Over long periods, however, they have not. In fact, prolonged periods of mispricing can be found for several similar twin-share structures, such as Unilever (see “chart“). This phenomenon occurs because large groups of investors prefer (and are prepared to pay a premium for) one of the twin shares. Rational investors typically do not take positions to exploit the opportunity for arbitrage.
Thus in the case of Royal Dutch/Shell, a price differential of as much as 30 percent has persisted at times. Why? The opportunity to arbitrage dual-listed stocks is actually quite unpredictable and potentially costly. Because of noise-trader risk, even a large gap between share prices is no guarantee that those prices will converge in the near term.
Does this indict the market for mispricing? We don’t think so. In recent years, the price differences for Royal Dutch/Shell and other twin-share stocks have all become smaller. Furthermore, some of these share structures (and price differences) disappeared because the corporations formally merged, a development that underlines the significance of noise-trader risk: as soon as a formal date was set for definitive price convergence, arbitrageurs stepped in to correct any discrepancy. This pattern provides additional evidence that mispricing occurs only under special circumstances — and is by no means a common or long-lasting phenomenon.
Markets and Fundamentals: The Bubble of the 1990s
Do markets reflect economic fundamentals? We believe so. Long-term returns on capital and growth have been remarkably consistent for the past 35 years, in spite of some deep recessions and periods of very strong economic growth. The median return on equity for all U.S. companies has been a very stable 12 to 15 percent, and long-term GDP growth for the U.S. economy in real terms has been about 3 percent a year since 1945.(U.S. corporate earnings as a percentage of GDP have been remarkably constant over the past 35 years, at around 6 percent.) We also estimate that the inflation-adjusted cost of equity since 1965 has been fairly stable, at about 7 percent. (Marc H. Goedhart, Timothy M. Koller, and Zane D. Williams, “The Real Cost of Equity,” McKinsey on Finance, Number 5, Autumn 2002, pp. 11–5.)
We used this information to estimate the intrinsic P/E ratios for the U.S. and U.K. stock markets and then compared them with the actual values.(Marc H. Goedhart, Timothy M. Koller, and Zane D. Williams, “Living with Lower Market Expectations,” McKinsey on Finance, Number 8, Summer 2003, pp. 7–11.) This analysis has led us to three important conclusions. The first is that U.S. and U.K. stock markets, by and large, have been fairly priced, hovering near their intrinsic P/E ratios. This figure was typically around 15, with the exception of the high-inflation years of the late 1970s and early 1980s, when it was closer to 10 (see “chart“).