Not just all-in on BRICs
Joseph Mariathasan and Martin Steward research some of the top performers in emerging market equities and finds considerable diversification, not only between top-down and bottom-up approaches, but between valuation methodologies
While the crash of 2008 showed that emerging stock markets are not decoupled from problems originating in developed markets, the economies are decoupled in the sense of having sustainably higher growth rates. That is one of the reasons for a key dynamic in the management of emerging market equities, namely how much credence should be given to country selection and how much to pure stock selection.
Despite increasing globalisation, stocks in many sectors are still much more closely correlated with their domestic country index than with the global sector index. With 8,500 or more stocks to consider, managers have to be able to screen the universe. While few managers would claim to run active emerging market equity portfolios without any stock selection, there are some successful fund managers such as First State and Polunin that have run portfolios for years without any country selection, while others like Robeco and Baring Asset Management attribute as much as 50% of their outperformance to country selection.
Proponents of country models such as Paul Wimborne, director in emerging market equities at Barings, argue that just as one would not like to have a good house in a bad neighbourhood, one should exclude even good companies if they are in a ‘bad' country. Barings's process therefore starts with country selection, then follows with stock selection to form a concentrated, 55-stock portfolio, and concludes with risk management that determines position, sizes and checks the cohesiveness of the portfolio.
The country selection covers the 20 main markets in the MSCI universe and is based on a mixture of qualitative and quantitative factors covering GDP growth, liquidity (focusing on the interest rate cycle and money supply), currency, management and valuation. As Wimborne explains, the overall score is not mathematically derived from the others but is a qualitative view that can change quickly: "In 2007, Thailand had a military coup which meant that the management factor overrode all the others," he notes.
Robeco's country selection approach is also based on five factors which head of emerging markets, Wim-Hein Pals, says consists, first, of macroeconomic and political effects (GDP growth, monetary policy, FX reserves, the interest rate environment and political risk), combined to give a macro score; second, earnings expectations (the aggregate figures for a country based on IBES data); third, valuations (based on standard ratios such as price/earnings, price/book and dividend yield); fourth, price momentum; and fifth, sentiment, as suggested by fund flows and debt spreads.
Unlike Baring, Robeco does occasionally own a stock in a country that is unattractive, although with a 140-stock portfolio, an individual name is necessarily less significant - but once a country weight is decided the firm will only pick what it perceives to be the best stocks in that country: "The Indian manager has to outperform the Indian index," says Pals.
Baring's country bias is towards Brazil, Indonesia and Turkey - which Wimbourne describes as "all having large populations, and in the last decade, have made significant progress in legal, political and economic reforms" - while being underweight countries that rely on OECD demand such as Korea and Taiwan.
Robeco, by contrast, has its largest overweight in Korea: "This is driven by the domestic recovery and also the Chinese angle, with large blue chip Korean companies having very strong links to China and benefit from growth there," explains Pals.
A bottom-up stockpicker like First State would argue that even if a country is facing problems it is better to maintain a focus on its good quality companies. "At one end of the quality spectrum, falling input prices and cheaper currencies will be positive for the long-term profitability of a wide range of companies in our portfolio," says joint deputy head of emerging markets, Jonathan Asante. "And in a downturn strong managements really earn their keep and create stronger franchises."
Still, while the pure stock selectors have little or no top-down country views, their stock selection obviously throws out preferences by default. First State has very low exposure to China and Russia because these two economies are "dominated by government", making it difficult to identify companies with the characteristics that First State looks for.
Asante also sees risks in the valuations. "China's growth rate of 8% or more is approaching a bubble," he says. First State does, however, retain a large exposure to a Hong Kong-based, family-owned company with large gas distribution activity in mainland China.
In contrast, Barings' Wimborne is bullish on the prospect of monetary and fiscal stimulus feeding into asset prices in China, which leads him to focus on banks, property insurance and consumer discretionary stocks. While China has not been on the top of Baring's country list, the high beta exposure of the stock picks means that China accounts for 20% of its risk budget.
Just as for First State, Russia's reputation for poor corporate governance makes it less attractive to JPMorgan Asset Management, according to its product specialist Claire Simmonds. This, combined with inflationary trends, informs their underweight. Again, Wimborne at Barings dissents. "Russia was sold off aggressively last year, but it went into the crisis in a good position with large FX reserves," he observes. "Whilst the US dollar debt levels were a source of concern, as the oil price increased a lot of those concerns dissipated."
Jonathan Neill, chief strategist at FPP Asset Management in London, offers a typically forthright assessment of this sort of debate: "We can waffle all day about whether corporate governance in Russia is better nowadays, about whether Indonesia is a basket case," he says. "The point is how much you pay for assets within those countries."
FPP's highly systematic quantitative approach not only strips out any emotional aspect of stock- or ‘theme'-picking, but also brings country and stock-picking side by side. It begins with Neill's concept of ‘appraised value' (AV), which values all listed emerging market companies, regardless of sector, according to their rate of equity growth, the return on that equity, and the level of dividends, all divided by book value: three figures are generated - one of the past five years; one for the next 12 months and one for the next two years. These are aggregated to give the same ratios at country level, and then divided by earnings yields for each country, which results in an initial country portfolio weighting.
To this assessment of absolute value, FPP adds an assessment of value relative to bond markets, which is used to multiply the initial country weight to provide new recommended country weights. These are multipled by the MSCI Emerging Markets weights - "so that we don't end up with 40% in Columbia", as Neill puts it - while a country scoring gauge based on interest rates, debt levels, exchange rate and stock market momentum also serves as an overlay to country weightings. The stock valuations that generated the country weights are then used to arrive at stock weights within the country allocations, and the whole thing is rebalanced quarterly.
The concluding index adjustment might stop 40% going to Columbia, but it does not tie the portfolio back to its benchmark. FPP's 10% Russia position is a "massive overweight" against MSCI's 1.8%, for example; Taiwan and Thailand also score well; whereas the fund's Hong Kong and China exposure is almost three times smaller than the index's and India exposure is almost half.
"Our country-scoring gauge often helps us to decide when shares are cheap enough to justify investing in a country that's in a bad state," says Neill. "In 2008, Russian appraised valuations were getting lower and lower and the process eventually recommended a 20% allocation, but our country-scoring gauge warned us to go in gradually - interest rates were rising to protect the falling currency. As soon as rates started falling and the currency started rising we struck - and went straight to 20%."
Whether stock selection accounts for 50% or 100% of the target alpha, all fund managers are faced with the issue of how to allocate scarce resources in undertaking due diligence on potential purchases, or indeed, sales. Neill, of course, has "no interest in visiting any of these companies" and insists that he makes "better decisions now than when I used to visit companies earlier in my career". Polunin (a small London boutique with just $350m or so under management) and First State (with over $6.5bn) represent the opposite extremes in fund manager size, yet both have good longer-term track records based on pure stock selection. However, their approaches to bringing down the pyramid of choices to a more manageable level for detailed due diligence are very different.
At First State, the focus is on screening out companies that appear to have weak or shareholder-unfriendly management. "The future is uncertain and macro economics is not good at forecasting earnings of companies," explains Asante. "A good company will still do OK, even when the tide goes out. We are very favourably inclined towards family-owned companies while we are not keen on state-controlled companies." First State also rules out companies that take on excessive risk through foreign currency borrowings, or whose accounts cannot easily be understood. This brings down the universe to 250 companies for ongoing monitoring: the GEM portfolio has 100 stocks; the ‘Leaders, portfolio 50, while its most concentrated portfolio has 30.
Polunin's approach is unusual in its focus purely on replacement cost as the best valuation metric for companies in emerging markets. "Defining the industrial sector as the common ground for companies in emerging markets gives the best vantage point," explains Douglas Polunin, its CEO. "Only companies with the most extreme values in each sector, the strongest balance sheets and the most favourable sector outlooks are considered." Polunin undertakes research on around 10,000 companies from a database of 16,000, which includes some 1,000 company visits a year. Most fund managers are just focused on quarterly earnings, he observes, whereas Polunin is "only interested in the long term-attributes of the business and how the assets are set up".
Polunin uses time-series software to store information on replacement costs across all sectors; a company can then be broken down into component businesses and valued accordingly. "With retail stores, we ask questions like, ‘Do you own the sites or have leases?'" explains Polunin. "That makes a huge difference to margins. If an earthquake takes out a polythethylene factory in Taiwan, we can find out which firms produce polyethylene and which ones have the greatest upside."
The firm screens out the cheapest quartile in each sector, which leaves 2,500-3,000 companies. These are passed through a liquidity screen to ensure that a position can be liquidated in seven trading days, reducing the list by half again. "We end up with a list of 1,500 companies whose replacement value may be $2.5bn-5bn, whilst the market may be valuing them at $500m," explains director Julian Garel-Jones. "[Polunin] argues that valuing companies using this industrial logic generates better long-term returns and ignores short-term noise. Companies are sold when their valuations reach the median average for the sector. With a 100-stock portfolio and an 80-120% turnover and a 6-12-month horizon, it is using long-term valuations for a short-term holding." As the firm readily admits, while the strengths of the process are that it covers periods of stress such as Brazil in 1999 and the globe generally in 2008 very well, it avoids overvalued currencies and sectors, and the replacement cost and relative value methodology has huge advantages over subjective valuation methodologies such as price-to-book, the weaknesses are that the process underperforms during bubbles such as the oil and commodity-driven markets in 2008, as well as during periods of very strong inflows with little market breadth that result in momentum-driven price rises.
JPMorgan also has a pure stock-focused approach, but with $13.5bn in its flagship GEM product alone, its size constrains the universe they can consider. While frontier markets are managed in separate regional and country funds, the GEM strategy has a universe of 1,000 stocks based on the MSCI index plus some acceptable off-index stocks. Its approach relies heavily on ratings from individual analysts given free rein to use either an earnings or asset-based approach. "The objective is to understand the four drivers of return - earnings growth, dividends, currency movements and change in valuations," explains Simmonds. The strategy has a long-term time horizon of five years and very low turnover, averaging 30% in most years and just 10% in 2009.
Baring's stock selection is also based on leveraging its 30-plus team of emerging market analysts and fund managers in London and Hong Kong, who use a scoring system looking at the same factors that the firm applies at country level (growth, liquidity, currency exposure, management and valuation) to give a traditional GARP approach. Some 70% of analysts' bonuses is tied directly to the performance of their stock picks, whether or not they have been selected for portfolios. By contrast, Robeco's stock selection approach uses a quantitative model as well as qualitative analysis. The quant model is based on three factors - valuations, earnings momentum and price momentum. "This acts as a screen and also as an idea generator," says Pals. "We then do fundamental analysis, have meetings with management, and so on."
Emerging market equities have risen from 4% of the MSCI Global Equity index in 2002 to 12% at the end of July 2009. Emerging market economies now account for 50% of global GDP. Few would dispute that these proportions will go any other way than up. This growth will be driven by the rise of domestic demand, as opposed to the historical drivers of emerging market economies - exports to OECD countries.
Not surprisingly, managers often focus on this theme. JPMorgan's portfolio, for example, has a strong domestic demand bias, and underweights cyclicals such as IT, which Simmonds feels have lower earnings visibility. More recently, it has been adding to financials across the globe. "Financial risk was significantly discounted in emerging markets, but their bank systems were still functioning and subject to a structural growth trend," Simmonds says. In addition, financial stocks are closely tied to the domestic demand story. Baring's overweight positions in Brazil, Indonesia and Turkey have a common factor. All three countries have made progress from a situation of inflation higher than 70% and interest rates higher than 40%: "As rates come, down, you start to see the emergence of a consumer finance market that leads to the growth of domestic demand," observes Wimborne
Again, FPP's Neill pours cold water on many of these well-worn emerging market themes. "It's really difficult to tell which themes are going to be the strongest and which companies and management teams are going to be best-positioned to exploit those themes," he says. "In fact I think it's impossible. That's why I prefer our ruthless tools." Correspondingly, perhaps, while his strategy is stock and country-driven, sector weights fall out by default and appear rarely to drift from the benchmark. This makes sense now, while stocks are more closely correlated with domestic markets than with global sectors. But will that dynamic change as these companies become the new global titans? Perhaps, Neill concedes. "But countries tend to be dominated by one or two key sectors. Technology in Taiwan, for example. Technology and finance in Korea. Half the Brazilian index is oil and iron ore. So the sector work is implicit in what we do at country level."
Still, the structural growth of domestic demand is the driver for increasing emerging market allocations. Should institutional investors adopt a backward or a forward-looking weighting of their equity portfolios to emerging markets? One would hope that it will be forward-looking - but the problem they might face if they want to increase exposures is that some of the best fund managers are now closing to new business.