Stock Star Rating Performance Update

As an investor, I have always valued candor very highly when assessing both CEOs and portfolio managers. So, I will be as candid as possible in reviewing the overall performance of our stock picks in 2007: We had a bad year. Our 5-star calls underperformed just about any relevant benchmark, and they also posted poor absolute returns.

What went wrong? Broadly speaking, two themes hurt us. First, we underestimated the amount of bad lending behavior during the housing bubble, which had repercussions across our coverage universe--from homebuilders to mortgage lenders to bond guarantors and beyond. Second, the spillover effects of a weak real-estate market on consumer spending were much greater than we anticipated, causing us to re-evaluate the prospects for many retailers and restaurants.
 

Calendar-Year Trailing Returns
Summary	Buy at 5, Sell at 1	Buy at 5, Sell at FV	Buy at 5, Sell at 3	S&P 500	S&P 500 Eq-Weight
2007	-4.5%	-8.4%	-8.7%	5.5%	1.5%
2006	17.4%	20.2%	23.3%	15.8%	19.0%
2005	7.6%	9.4%	9.2%	4.9%	7.4%
2004	21.4%	26.7%	32.8%	10.9%	17.1%
2003	42.3%	47.2%	44.3%	28.7%	44.8%
2002	-32.8%	-37.0%	-36.8%	-22.1%	-16.2%
2001*	2.0%	3.9%	5.0%	-11.9%	-1.2%
Data from Abacus Analytics, through 12-31-07. * Morningstar began rating stocks on 08-06-01.

Our poor 2007 returns drag down our trailing numbers, as you can see in the table below. Although we are still ahead of the S&P 500 over the long term, our lead has been cut by quite a bit.

Time-Weighted Returns
	Buy at 5, Sell at 1	Buy at 5, Sell at FV	Buy at 5, Sell at 3	S&P 500	S&P 500 Eq-Weight
Trailing 1-Year	-4.5%	-8.4%	-8.7%	5.5%	1.5%
Trailing 2-Year	5.7%	4.7%	5.9%	10.5%	8.4%
Trailing 3-Year	6.5%	6.5%	7.2%	8.6%	8.3%
Trailing 4-Year	10.1%	11.3%	13.1%	9.2%	10.4%
Trailing 5-Year	15.6%	17.7%	18.8%	12.8%	15.9%
Since Inception	5.7%	5.9%	6.6%	4.9%	10.1%
Data from Abacus Analytics. Data through 12-31-07.

Let's start with housing.

With the benefit of 20/20 hindsight, I think it's clear that much of the unsustainable demand and uneconomic lending that occurred during the housing boom can be traced back to the widening gulf between lenders and borrowers.

Way back in the days of yore, a real estate lender assessed the value of a piece of property as collateral, made a judgment about the borrower's ability to repay the loan, and then held that loan for a lengthy period of time. If the property's value was misestimated, or the borrower turned out to be a poor credit risk, the lender suffered the consequences. The advent of  Fannie Mae (FNM) and  Freddie Mac (FRE) made it easier for lenders to sell loans into the secondary market, but the standards imposed by those two institutions on the kinds of loans they would purchase maintained some degree of oversight on the market.

But as the demand for nonconforming loans--that is, those outside Fannie and Freddie's guidelines--increased, lenders were happy to step up to the plate, make the loans, and wash their hands of the loans' future performance by selling them to investors. This latter point is crucial, and a recent paper from the University of Chicago puts some analytical meat on a point that seems intuitively clear--when a lender has little economic stake in the future performance of a loan, he or she will have little incentive to make loans that are more likely to be repaid.

The bottom line is that poor lending and underwriting decisions were being made in many, many parts of the mortgage market, and we did not anticipate the consequences of those decisions. These decisions created unsustainable levels of demand for real estate, which made the downturn that much worse for homebuilders, and they created large losses in many financial institutions as loan performance worsened.

That's the general backdrop. On a more specific level, we entered 2007 with fairly downbeat forecasts for most of the financial services companies we cover, expecting that the benign credit environment of the previous few years would deteriorate substantially. In fact, the average price/fair value ratio for the financial services sector was 1.10 at the beginning of 2007, meaning that we thought the sector was about 10% overvalued, and only 4% of financial services companies were rated 5 stars.

However, we were expecting credit metrics to revert back to long-run historical averages, and it now looks like they'll shoot way past that level. We also underestimated just how much of the financial services world was tied to mortgages in one way or another, which meant that we did not anticipate the massive security write-downs that have plagued the largest banks and caused the bond insurers to implode.

Finally, as is often the case with levered companies that start experiencing liquidity troubles, the market can quickly shift from valuing a company as a going concern to valuing it as a collection of assets and liabilities that may be worthless. We were behind the curve in recognizing this change for some of our worst performers--at a certain point, market perception can become reality if it impairs a company's ability to raise capital.

Our poorly timed bullish calls on homebuilders reduced the performance of the Buy at 5, Sell at Fair Value portfolio by 3.8 percentage points in 2007, a figure that rises to 8 percentage points once you factor in the impact of our ratings on mortgage-related financial services firms. The portfolio would have been roughly flat absent these calls.

Consumer stocks also hurt our performance in 2007, though to a much lesser extent than housing and mortgage-related stocks. On a very big-picture level, consumer spending tends to follow employment, and the relatively strong job market throughout 2007 lulled us into a sense of false security regarding consumers' likely reaction to the declining real estate market, and so we rated many retailers and restaurants 5 stars as the sector rolled over in the second half of the year. The picture has turned decidedly worse over the past few months, and although we still think many consumer-related stocks are excellent bargains at the moment, we did reduce a number of our fair value estimates in early 2008. We turned bullish too early, generally speaking.

On the bright side, we had some excellent stock-specific calls that helped our performance.  Monsanto  and  Mastercard (MA) continued to be big winners for us, both more than doubling during 2007. We still love both of these wide-moat businesses, but their shares are only fairly valued at the moment. Speaking of good businesses, our bullish stance on for-profit education stocks paid off nicely in 2007, with  DeVry (DV) and  Apollo  each rising about 80%, and  Strayer (STRA) popping more than 60%. Steel stocks also helped our performance, with Brazilian steelmaker  Gerdau (GGB) rising more than 70%, the newly merged  ArcelorMittal (MT) jumping almost 90%, and low-cost minimill operator  Steel Dynamics (STLD) rising 60% in the second half of the year.

I'll end this performance review on a bullish note. While we clearly erred in our assessment of most housing-related stocks last year, we're working hard to not anchor on that experience--otherwise, we could easily become too conservative and miss the bottom whenever it does come. Moreover, it's important to remember that stocks do not become cheap when the headlines are good, and we try to add value for our clients by being willing to step into seemingly ugly situations. Last year, many of these contrarian calls hurt our performance, but we continue to believe that, over time, buying assets for 70 and 80 cents on the dollar is a rational route to solid investment returns.

At the moment, the median stock we cover is about 12% undervalued, while large-cap, high-quality companies look even cheaper, trading at 15%-20% discounts to our estimates of intrinsic value. In our opinion, this is a better time to be a net buyer of stocks than a net seller, especially because the cheapest assets--wide-moat mega-cap stocks--are also the ones best positioned to serve as the bedrock of a long-term portfolio.