The past year was a lousy one for investors in emerging markets. The MSCI Emerging Equity index fell 19.3% in the 12 months ended September 30, 2015, the J.P. Morgan ELMI Plus index of emerging currencies was down 12.5%, and the J.P. Morgan EMBI Global index of emerging sovereign hard currency debt was down 2%. By comparison, the S&P 500 was down 0.6%, the Russell 2000 up 1.3%, MSCI EAFE down 8.7%, and the Barclays U.S. Aggregate Bond index was up 2.9%. It would seem to qualify as a nightmare year for emerging, but those of us old enough to have been paying attention back in the 1990s can remember what a true emerging nightmare is. In the year to September 30, 1998, MSCI Emerging was down 48% and the J.P. Morgan EMBI Global down 24%, against a rise of 9% for the S&P 500 and 11.5% for the Barclays (then Lehman) U.S. Aggregate Bond index. So those past 12 months could have been worse. But considering that we thought that emerging assets were some of the cheapest around a year ago, Lord knows it could have been better.
And this wasn’t the first bad year for emerging, or the second, or the third. Table 1 shows the 1-, 3-, 5-, 10-, and 15-year performance for emerging versus EAFE and the S&P 500, and it is abundantly clear that things have been quite bad for half a decade now.
Losing to the rest of the world by 14.7% in a year is bad enough, but losing by 12.2% per year for five years is a lot worse. A dollar invested in MSCI Emerging has turned into $0.83 over that period, while a dollar invested in MSCI EAFE has grown to $1.21 and a dollar in the S&P 500 to $1.87.
Is it any wonder investors are questioning why they allocate to emerging markets in the first place? Even going beyond the woes of emerging, we are starting to hear some investors asking whether holding non-U.S. stocks is at all necessary. As market historians we can say that the timing of such sentiments tends to be bad – no one seems to ever decide to give up on an asset class after it has just had good performance, and the last burst of “why bother with non-U.S. stocks” occurred just before the top for the S&P 500 in 2000. But just complaining that investors got it wrong last time they voiced these sentiments does not qualify as thoughtful analysis. Relative to what we were thinking five years ago, emerging equities have done surprisingly badly, and the U.S. equity market has done surprisingly well. Was that the luck of the draw, which has no bearing on future returns? Was it a temporary phenomenon that will soon reverse? Or does it tell us something important about emerging being a value trap and/or the U.S. being extraordinary that we need to take into account in our forecasting of the future?
The short answer to these questions is that while emerging markets “deserved” some of their bad luck over the last several years and the outperformance of the U.S. has made some sense, we do not believe that emerging is a value trap, nor do we believe that the U.S. has proved itself particularly extraordinary. There are some reasonable models for valuing the U.S. stock market that make it noticeably more attractive than our overall seven-year forecast estimate of -0.6% real, but others that make it look meaningfully worse. The different valuation models for emerging are actually saying surprisingly similar things today with regard to expected returns, and one of the major headwinds for emerging over the past five years, currencies, may well soon turn into a tailwind. None of this is to say that emerging doesn’t have its share of problems or that the U.S. may not pull a rabbit out of its hat, but we do not currently see any reason to assume the worst from emerging or the best for the U.S.
Is EM a value trap?
What would it mean for emerging equities to be a value trap? If we are valuing these equities on a normalized earnings basis, which GMO’s forecasts effectively do, two possible reasons emerging would be a value trap is if “true” normalized earnings are a lot lower than our models make them out to be, or if the return to the stocks is not commensurate with an assumption that outside shareholders get the benefit of corporate earnings. There is also a third potential problem that is a little harder to quantify, but still important. If there were some sort of malign interaction between the equities and currencies such that we should expect foreign currencies to fall against the dollar with no corresponding increase in local real returns, we could find emerging equities winding up a value trap from a U.S. dollar perspective even if not from a local one. Because any of these problems could trip us up, let’s take them in turn.
We use a variety of ways to come up with the normalized earnings power for stocks, but for this purpose, let’s focus on ROE and book value as the measure. We generally assume that book value grows slowly over time while ROEs are volatile and bounce around with the state of the business cycle. This means that to calculate normalized earnings we need to come up with a normal ROE. This is not a trivial exercise, but here is how we have done it in this case. Exhibit 1 shows the ROE for emerging markets and emerging markets ex-financials and resources over time, along with our assumed normal ROE.
We can see the impact of the resource companies and financials on the profitability of emerging stocks, but it isn’t overwhelming. We prefer excluding resource companies for analyzing the profitability of emerging today because not only are they quite unprofitable now, but from 2005-2012 their profitability was significantly better than other stocks and, hence, dragged up the averages. As a result, we prefer to build a separate forecast for the resource companies taking into account the fact that their profits are likely to be much lower than they averaged over the last decade. So the profitability line we are more interested in is for the group excluding resources, and on that basis current profitability looks slightly better than normal.1 Now, the actual way we come up with our profit normalization uses book and other measures to help us estimate economic capital, for it is really return on economic capital that should be mean reverting. Table 2 shows the different estimates for return on economic capital that come from the four models we use to proxy for that unobservable quantity.
The four estimates are actually quite close to each other for the groups excluding financials and resources, which gives us some comfort that we are not doing anything too aggressive with our normalization of earnings. The figures would all be far friendlier if we included the financials and resources, but our best guess is that doing so would be too friendly.
We therefore estimate that emerging normalized profitability is a little better than current levels on an overall basis and a good bit worse excluding financials and resource companies. It is hard to be supremely confident in this, but for true normalized earnings to be significantly lower than we are estimating, it would mean that today’s environment is currently a very good one for profitability compared to true “normal.” That seems out of keeping with the general state of most emerging economies today, and none of our proxies for profitability suggests that it is true.
But profits only tell part of the story. We know that in parts of the emerging world, earnings yields overstate the likely long-term returns to outside shareholders. In the case of Gazprom, for example, corporate investment decisions are made with a lot of concern for the Russian government’s view of its strategic priorities and much less worry about the return on capital employed. If emerging equities are generally characterized as acting like Gazprom, we would see returns to emerging that are systematically well below what one would expect from the stated earnings of the companies. So let’s look at the results. We’ve got decent financial data on emerging companies going back 20 years, so Exhibit 2 shows what has happened.
This exhibit takes a little explaining. The average earnings yield over the 20 years has been 6.7%. So, we might have expected real returns to be 6.7% over the period. But the earnings yield of emerging has risen from 4.8% to 7.6% since 1995. A rising earnings yield equates to a falling price, and the effect of that change has been to suppress returns by 3% per year. That means a fair return to emerging would have been 3.7% real, and the actual return has been 3.3% real. The 0.4% gap is something we refer to as “slippage” and it says that emerging stocks have indeed done a bit worse than one would expect given their earnings yield. However, given the long-term data for equities around the world, we assume 0.5%/year slippage in our forecasts, so emerging equities have been almost exactly in line with what we expected from them. The one fly in the ointment for this analysis, though, is the fact that these returns are local real returns, so there remains the possibility that currencies could have led to further problems for emerging equities.
Is it just a currency thing?
Falling currencies have indeed been a significant driver of losses in emerging equities over the past few years. In the 12 months ending September 30, for example, the local return to emerging stocks was -7.1% while the loss from the currency movements cost 13.1%. This might tempt one to think about currency hedging their holdings in emerging markets, but this turns out to be a bad idea historically. In fact, it seems to be a spectacularly bad idea, as we can see in Exhibit 3.
This is a truly striking chart. Since 1995, the rather anemic returns of +3.0% real in U.S. dollars for emerging equities turn into a truly depressing +0.9% real. So much for the idea of curency hedging! Actually, this speaks to a broader question. If the return to holding the currencies has been so close to that of holding the equities, is it possible that just owning emerging currencies is the right way to play emerging markets?
In order to answer this question, we need to dig into the sources of return for emerging currencies. Currencies are pretty simple assets, so the breakdown is straightforward. The two sources of returns to a currency are the real interest rate earned and the change in the real exchange rate. While you could look at both in nominal instead of inflation-adjusted terms, for countries whose inflation rates have differed considerably from the U.S., that could easily give quite misleading answers. Exhibit 4 shows the returns to the J.P. ELMI Plus Emerging Currency index broken down into the real interest rate and FX components.
