Bill Nygren of Oakmark is out with his Q2 shareholder letter to investors. You can see the letter from Oakmark’s David Herro here. Bill Nygren discusses the Fed and risk free rates. Nygren explains why he uses the seven year treasury as the risk free rate. The note from Bill Nygren can be found below.
|At Oakmark, we are long-term investors. We attempt to identify growing businesses that are managed to benefit their shareholders. We will purchase stock in those businesses only when priced substantially below our estimate of intrinsic value. After purchase, we patiently wait for the gap between stock price and intrinsic value to close.|
A little over a month ago – May 21, to be exact – the stock market was on track for another great quarter, adding to the 11% gain it achieved in the first quarter. Since the end of March, the S&P 500 had gained 100 points (or 7% including dividends) and was trading at another all-time high price. But in the ensuing weeks, the long-awaited correction arrived, and the market quickly fell right back to where it started the quarter.
What happened? Most market commentators pointed to the declining bond market. Many times over the past few years our portfolio managers have said that the bond market was not offering an adequate return for long-term investors. We believed that most people were simply attracted to bonds’ strong recent performance; in short, many buyers were just riding a momentum wave. Further, some traders believed that, because the Federal Reserve was buying so many bonds, bonds could not decline in price (often referred to as the “Bernanke put”). Effectively, the marginal buyer of bonds was not a long-term owner, but was a “renter” of the bond, who expected to sell either to a greater fool or to a forced buyer.
That changed a few weeks ago when word spread that the Fed might reduce its bond purchases, and the “renters” reversed course and started to sell. Bond prices went down. Investors who viewed their bond assets as the safest part of their portfolio were alarmed at their losses. Retail investors redeemed their bond funds, which in turn caused more selling. As an example of the losses, let’s look at Apple’s 30-year bond, issued at the end of April at a price just below 100. Because of strong demand, that bond traded up to 104 in early May. Less than two months later it traded at 85, down 18% from its peak – a loss equivalent to more than five years of coupon income. The stock market’s 6% correction seems mild in comparison.
So a logical question is, “Has the bond market decline made Oakmark less positive on stocks?” At the risk of spoiling the suspense, our answer is no; we still view equities as undervalued. Here’s why.
At Oakmark, one model we use to estimate value is what is commonly called a Dividend Discount Model – the most basic model for calculating a “fair” price for a stock. Though we rely much more heavily on private market activity for our valuation estimates, our Dividend Discount Model provides a good reasonableness check for our more sophisticated models.
The model simply says that the combination of dividend yield and dividend growth must equate to an investor’s required return. The required return is a combination of a “risk-free asset” (typically a government bond) and a risk premium. The premise of the Dividend Discount Model is that a rational investor must get paid to accept the risk inherent in owning a stock rather than owning a risk-free bond. Academics call that incremental payment the “risk premium.” The lower the bond yields and risk premiums are, the higher the valuation and vice-versa.
Though simple in concept, estimating the risk premium demanded by investors is not as easy as it sounds. First, not all investors require the same risk premium. Second, they don’t measure against the same risk-free bond. Third, the premium doesn’t have to be stable over time. That’s part of what causes the model to only be good at giving broad estimates of value, as opposed to precise values.
At Oakmark, we decided to use a seven-year U.S. government bond as our “risk free asset” because that matches the timeframe we use for valuing our equity investments. Our historical analysis suggests that equity investors usually demand that an average stock earn about five percentage points more in expected return than that government bond. (More precisely, we use 450 basis points over the average seven-year AA-rated industrial bond, which averages about 50 basis points more than the government bond, but this commentary is easier to follow without that complication.)
In the early 1990s, when the seven-year U.S. government bond yielded 7%, our model said that an average stock needed a total expected return of 12% to be equally attractive. As bond yields fell, to 6%, 5% and 4%, our bogies fell to 11%, 10% and 9%. In early 2009 when the seven-year went under 3%, we said enough is enough. We simply didn’t believe that bond investors – buyers that were willing to hold their bonds to maturity – were really accepting a return of less than 3%. Renters maybe, but not investors. So we stopped lowering our average equity discount rate when it hit 8%. Had we not done that, last summer, when the seven-year reached a yield under 1%, our equity value estimates would have skyrocketed. Then, when rates last quarter went back up to 2%, we would have brought values back down, albeit to a valuation level higher than the one we currently use.
For four years we have been saying that bond prices are not being set by long-term investors, so instead of using the actual interest rate, we’ve used a 3% floor. This way, our equity valuation estimates have not inflated to what we see as unsustainably high levels. Instead, we believe that stocks were, and continue to be, somewhat undervalued relative to a seven-year government bond that yields 3%. Even after the recent increase in interest rates, the seven-year has moved only a little more than halfway back toward our floor of 3%. So the declining bond market has not at all dimmed our enthusiasm for equities. We are effectively already factoring in the assumption that rates continue their upward march until the seven-year hits 3%.
Should interest rates rise further than that, to levels above 3%, we will have to examine the cause of that increase. If rates rise because of higher inflation expectations, our earnings and dividend growth projections will likely also rise, largely offsetting the valuation impact. If, however, higher rates are not accompanied by expectations of higher inflation and earnings growth, our valuation estimates would need to decline.
Fortunately, with the maturing of the TIPS (Treasury Inflation Protected Securities) market, we can easily derive investors’ estimate of future inflation. That presently stands at about 2% per year, which we use as an input in our growth estimates. If rates rise above 3%, we will be