce the stated value of some of their “intellectual property”.  But some market analysts, who use GAAP rather than operating earnings, subtracted that $45 billion from the total earnings of the other S&P 500 companies in calculating overall earnings and the resultant P/E ratio.  Although this is technically accurate GAAP accounting, it is also a place where GAAP technicalities obscure an accurate picture of earnings from current operations.

     The Value Line approach more realistically reflects operating earnings.  Each week, they report “the median of estimated price-earnings ratios of all stocks with earnings (emphasis added).  That number was 14.9 at the end of March.  While that number may not reflect the hit to earnings from executive stock options,  it is still a pretty reasonable measure.  Plus Value Line’s historical data is readily accessible, so it is easy to compare periods as I did above.  Sorry to be long-winded here, but you read so many different  accounts of the market’s P/E ratio in business articles that I thought it was worth some elaboration. 

     Some people believe that the market will not bottom until we’ve retraced the entire bubble of the 1990s.  We can conjure up any number of bogeymen to make that case.  Overextended consumers.  Government deficits will trigger higher interest rates, and that will hurt the market.  The trade gap will continue to widen, causing a run on the dollar, and sales of US assets by foreign investors.  Terrorism.  Iraq.  North Korea.  Real men want to invade Teheran.  Etcetera.  It’s hard to judge when these are truly negative factors, and when they are simply part of the proverbial “wall of worry” that a bull market climbs.  Just for perspective, consider some other recent bogeymen:  Ravi Batra’s best-selling “The Great Depression of 1990”  (1985), the ’87 crash, Persian Gulf crisis (1990), market plunge on Gorbachev ouster (1991), Fed raises interest rates six times (1994), impeachment (1998), etc.  Since I can’t divine the geopolitical future and since there are good businesses even in a bad economy, my energy is better spent looking for promising investment opportunities.

     We did a decent job of that in the first quarter, and outperformed the market by a respectable amount.  But as usual, there were certainly some regrettable investments.  We did well betting on a turnaround in the oil refining cycle, buying Tesoro Petroleum at $4.69; it ended the quarter at $7.40. Also in the energy sector, we did well in ConocoPhillips, an oil major which has little exposure to the Persian Gulf region.  We bought the stock late last year, and it rallied 10.8% for the quarter as crude oil prices rose. 

     It is worth noting that the Nasdaq, which led the bear market decline, has recently outperformed the S&P 500.  In fact, it has been above its 200 day moving average for a month.  We gained from a turnaround in EMC stock, which rallied 17.8% for the quarter.  Some other tech stocks showed signs of life, notably Flextronics (+6.5%) and Intel (+4.6%).  Open Text (+21.3%) produces knowledge management collaborative software and is a newer tech holding that has done well.  Oracle was essentially flat. 

     Some of our best gains were in conservative stocks with high dividends.  For instance, we got 14.3% appreciation in PSE&G shares, which also have a dividend yield of 5.9%.  Similarly, old reliable Thornburg Mortgage gave us modest appreciation of 2.6% while sporting a dividend yield of 11.3%.  Similarly, Mack Cali rose by 2.2% while providing a yield of 8.1%. 

     We continued to profit from the robust housing sector.  Shares of Hovnanian appreciated by 9% for the quarter.  We’ve also done well in shares of banks that emphasize mortgage lending.  It’s interesting that as pundits continue to speculate about a peak in the housing market, Warren Buffett steps in and buys Clayton Homes.  I think the stocks we own are equally good values. 

     Obviously, it wasn’t all good news.  Cisco has just been stuck in a range.  Maytag seemed like a good investment, given its valuation and all the activity in the housing sector.  But it lost a third of its value after warning on earnings.  Michaels, which has been an old reliable, finally cracked and shed 20% of its value.  We still own it, and sense that their slowdown may have more to do with short term difficulties in replenishing inventory than with a falloff in demand.  We moved a little too soon in buying Charles River Labs (pharmaceutical product development) just under $30; it ended the quarter at $25.52.  And despite record earnings, Universal Forest Products fell 27% in the quarter, in part due to fears that health claims stemming from treated lumber will be a new avenue for tort lawyers.  I think that’s probably far-fetched, but litigation fear has kept the lid on a lot of stocks.  I thought Webex’s audio and video combination was the next great wave in interactive communications, particularly in a world with declining business travel.  But solid gains melted all in a day, and I’ve cut our position substantially.  [some sold all]

     We added some promising new positions.  For example, we bought xx shares of UT Starcom at $21.05.  The company provides telecommunications equipment in the Chinese market.  In 2002, its net sales rose 57% to $981.8 million.  Earnings per share were $0.94 in 2002, an 80% increase from a year earlier.  When we bought in, the consensus among analysts was for earnings of $1.26 in 2003.  Assume for a moment that the company can hit that and then maintain earnings growth of 20% for five years.  If it does so, it is reasonable to assume that the P/E would be at least 20 after five years.  If we buy those assumptions, we can calculate that the stock would provide us with a compound annual return of about 18% going forward.  There are many risks, such as the possibility of a competitor coming in with a far better, next generation product.  But based on what we know now, this seems like a very promising investment.  And you get a sense of the earnings analysis I do before committing capital. 

     I’ve done similar analysis to project similar returns on other recent buys.  For instance, we initiated a position in Headwaters, a leader in producing alternative energy products.  Countrywide Credit does mostly prime credit mortgage loans, and is a new addition to our holdings in the financial sector.  We about doubled our money in Owens Illinois last year, and with the stock’s recent decline, have bought a big dip which so far has become a somewhat bigger dip.  Some of the insurers seem cheap.  We bought a newer company with no skeletons (but a less diversified portfolio) called Montpelier RE at $xx, and more recently added WR Berkley.  In each of their lines, Berkley had net-premiums-earned gains of at least 75% last year.  Again using past and projected earnings data, my model projects a compound annual return of about 18% on this stock too. 

     In sum, I think we’re again running with a portfolio that has a good chance of outperforming the market.  And given valuations, the market has a good chance of outperforming other investments and its own recent history.  Keep the faith!  And thank you for your continued confidence.