What is smart at one price is stupid at another." -Warren Buffett
The S&P 500 returned 31% in 2019. With the exception of its 32% return in 2013, this was the S&P 500’s largest annual gain in the past 20 years. The Oakmark Fund produced a return of 27% for the year and Oakmark Select returned 28%. Compared to almost anything other than the S&P 500, those are very good numbers and exceeded almost all beginning-of-the-year predictions for what a mutual fund would return in 2019. Still, for the third straight year, both Oakmark and Oakmark Select returned less than the S&P 500.
We are encouraged that the past quarter showed signs of a turn—both funds outperformed a strong market. But given that we and our shareholders expect our Funds to outperform the S&P 500 over the long term, we wanted to focus this report on our relative performance as opposed to our strong absolute performance. We hope this will help our shareholders answer this important question: “Did Oakmark trail because value investing is, as a strategy, underperforming? Or is Oakmark doing a poor job of implementing its strategy?”
To answer these questions, let’s first look at how value performed in 2019. You’ve probably heard of the “Dogs of the Dow” theory, which states that the 10 cheapest stocks in the Dow Jones Industrial Average (based on dividend yield) tend to outperform that index over the following year. And from 2000 through 2018, this held true: the “Dogs” outperformed the Dow by an average of 150 basis points per year. But in 2019, the 10 “Dogs” underperformed the other 20 stocks in the Dow by a whopping 1770 basis points, returning 13% versus 30%.
Using another value measure, if, at the beginning of 2019, you had bought the 50 cheapest S&P 500 stocks based on price-to-book value, you would have underperformed the rest of the S&P 500 by over 500 basis points for the year. Similarly, if you bought the 50 stocks with the lowest expected 2019 EPS growth, your portfolio would have underperformed the 50 highest expected growth stocks by 830 basis points over the year. Consistent with that, the Russell 1000 Value Index, an index composed of lower priced stocks relative to earnings and book value, underperformed the Russell 1000 Growth Index by 990 basis points in 2019. Given these results, it’s no surprise that Morningstar reports1 that large-cap value funds as an aggregate underperformed large-cap growth funds by 690 basis points in 2019.
Undoubtedly, we made our share of mistakes at Oakmark in 2019. But the data suggest that our much bigger problem was that investors were not very concerned about valuation levels. Though this can be frustrating, it also gives us the opportunity to start the year with a portfolio of stocks that our research suggests is at a larger discount to the market than is typical.
To examine these opportunities, let’s take a step back and compare U.S. equities with the bond market. Many investors think a 10-year U.S. Treasury bond is a riskless investment because the U.S. Treasury not paying its debts is unthinkable. And if you rule out default, except for inflation risk, it would be risk-free—but only if it is held to maturity. However, if you hold that bond for a shorter period, its total return will be a combination of its coupon yield plus the change in value caused by interest rate changes.
For example, if interest rates rise to just 2.1%, a 10-year Treasury bond that currently yields 1.9% will generate a negative one-year return. And good luck to those pension funds relying on 30-year Treasuries repeating their 7% annualized return from the past decade. That will only happen if their yield, starting at 2.3% today, goes to negative 1.2% a decade from now. I guess nothing is impossible, but this outcome seems highly unlikely. Fixed income investors who ignore the impact of interest rate changes have a lot in common with equity investors who ignore the impact that movement in P/E multiples have on stock prices.
I sometimes get frustrated with legal edits that don’t allow me to say things like, “Alphabet is a great business.” Despite the company’s demonstrably superior financial metrics, that statement is an opinion, not a fact. So, when my writing comes back from editing, it is often filled with new insertions like, “we believe,” “in our opinion,” “it could be the case” and so on. At times, it feels as if I have to write, “Two plus two, in our opinion, equals four.” (To be clear, our lawyers aren’t to blame. Rather, it’s our industry’s history of bad actors who stretched the truth that have led to increased regulation.) Because of this, I’m excited that I can write something definitive about stock prices: a stock’s price always equals its price-to-earnings ratio times its earnings-per-share, or P= P/E x EPS.
As value investors, we pay close attention to P/E and base most of our investments on the premise that a stock’s current P/E ratio is too low. If a stock moves to what we believe is a fair multiple, the result is a higher price. Occasionally, we have a strong non-consensus view on earnings potential, such as when we believed that Baxter’s new management team had an opportunity to nearly double margins. Likewise, as the 2008 recession came to a close, we believed that earnings would get back to “normal” over our seven-year time horizon—a decidedly more positive outlook than most investors had at the time.
Usually, however, we don’t quarrel much with consensus earnings forecasts. Instead, we believe that our stocks will benefit from higher P/E multiples. That was our view in 2000 when we avoided technology stocks that were selling above the S&P 500’s 30 times multiple and instead owned single-digit P/E stocks, such as consumer packaged goods, industrials and financials. Today, you can see this same logic at work in our bank and cyclical holdings, with many selling at single-digit P/Es, and our avoidance of utilities, consumer packaged goods and REITs that trade at P/Es in the 20s.
Although the formula P=P/E x EPS highlights that estimating future P/E is just as important as forecasting EPS, investors typically alternate being obsessed with one factor and then the other. The collapse of the tech bubble in 2000 was a time when investors stopped paying higher and higher prices for the fastest growers and quickly pivoted to low P/E stocks. And today, just like during the height of the tech bubble, analysts are focusing much more on a company’s earnings than on the company’s appropriate P/E multiple. It’s not the analysts’ fault. After all, their job is to earn commissions from their clients, and today, most of their clients are paying them to focus on earnings predictions.
But this focus on earnings instead of valuation has led to some very—shall we say—interesting analyst reports, including the following “takes” we’ve seen on our own holdings:
- One analyst wrote that he believed, as we do, that DXC’s new CEO will restructure the company and largely eliminate the quality gap between DXC and its public peers over the next three years. Yet, in the same report, the analyst set the company’s target P/E at a 30% discount to its peers—unchanged from its historical average, despite its improved competitive stance.
- An in-depth report on Lear highlighted the company’s many advantages compared to other auto parts businesses that sell for between 5 and 11 times EBITDA. But then the analyst computed Lear’s new target price using a multiple of 4.8 times EBITDA. Why? That was left to the reader’s imagination.
- A report on CBRE Group touted the company’s improved business mix. Over the past few years, CBRE’s maintenance outsourcing segment has grown rapidly compared to its more cyclical brokerage segment—historically the larger part of the business. Importantly, the market tends to value recurring income, like that from service businesses, at a much higher P/E than businesses based on one-time transactions. Nevertheless, this still concluded that CBRE is fairly priced because its current P/E is approximately at its 15-year average.
- A report on Constellation Brands kept the company’s target P/E the same—at 17 times—despite the company’s recent purchase of a large interest in Canopy Growth Corporation. Canopy’s losses reduce Constellation’s reported EPS by about $0.85 so the analyst is inadvertently valuing Constellation’s Canopy investment at negative $-14 per Constellation share, despite its market value being positive $14.
- Another report noted that big banks are safer and more competitively advantaged today than at any time in recent history. Yet it concluded that these banks are fully valued at their current price of 10 times earnings—which is a P/E roughly the same as their 30-year average. The report never explained why the improved business fundamentals shouldn’t be rewarded with a higher P/E multiple.
The largest industry weighting in our portfolios, financials, demonstrates why we believe our Funds will benefit when valuations become a bigger determinant of prices. In the Oakmark Fund, for example, we own 10 stocks in the financials sector, comprising about 30% of the portfolio. Their median P/E on expected 2021 earnings is 9 times, compared to the S&P 500 at 16 times. Median price-to-book is 1.2 times and dividend yield is 2.3%, compared to the S&P 500 at 3.6 times book and a 1.9% yield. So, on earnings, assets and yield, the banks appear much cheaper than the S&P 500. Normally, stocks that look that cheap are expected to grow much slower than the market or even experience declining earnings. In this case, however, we expect our median financial stock to have annual EPS growth of 8%, which exceeds the consensus expectation for the S&P 500. To us, faster growth, higher yield and cheaper price translate to win, win and win. We believe that the market will eventually reflect our view by narrowing the gap between the S&P 500’s and the financials sector’s P/E ratios. Our portfolio is filled with stocks whose stories sound similar and our research leads us to believe are selling at bargain prices—relative to both other stocks and to the absolute returns we expect in assets other than equities.
One year ago in this
commentary, after the market fell 14% in the fourth quarter, I wrote:
“The stock market looks more attractive to us than it usually does, and the divergence among individual stocks allowed us to structure a portfolio that we believe is more undervalued relative to the market than it usually is. Though the decline has made watching the market painful, we are all gritting our teeth and adding to our personal holdings.” With hindsight, we were right about the market being unusually attractive, but we have yet to prove that our portfolio was more attractive than the market. It is frustrating when market performance doesn’t reflect our estimates of business value, but that’s what creates opportunity. Since our longest tenured mutual fund, the Oakmark Fund, started in 1991, its annualized return has been 12.5% versus 10.0% for the S&P 500.
Yet during those 28 years, our trailing three-year return has lagged behind the market 49% of the time. That number falls to 35% for 5-year and just 22% for 10-year time periods. We understand that patience is in short supply when a fund underperforms. In addition to our strong long-term record, consider a few other issues when evaluating our recent returns.
First, most value funds have underperformed over the past three years at least in part because investors have shown little concern for valuation as some high growth stocks have surged. Second, the relative values that are available today in sectors like financials (our largest exposure) seem historically unusual. And, finally, our investment philosophy and team have been remarkably consistent throughout our history. In our view, this consistency is a major factor behind our long-term outperformance.
We believe our long-term returns have been higher because we have applied our value approach consistently as opposed to following current market trends. Based on what we’ve seen in the past, we believe today’s market offers the opportunity to profit from a potential narrowing of the gap between business value and stock price. That’s exactly what we’ve been exploiting for the past 28 years.