Old Friends

Two Concerns About the Cape Ratio
In yesterday's English edition (registration required) and today's American edition, Financial Times weighs in on the debate between the professors Shiller and Siegel about U.S. stock prices. As previously covered here, Robert Shiller's cyclically adjusted price/earnings ratio (Cape ratio) shows the U.S. stock market to be expensive by 100-year standards. Jeremy Siegel responds that yes, the ratio does, but due to changes in U.S. accounting practices, the first 85 years' worth of data in the chart are noncomparable--and that stocks look acceptably priced by the standards of the past 15 years.

The two editions' articles are not only different, but they differ in their conclusions, although the articles were written by the same author. (And you thought I was confusing!)

Monday's story (which also discusses another long-term valuation ratio called Tobin's q), suggests that Siegel's argument is "nonsense" stated by those who "wish to assert that shares are always cheap," in the words of Andrew Smithers, a Cape devotee who is the story's only source. The article concludes that "the conventional academic wisdom that U.S. stocks are overvalued still looks convincing." 

Today's article also includes a Smithers quote but spends more ink on the work of three professors (Dimson, Marsh, Staunton) from the London Business School who doubt the usefulness of the Cape ratio. That ratio, and other valuation measures like it that rely on the concept of mean reversion in stock prices, is useful when looking backward at an existing data series, claim the professors, but not helpful as a practical signal at the time of investment. They write, "Sadly, we learn far less from valuation ratios about how to make profits in the future than about how we might have profited in the past."

The professors make this statement not only for the United States but also for 19 other countries, as they tested various trading strategies based on stock-price mean reversion on a 20-country database. As a general rule, they write, "Without the benefit of foresight, the evidence on mean reversion is weak. Market-timing strategies based on mean reversion may even give lower, not higher, returns."

To summarize, there are two reasons not to implement the Cape ratio's bearish signal by reducing stock exposure:

1) The signal might not be bearish (Siegel's argument);

2) The signal might not give useful practical information (the London Business School professors' argument).

That's more than enough for me. The odds are pretty high that at least one of those two items is correct.

Twice the Data
My column on the relationship between a country's corruption level and its 20-year stock returns drew some fire. Because MSCI's best, most accurate measure of stock market performance, Net Returns, is relatively new, my data set included only 21 countries, 20 of which scored low for corruption. (The only exception was Italy.) Thus, the critics said, I didn't demonstrate that as a global rule, low levels of corruption were associated with higher stock returns. Rather, the chart (shown below) illustrated that among the least corrupt of countries, the very cleanest fared the best. The criticism was fair; those are indeed not the same thing.

  - source: Morningstar Analysts

Therefore, I returned to the MSCI data, this time using all trailing 20-year Gross Returns. The Gross Returns calculation is not quite as good as Net Returns, but it's close enough. Switching to that data set expanded the country list to 44 from 21. Most of the additional countries were emerging markets with average to below-average corruption scores, such as Peru, India, China, and Mexico.

The new and expanded graph is below. (Sorry for the eye test; it's difficult to publish a chart with this much data in this column's format.) The pattern is weaker, but it still exists. The biggest outliers are Colombia, Peru, and Brazil. They had top performances despite having relatively high levels of corruption. Overall, though, the cleaner countries once again performed better. Of the 10 least corrupt countries, six had top-quartile gains, and none finished in the bottom quartile. 

  - source: Morningstar Analysts

Unlike with the developed markets, there was no relationship among the emerging markets between stock returns and the level of corruption as measured by Transparency International. The most scrupulous of the new countries (Chile, Taiwan, Israel, Poland) didn't outgain the least scrupulous (Pakistan, Indonesia, Philippines, Mexico). However, overall the emerging markets trailed the developed markets. Thus, the addition of the emerging markets sent a mixed message, as it weakened the case from one perspective and strengthened it from another.

The argument strengthens if performance is measured on a risk-adjusted basis, as it properly should be (and would be if this were a formal paper). Over that particular 20-year period, the countries that had relatively safe, stable, and noncorrupt governments, with stock markets that showed relatively low volatility, were markedly better investments than were the higher-risk countries. Less risk, more return--lack of corruption provided something of a free lunch.

Of course, as with the Cape ratio, it's easier to profit when looking backward. Will there be a link between a country's honesty and its stock returns over the next 20 years? Not sure. But it's a subject worth considering. Traditionally, international investors have paid far more attention to prospective gross domestic product growth than to the presence of corruption. (As Morningstar's Gregg Wolper points out, in the early 1990s there were three closed-end Mexico funds, two closed-end Brazil funds, and one each for Malaysia, Indonesia, and the Philippines--but none for Denmark, Finland, or Sweden. There still are none for Denmark, Finland, or Sweden.) Those priorities seem to be misplaced.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.