Martin Steward finds that a significant change in market leadership after 2009 determines who stays at the top of the emerging equity performance tables - and who falls
You could waste a lot of time looking for the shares or head offices of brewing giant Anheuser-Busch InBev in the emerging markets’ stock exchanges or capital cities; it is headquartered in Belgium and listed on the Euronext BEL20. So what on earth is it doing as the largest holding - at 5.9% - in the First State Global Emerging Markets Select fund?
It might have something to do with the fact that the company has been managed by Brazilians since the 2004 ‘reverse takeover’ of Belgium’s Interbrew by Brazil’s AmBev, or that half of its revenues come from emerging-market drinkers (its tipples account for more than 70% of the Brazilian and Argentinean beer markets).
“Where companies are listed is not so important to us as where they make their profits,” says Glen Finegan, portfolio manager in First State’s Asia Pacific and global emerging markets team. “We also hold Coca Cola Hellenic Bottling, listed in Greece but very big in Russia and with a great franchise in Nigeria.”
Anheuser-Busch InBev is priced as a developed-market stock - 13-times earnings - despite heroic achievements in efficiency savings and the stupendous growth prospects one associates with emerging markets. Indeed, First State has launched a new global fund to get around the fact that its key emerging consumer theme has become overbought in emerging markets but remains under-priced in developed markets. Its largest holding is Unilever. Hindustan Unilever is a major holding in its emerging markets portfolios, but Finegan says he is really backing new management at the parent company - because it is focusing strategy on emerging consumers.
“Similarly, Tesco can’t be held in the GEMs funds, because nowhere near 50% of its revenues come from emerging markets, but one-third of its floor space is now in emerging markets,” he says. “They’ve been extremely successful in Thailand and are partnering with the Tata Group to enter India. Tesco is trading on quite a low multiple because UK stock market investors worry about maturity at home - but if they make it work in India it begins to look very, very different.”
First State is not alone in these off-benchmark positions. Brandes Investment Partners can allocate up to 20% of its Emerging Markets Equity portfolio to markets categorised as developed by MSCI. Today that includes 2.1% allocated to Austria. It also enjoys a lot of flexibility to invest in small and mid-caps - a key source of performance over the past three years.
Brandes takes a classic Graham-and-Dodd value approach in all of its strategies, so if the Anheuser-Busch InBev and Tesco examples are indicative, it stands to reason that these off-benchmark opportunities should look attractive. For First State, it also fits with the over-riding concern with ‘quality’ within the emerging markets team (which has operated as an autonomous boutique since 2004).
“Our starting point is that we are investing in risky places, so, as far as possible, we want to de-risk our portfolio with a conservative approach to buying the long-term trend of improving living standards,” Finegan explains. The consistency of the team’s placing in the Mercer performance tables over all time horizons attests to a style that takes wealth preservation seriously. “That results in a focus on the whole grey area around quality, alignment and integrity that we find the industry does not really focus on.”
Most emerging market companies that are not majority-owned by governments are owned by individuals or families, he observes, and it is important to understand how those people treated their minority shareholders in the past when their backs were against the wall - during the crises of the late 90s, for instance. That’s recognisably the ‘G’ in ‘ESG’ - but Finegan also stresses the other two. ‘Quality’ can be defined as a strong brand, good cash flow, healthy balance sheets - all of which a tobacco company that sets its stalls outside schoolyards could happily show to potential investors. But that model is not only “morally flawed”, says Finegan, but vulnerable to the same changes in culture and regulation that we have seen in developed markets. Similarly, if management is willing to turn a blind eye to child labour in its supply chain, why expect them to treat you any better?
“Those issues can give you advance warning about who might be the wrong sort to be in partnership with when things get tough,” he reasons. “We are much more interested in companies that are part of the solution, whose business is founded on helping living standards to improve - that’s a big part of how we assess quality.”
He points to the rise of modern retail in Africa. On one level, this recognises that the growing affluent class in Lagos would rather shop in a supermarket than from a roadside shack. But it is also about how the problem of food waste can be addressed via modern cold-chain solutions and distribution efficiencies. Shoprite Holdings of South Africa - the leading modern retailer in Nigeria and Ghana - is a major position in the fund. It only dropped out of the Top 10 because its success started to get priced-in, at which point South African consumer staples firm Tiger Brands took its place; and Tiger Brands is expanding in sub-Saharan Africa because Shoprite, its major customer, needed it to supply its new stores. Slowly but surely, a whole new, efficient distribution industry spreads across Africa.
The philosophy also helps explain why the fund is overweight consumer staples but underweight consumer discretionary; many of the latter companies support sales with credit books, Finegan observes - and his team doesn’t consider a borrowing binge to be a sustainable model for growth in the emerging world anymore than it has been in the developed.
The shared benchmark-agnosticism at First State and Brandes gives their portfolios a lot of first-glance similarity. First State’s Select fund has about 50 stocks, with around half of its assets in its Top 10. Its tracking error has been around 7.2% and its beta, 0.86. Brandes’ Emerging Markets Equity fund holds 50-70 stocks, with about 30% of its assets in its top 10. Its tracking error comes in at 9-10% and its beta at around 0.90.
By contrast, FPP Asset Management’s Global Emerging Markets fund is neither benchmark-agnostic nor high-conviction. It has some off-benchmark positions, but they are much less significant than First State’s or Brandes’. As a result, its numbers could not look more different. Tracking error has been 4.8% and beta 0.97. While 24% of its assets are currently in the Top-10 stocks, the portfolio contains more than 250 positions; stocks are only excluded if FPP’s stock-weighting process allocates less than 0.2%.
“If you are going to try and outperform an index you must have a view on all the constituent stocks in that index - and most managers are not doing that,” says founding partner Jonathan Neill.
Indeed, most managers make a virtue of ‘high conviction’, ‘drilling down’, ‘wearing out the shoe leather’ - or checking out how Indonesian management treated its shareholders during the Asia crisis. Neill is having none of it.
“I am very suspicious of the company visit,” he says. “People don’t get to run major companies unless they’re very charismatic and persuasive - and I made many mistakes in the past due to being misled by those people. I take issue with anyone who says they understand everything about a company - let alone 40 of them - and even if you think you do, it’s hard to keep on finding examples that are good enough to warrant 5-10% of your clients’ assets, year after year.”
Instead, for 20 years Neill has used a 100%-systematic process, ‘FPP Appraised Value’, to rank every stock in the MSCI Emerging Markets index. The process favours companies that deliver the highest return on equity, the highest organic earnings growth and the biggest cash dividends at the lowest multiple of book value. This ‘price-to-appraised value’ ratio (P/AV) is calculated for the past five years and, using analysts’ estimates, the next 12 months and the next two years. The same three calculations are done for earnings yield - and the average of those three scores is divided by the average P/AV ratio.
These bottom-up, single-stock scores in aggregate generate an initial country weighting. Then there is a relative value overlay: country weights are corrected by favouring those countries where earnings yields are higher than short-term government bond yields. This is followed by another overlay that brings any significant active country weights back closer to the benchmark. This recommended country weighting, is then subjected to another, top-down, ‘common-sense’ overlay: countries are scored to favour those with downward interest-rate momentum, low debt-to-export ratios and positive current account balances, the lowest real currency appreciation against the US dollar since 1997 and year-to-date, and stock market underperformance over three years. Countries that score above the average on these criteria have their recommended weights implemented in-full; those that fall below average have their weights cut in half. The process is repeated monthly.
This is not a pure value strategy - ‘Appraised Value’ accounts for earnings growth and “we don’t want to be afraid to pay a high multiple of book value for a fast-growing company”, as Neill puts it. But both its bottom-up and top-down elements exhibit a significant value bias and it is certainly a long way from the thematic style that drives a lot of emerging market growth strategies. Neill dismisses ideas like buying consumer staples because of the growing emerging middle class as “a kindergarten method of investing”.
As such, FPP’s value metrics align it more closely with the Brandes Graham and Dodd style than with First State’s. Its portfolio P/E ratio is 11.8-times, and its dividend yield 3.3% (close to the index). The Brandes portfolio has 5.8-times and 4.1%. First State’s is 16.5-times and 2.9%.
When there is clear growth momentum in the markets, both FPP and Brandes have underperformed. Brandes started to sell out of its energy and resources positions in late 2006, missing the last hurrah of the pre-crisis commodities boom. “We felt they were finally trading at their long-term value,” says institutional portfolio manager Chris Garrett. The fund went on to underperform its benchmark by more than 21 percentage points in 2007. FPP’s fund underperformed by eight and 14 percentage points in 2006 and 2007, respectively, for the same reasons - re-allocating away from commodity exporter countries and towards commodity consumers.
However, Neill claims that the systematic process did not signal this move at all, and that it was an entirely discretionary superimposed decision. “It still hangs over my head,” he admits, vowing that it will never happen again. Left to its own devices, the process would have held on to those resource stocks: Neill reminds us that it will never sell a stock unless it breaks the 0.2% lower recommended weight.
But when it is not about the sheer number of stocks, where FPP otherwise diverges from Brandes it appears to result from the earnings growth and top-down elements of the FPP strategy. Gerardo Zamorano, director of investments at Brandes, says that his fund is picking up high yields from companies that they regard as having strong balance sheets because “investors are punishing some companies because their earnings aren’t growing”. As an example, while Brandes is underweight commercial banks, it is still its biggest sector allocation and nowhere near the underweight in financials at the other two funds. It is the biggest underweight at FPP, and the reason Neill gives reads like the opposite of Zamorano’s logic.
The shares of Cathay Financial Holdings sell at more than 200% FPP Appraised Value. But when Neill suggests to investor relations that this looks expensive, they point to their balance sheet and list all their assets. “I don’t care if they own every company in Taiwan, three harbours and 500 office buildings,” he says. “A classic value manager might say that’s a tremendously valuable share because the assets are worth four times the share price. I disagree. Until they start doing something with those assets for you, as a shareholder, it’s not worth anything.”
The contrast is more stark when we turn to Brazil, which is Brandes’ biggest country overweight and FPP’s biggest underweight. There are plenty of investors who will tell you about their top-down concerns about Brazil. Brandes looks through all this. “If you’re presented with a company that’s trading at 6-7 times earnings it’s tough to say that you don’t want to touch it because of macro concerns,” says Zamorano. FPP also scores Brazil high in macro terms, but punishes it for its low earnings-yield premium over bond yields and, especially, for its poor long-term score on ‘Appraised Value’. Brazil, it seems, is a country that looks good to a pure value manager, but not so good to one that has more of an eye on earnings growth.
Like most pure-value managers, Brandes does not have a top-down process for risk management. It relies on the asymmetric risk that generally comes from buying companies on low multiples. The big risk attendant on that, of course, is the value trap - and to some extent FPP’s top-down overlay helps it to avoid these traps, or at least better time its entry.
The example Neil cites is Russia in 2008 - the market just kept selling off, to the point where the ‘Appraised Value’ process was recommending a 20% weight. It was the top-down overlay, which did not like the rate rises implemented to defend the rouble, that prevented too early an entry.
Might India turn out to be a similar example? FPP’s 5% underweight against the index is almost identical to Brandes’ - and indeed, it scores very badly on ‘Appraised Value’, the part of the process closest to what Brandes does. But the initial recommended weight from FPP gets cut even further, thanks to the country’s dreadful earnings yield-to-bond yield ratio, rising rates and steeply negative trade and current account balances. “It’s been atrocious,” says Neill.
Zamorano notes that, top-down, India does indeed look like the most expensive emerging market, and he compares it with the situation in 2007 when the Brandes fund had almost zero allocated to China. But Garrett, his colleague, notes that 2011 was the first year in five that the fund had anything at all in India.
“Our investment committee, driven by the work of our research group, was beginning to see value emerge where none had existed before,” he says. “The best opportunities we’ve seen have been on the utilities and infrastructure front, but we have also looked at oil and gas, telecoms, insurance, consumer products.”
There seems to be a good chance that the Brandes fund’s India position will look much more like First State’s (a 1.9% overweight) than FPP’s by the end of next year.
This is not to say that First State’s overweight in India is particularly large. It is not - and this is indicative of another point of difference between our three strategies. First State’s biggest country overweight is South Africa - a non-BRIC market. Brandes’ biggest country overweight is Brazil; and FPP’s two biggest are China and Russia. This BRIC/non-BRIC split maps onto sector weightings in interesting ways.
Where First State does own Brazil, it is through companies like the utility Tractebel Energia, and it is the quality theme that tells again: “The parent company has a great track record of sharing profits equally,” says Finegan.
Ask Finegan about his second-biggest sector underweight, financials, and he will tell you how much he loves Standard Bank, a top 10 holding “in a great position to benefit from the growth of retail banking in Africa”, or Siam Commercial Bank, which he thinks can benefit as the Thai economy at long last enters its first lending cycle since 1998. But in the BRICs, banks he likes in India, such as HDSC, are simply too expensive, and banks in China and Brazil he regards as extensions of government, lending against projects with questionable economics on the one hand, or the hyper-cyclical agricultural sector on the other.
“You can argue that that’s precisely what a state-owned bank like Banco do Brasil should be doing,” he concedes. “I just don’t want to be a minority shareholder in it.”
One can hear similar sentiments from the Brandes team. Zamorano points out that its fund is pretty neutral on China as a country, but has an exposure that is radically different from the index in terms of sectors. “We are zero in banks and construction in China,” he says. “Loan growth has been very rapid over the past few years, with a lot of that credit heading into construction. Now you have a lot of empty apartments and over-priced units. Right now we’re finding plenty of good-value financials in other countries like Turkey, Brazil and Korea.”
The difference, you will notice, is Brandes’ turn to Brazil for opportunities. And even more illustrative of the defining difference between the First State strategy and the other two is hinted at in Zamorano’s mention of Korea - the most significant non-BRIC emerging market. This is First State’s third-biggest underweight. By contrast, although it is a slight underweight for Brandes, it remains the firm’s third-biggest single-country weighting; and for FPP, it is the third-biggest overweight behind China and Russia.
This might well be the indicator that confirms that, for all their distinctions, the latter two share a value DNA. Korea scores well on ‘Appraised Value’ at FPP, and hits the ball out of the park when it comes to the earnings yield premium over bond yields.
“South Korean shares have often traded cheaply, and people say there’s some kind of discount applied - whether that’s a North Korea discount, too much Samsung in the index, or just general distrust,” says Neill. “Maybe it is a value trap - but it hasn’t always traded cheaply and, in any case, the past is not a forecast of the future.”
Alongside a Korean equity ETF, Neil’s Top-10 holdings include a 7.1% position in Samsung Electronics (counting both ordinary and preferred shares) - an enormous single-stock overweight for a 250-name portfolio. “You mustn’t hide from the index,” Neill insists. “Some managers say we mustn’t own so much of certain shares that are big index weights. But if Samsung outperforms the Korean market and you don’t own at least the index weight, you are automatically doomed.”
Finegan does not like Samsung at all. “It’s corporate structure is complex to the point that you have to conclude it’s a deliberate attempt to make it impossible for minority investors to understand what’s really going on,” he says. “We’ve owned it, but never as much as we could have done, given its relative size in these markets.”
Admittedly, he cites the company as an example of how his quality philosophy can screen out successful companies - “our risk is that we miss some very-well-run companies like that because we are overly-focused on certain governance issues”.
Samsung has indeed outperformed strongly again since September. But he maintains that this is an exception, and in his argument that value investing in emerging markets is “probably a flawed strategy”, he offers a litany of “poor-quality, state-owned natural resource companies and Chinese banks” that make up the bulk of emerging market equity - and a large chunk of the FPP and Brandes portfolios.
“Gazprom is cheap, Petrobras is cheap,” he says. “But deservedly so. I don’t believe there are hidden jewels out there trading on very low P/Es. A purely quantitative value strategy would probably leave you with a poor-quality portfolio which would struggle in difficult market environments.”
We have heard Neill’s response to that. At Brandes, as one might expect, they mount a passionate defence of Graham and Dodd for emerging markets. “These economies are growing, but the vagaries of individual companies and their prospects are really no different from those you see in companies elsewhere in the world,” argues Garrett. “They may outperform and they may underperform - they just happen to be doing so in a context of a growing economy.”
Zamorano says that long-term value investors have to look through short-term volatility, and use it to establish positions against the over-reaction of global investors in emerging markets - citing the sell-off following the Lula election as one of the great opportunities to build holdings in Brazil as an example.
One can see the difference starkly revealed in the performance tables from Mercer. At the end of Q2 2011, all three of our strategies sat happily in the top six of their peers over three years of performance. But already the top performer, Brandes, had slipped to 117th out of 192 strategies over the three-month horizon; and FPP had dropped to 189th. First State maintained a respectable 12th position. By the time we get to the end of the turmoil of Q3, the trend is confirmed: First State has three strategies in second, third and fourth places over three years; Brandes has fallen out of the Top 10 and slipped even further down the short-term rankings; and FPP has dropped out of the upper quartile of 195 strategies altogether, even over three years.
Brandes and FPP look to have been victims of a significant change in market leadership during 2010 and 2011 - illustrated in figure 3 as the underperformance of value relative to growth. The chart suggests that we might be seeing the first hint of another turnaround in that leadership. If so, the underperformance from Brandes and FPP may prove to be the ‘negative alpha’ generated while picking out the unloved stars of tomorrow. If not, we might conclude that the ‘quality’ favoured by a strategy like First State’s is better suited to a new, more inimical global economic environment - from which not even the fast-growing emerging markets can de-couple themselves.