We aim to buy stocks that trade at a substantial discount to our estimate of intrinsic value.
But what drives a stock to become undervalued? Often it is something idiosyncratic.
For example, the market can underappreciate the potential impact of a change in management, or the total addressable market of a new product.
Occasionally, however, mispricings can occur for reasons that are not specific to any single company, but instead stretch across areas that are connected by some common thread.
Consider the following example. Mitsubishi and Kikkoman are both Japanese companies. Historically they have traded at similar multiples, mainly as a result of comparable levels of underlying growth and profitability. Since 2010, however, their valuations have markedly diverged.
Unless we can find a good explanation for this valuation gap, it looks like a mispricing. Kikkoman is a food company, its main product being soy sauce and Mitsubishi is a conglomerate with stakes in a wide range of businesses.
While Mitsubishi is harder to analyse, we find no acute reasons behind its low valuation. It looks too low given our expectations for future cash flow. Similarly, we struggle to see any idiosyncratic reasons for Kikkoman’s spectacular upward re-rating.
Kikkoman’s valuation simply looks too high, and Mitsubishi’s too low. If the mispricing is not idiosyncratic, what is driving such an unusual and extreme disconnect?
One clue is in the range of outcomes for the companies’ earnings. Mitsubishi’s future profitability is difficult to predict. The business is complex and many of its earnings streams are sensitive to exogenous variables.
Kikkoman is more stable, its historical return on equity having varied by only 3 per cent from year to year. It is possible to forecast Kikkoman’s cash flow five or even 10 years out with some degree of certainty.
In the current market environment, this difference in predictability has had a tremendous influence on the companies’ valuations.
To understand why, we need to borrow a concept from the world of fixed income: the term premium.
Yields for long-term government bonds can be broken into three parts: inflation expectations, the expected path of real interest rates, and the term premium.
The term premium can be thought of as the compensation offered to investors for taking on a long dated risk. Logically, the term premium should always be positive, but today, term premiums in most developed markets are near zero, and some, astonishingly, are negative.
A negative term premium implies that investors are paying for the privilege of taking on term risk.
Yields can be low for good reasons, but it’s hard to imagine a good reason for the term premium to be negative. This looks like a real inefficiency—a mispricing.