Monday, July 29, 2013

Developed versus Emerging Markets: Convergence or Divergence?

In my last post, I looked at country risk first from both a bondholder perspective (with ratings, default spreads and CDS spreads) as well as an equity investor perspective (with my estimates of equity risk premiums by country). While default spreads in sovereign bonds and differences in CDS spreads are explicit and visible to investors, the question of whether equity markets price in differences in equity risk premiums is debatable. In fact, there are quite a few analysts (and academics) who argue that country risk is diversifiable to global investors and hence should not be priced into stocks, though that argument has been undercut by the increasing correlation across equity markets. In this post, I look at the pricing of stocks across different markets to see if there is evidence of differences in country risk, and if so, whether market views of risk have changed over time.

Stock Prices and Risk Premiums
Holding all else constant, stocks that are perceived as riskier should sell for lower prices. That can be illustrated fairly simply using a basic discounted cash flow model. Consider a firm that pays out what it can afford to in dividends and is in stable growth (growing at a rate less than or equal to the economy forever). The value of equity in the firm can be written as:

Rewriting the expected dividends next period as the product of the payout ratio and expected earnings, we get:
    

Now, assume that you are valuing two companies with equivalent growth rates and payout ratios, in US dollars, and that the only difference is that one company is in a developed market and the other is in an emerging market. If investors in the emerging market are demanding a higher equity risk premium, the emerging market company should trade at a lower PE ratio than the developed market company.

So what? A simple test (perhaps even simplistic, since holding growth and payout constant is tough to do) of whether equity risk premiums vary between developed and emerging markets is to compare the multiples at which companies in these markets trade. If emerging markets command higher equity risk premiums, you should expect to see stocks trade at much lower multiples (PE, PBV, EV/EBITDA) in those markets, relative to developed markets, for any given level of growth and profitability.

Market Convergence: The Pricing Story
To examine how developed market and emerging market PE ratios have evolved over time, I computed PE ratios for each company in every market each year from 2004 to 2012, with an update to June 2013. I eliminated any company that had negative earnings and divided the market capitalization at the end of each year by the net income in that year.

I then categorized the companies into developed and emerging markets, using conventional geographical (but perhaps controversial) criteria. I included US, Canada, Western Europe, Scandinavia, Australia, New Zealand and Japan in the developed market group and the rest of the world (Latin America, Asia, Africa, Middle East and Eastern Europe/Russia) in the emerging market group. In sum, there were 36,067 companies in the developed market group and 24,429 companies in the emerging market group. 

I considered various summary statistics (the simple average, a weighted average, an aggregate market cap to earnings) but decided to use the median PE as the best indicator of the typically priced stock in each market. In the figure below, you can see the median PE ratios for developed and emerging market companies by year, from 2004 through June 2013.

Prior to 2006, emerging market PE ratios were about 30% lower than developed market PE ratios, but after almost catching up in 2007, the banking crisis of 2008 caused a drop in emerging market PE ratios, relative to developed markets. In the years since, emerging market companies have clawed their way back and the PE ratio for emerging market companies exceeded that of developed market companies in 2012. The shift away from emerging markets in the first six months of 2013 has put developed companies into the lead again, though the developed market PE premium (over emerging markets) in June 2013 is significantly lower than the premiums commanded in the early part of last decade.

Deconstructing the Convergence
The convergence of PE ratios across the globe is striking, but it is worth noting that it is more attributable to a decline in PE ratios in developed market PE ratios than to a surge in emerging market PE ratios. In fact, this phenomenon is made more explicit if we look at the median price to book ratios across developed and emerging market companies from 2004 to 2013:
The convergence that we see in PE ratios is even more striking when it comes to price to book ratios, but note that the convergence is largely coming from the drop in price to book ratios in developed markets, not from a increase in those ratios of emerging markets.

Reasons for Convergence
The convergence in PE ratios and PBV ratios between developed and emerging markets is confirmed when we look at other multiples (EV/EBITDA, for instance). The question therefore becomes not whether there is convergence, but why the convergence is occurring.  There are at least three possible stories (and perhaps more).

1. Decline in profitability at developed market companies, relative to emerging market companies: It is possible that shifts in global economic power have made developed market companies less profitable than they used to be, thus lowering pricing multiples for these companies. One measure of profitability is the return on equity earned by companies, estimated by dividing net income  by book equity. The median returns on equity for developed market and emerging market companies, each year from 2004 to 2013, are contrasted below:
Note that emerging market companies have had higher returns on equity than developed market companies in every year. While the 2008 crisis has resulted in declines in return on equity across both groups of firms, developed market companies have almost caught up in terms of return on equity with emerging market companies, suggesting that it is not profitability that explains the PE/PBV convergence.

2. Declining differential equity risk premium (between developed and emerging market companies): A second potential explanation is that the differential equity risk premium between developed and emerging markets has decreased over the last few years. There is a fairly simple mechanism for backing out the implied costs of equity and equity risk premiums from the price to book ratios and returns on equity. If we assume firms are collectively in stable growth, the price to book ratio can be written as:

Moving the terms around allows us to restate the equation in terms of cost of equity:

To compute the costs of equity in US dollar terms, we will set the expected growth rate for each year to be equal to the US treasury bond rate in that year and derive the cost of equity for developed and emerging markets in that year. I know that assuming the same growth rate in developed and emerging markets is simplistic, but I will revisit this assumption later.

For instance, take 2004, when the price to book ratio for developed markets was 2.00, the return on equity for developed markets was 10.81% and the US T.Bond rate was 4.22%. The implied cost of equity for developed markets in 2004 is 7.52%:
Implied cost of equity in 2004 (developed) =((.1081-.0422)/2.00) + .0422 = .0752 or 7.52%
In the same year, emerging market companies had a price to book ratio of 1.19, a return on equity of 11.65% and a resulting implied cost of equity of 10.46%:
Implied cost of equity in 2004 (emerging) =((.1165-.0422)/1.19) + .0422 = .1046 or 10.46%
If you accept these estimates, emerging markets had an equity risk premium about 2.94% higher than developed markets:
Differential ERP = 10.46% - 7.52% = 2.94%
I repeated this estimation process for 2005 through 2013 to yield the following:

The last column is striking, as the differential ERP dropped close to zero at the end of 2012 before rebounding a little bit in the middle of 2013. In fact, at the 0.50% level in 2013, it is still well below historical norms.

3. Decline in differential real growth: Now, let's revisit the assumption that I made in the last section that both developed and emerging markets will grow at the same rate (set equal to the US treasury bond rate each year). You can take issue with that assumption, since emerging markets have not only more growth potential but have delivered more real growth that developed markets over the last two decades. If you assume higher growth in emerging markets than developed markets, the table above overstates the equity risk premium for developed markets, while understating the premium for emerging markets. I redid the table setting the growth rate in developed markets at 0.5% below the risk free rate, while allowing the growth rate in emerging markets to be 1% higher than the risk free rate; this results in 1.5% difference in annual real growth rates between the two groups. 


While the differential ERP is higher in every year, with the assumption of higher growth in emerging markets, the trend line remains unchanged with the differential value hitting a low at the end of 2012. Unless you assume a widening of the difference in expected real growth between developed and emerging markets between 2004 and 2012, which would be difficult to justify given the growth in size of emerging markets over that period,  a decrease in differential equity risk premiums seems to be the most likely explanation for the convergence in multiples across markets.

In summary, the shrinking differences in pricing between developed and emerging markets cannot be explained by profitability trends or changes in real growth but can be at least partially explained by narrowing risk differentials between the markets and the globalization of companies.

Implications
The trend lines in profitability, risk and pricing over the last decade are interesting from a macro standpoint but there are three general lessons/implications for investors:
  1. Reality check for expectations in emerging markets: For the last two decades, developed market investors have been lured into investing in emerging markets by the promise of higher returns in those markets, though accompanied with the caveat of higher risk. If the last few years are any indication, it is time for investors to adjust expectations for emerging market returns, going forward. Emerging market companies are no longer being priced to generate premium returns, but they are also no longer as risky as they once were (at least relative to developed market companies).
  2. Markets can still over shoot: While it is clear that emerging markets have evolved in terms of economic growth, political maturity and risk, it also remains true that there is more risk in these markets than in developed ones. Markets, however, often move in ebbs and flows, under estimating this differential risk in some periods and over estimating it in others. Thus, a reasonable case can be made that markets were being over optimistic about emerging market risks, when they priced stocks to generate roughly the same expected returns in developed and emerging markets at the end of 2012 (see the table in the last section) and that the correction this year is a reversal back to a more reasonable differential premium. For those who believe that the reasonable premium is that observed between 2004 and 2006 (when the average differential in ERP was 2.5% and higher), this would lead to the conclusion that there is far more pain to come in emerging markets. If you believe, as I do, that the norm is closer to that reflected in the average since 2008 (about 1 to 1.5%), the correction is about half way over.
  3. Think global, not localAs companies, notwithstanding where they are incorporated, increasingly become global competitors, it can be argued that equity risk premiums will converge across markets, since each market will be composed primarily of global companies exposed to risks around the world. For investors and analysts in developed markets, there is the unsettling reality that emerging market risk is now seeping into their portfolios, even if it is composed purely of domestic companies. For investors and analysts in emerging markets, there has to be the recognition that the automatic discounts that they apply to emerging market company multiples, relative to developed markets, may no longer be appropriate. I will return to this issue in a future post.
  1. Rediscovering risk in emerging markets: A country risk premium update
  2. Developed versus Emerging Markets: Convergence or Divergence? 
  3. Market Multiples: Global Comparison and Analysis
  4. Global Businesses and Country Risk: Investment Challenges and Opportunities (Still to come)

15 comments:

  1. Great Post! Could it be that investors globally are not willing to pay more than the salvage value (assets-liabilities). While this has been true for emerging market companies (due to illiquid information) it has also become the norm for developed market companies largely due to nervousness about the global outlook. My 2 cents.

    ReplyDelete
  2. UniverseofRisks,
    You may very well be right. I read it slightly differently. In a global marketplace, I think that investors are less willing to believe that firms will be able to keep their competitive barriers to entry up for very long. That, in turn, suggests that excess returns will fade quickly and that companies are collectively more likely to trade closer to book value.

    ReplyDelete
  3. Aren't we inveriably considering profitability in computing ERP as we use ROE?
    And on second thoughts, considering Behavioural Finance wouldn't asset managers allocate higher weightage to developed markets in certain macro economic conditions (current situation where premium is paid for earnings growth or cash flow visibility) taking the developed market PE multiple way higher than emerging market multiple.

    ReplyDelete
  4. As always, a great eye-opener and mind-enriching post.

    Just a typo in the COE formulas where you should show the growth rate subtracted from the ROE in the numerator in:

    Implied cost of equity in 2004 (developed) = (.1081/2.00) + .0422 = .0752 or 7.52%

    and

    Implied cost of equity in 2004 (emerging) = (.1165/1.19) + .0422 = .1046 or 10.46%

    ReplyDelete
  5. Professor I'd love to know what you're using as your general terminal growth rate in your models these days. 1.5% - 1% ? Jeremy Grantham at GMO really got me thinking when he forecasts that US GDP growth of just under 1% is their expectation. They're seen as one of, if not the best at asset allocation, but I would love your thoughts.

    ReplyDelete
  6. I use the risk free rate as my cap on terminal growth. Ensures internal consistency in my models and is easy to get.

    ReplyDelete
  7. I really wonder about the truth of valuations being based on future dividends. Dividends are not the only way that companies distribute earnings to share holders. Volatility can be harvested either by using options or simply by re-balancing stock holdings against treasuries etc. Then volatility actually provides the potential for reward by harvesting that volatility. That makes a nonsense of the whole risk versus reward concept. Let's face it many stocks NEVER pay a dividend and the company grows, the stock price bobs around for thirty years and then the company goes bust without ever having paid a dividend and yet share buy backs, share splits and employee stock options mean that all the considerable earnings over the life of the company have been transferred to those share holders who bought low and sold high. Remember share buy back activity is a key indicator that a stock price is peaking - that shows that share buybacks are a transfer of wealth from the company to PREVIOUS share holders.

    ReplyDelete
  8. Fascinating analysis. I think it only works though if ROE is greater than Expected Rate of Return, which it is by your definition. If you use an external source of expected returns (I experimented with Div yield + LT consensus EPS growth expectations as measured by IBES) effectively using a simplified DDR as a proxy for expected equity returns numerator is negative and therefore COE cannot be calculated. Am I getting something wrong here?

    ReplyDelete
  9. I wanted to thank you for this great read!! I definitely enjoying every little bit of it I have you bookmarked to check out new stuff you post
    irs tax liens
    creative real estate investing
    organ tax lean sales

    ReplyDelete
  10. Dear Professor,

    when computing cost of equity for investments in emerging markets should one compute the formula as rf+Beta*ERP+CRP or rf+Beta(ERP+CRP)? And why is one approach better than the other from a theoretical perspective? thank you in advance.

    ReplyDelete
  11. Why growth rate has been assumed as US bond rate?

    ReplyDelete
  12. hi Aswath. Your analysis last year on EM proved very helpful in helping to avoid some of the malaise. It would be great to hear your thoughts on where EM valuations are currently. A lot of strategists are saying it is time to be more country selective.

    ReplyDelete

Given the amount of spam that I seem to be attracting, I have turned on comment moderation. I have to okay your comment for it to appear. I apologize for this intermediate oversight, but the legitimate comments are being drowned out by the sales pitches and spam.