Genuinely active emerging market fund managers appear to be able to consistently outperform their benchmark, according to research from Copley Fund Research, which tracks 124 funds with assets of $248bn.
The findings will raise eyebrows, given the well-documented failure of fund managers globally to even match the returns of passive index-trackers such as exchange traded funds. Even Steven Holden, founder of Copley, who produced the data, admits he was “surprised” by his findings.
Andrew Clare, professor of asset management at Cass Business School in London, says “market folklore” suggested that, while “you may as well go passive for blue chips in the US and Europe,” it is in emerging markets that “active managers come into their own”.
“The general idea is that index tracking works better in efficient markets. The less efficient the market, the lumpier it is, and the greater requirement to have local knowledge, the higher returns might be for managers who can invest to analyse these markets,” says Prof Clare.
Nevertheless, he adds: “That is according to the folklore. I’ve not seen much evidence of that. On average, fund managers deserve their [poor] reputation.”
Copley’s research suggests that fund managers running global emerging market funds have managed to beat the passive alternative over both five and 10 years.
However, by far the best performance has come from “genuinely active” funds willing to build portfolios that differ markedly from the underlying index. Funds marketed as active but which may be viewed by some as “closet” trackers due to their reluctance to stray too far from benchmark weightings, have done far less well.
Copley defines “high active” funds as those that have an “active share” — the degree to which their portfolio differs from the index — of at least 75 per cent. They must also hold fewer than 75 stocks, suggesting a willingness to take reasonably punchy, concentrated bets.
Over the past 10 years, these funds have, on average, generated a cumulative return of 81.8 per cent after fees, which are typically around 150 basis points a year. The remaining “low active” funds have managed to return 45.6 per cent, while the passive alternative (the iShares MSCI Emerging Markets ETF) has made 39.3 per cent after fees, as the first chart shows.
Admittedly, the low-active funds look less impressive when compared with the benchmark, the MSCI EM index (rather than the ETF), which has produced a total return of 49.8 per cent over the past decade. However this barely dents the lead of the highly active funds and, as Mr Holden, points out, is something of a false comparison anyway as investors cannot invest in the benchmark itself, only in funds that attempt to track it and have fees and other costs.
This relationship, although far smaller in magnitude, also holds over five years: high-active funds have returned 11.4 per cent, low-active 7.9 per cent and the ETF 1.6 per cent.
Moreover, the “quality” and predictability of this outperformance has been strong. High-active funds have outperformed low-active funds in every calendar year since 2006, bar 2009, when a powerful emerging markets rally lifted pretty much all boats.
They have beaten the passive option every year bar none, although the differential has tended to narrow when EMs have performed strongly, such as in 2009 and so far this year.
“There is a real difference in performance in pretty much every time period except year-to-date, when all the rubbish has gone up as well. There has been a relief rally, and that’s not a good period for a stockpicker,” says Mr Holden.
“There is a definite positive correlation between active, concentrated funds and returns. The more active, the higher your returns are,” as demonstrated in the second chart.
A more circumspect Prof Clare adds: “High active minus passive is a fairly stable number. The only thing you would conclude from this is there does seem to be some benefit from the high active versus the low active approach in EMs.”
Furthermore, Proft Clare notes that the annual returns of the highly active funds exhibited a standard deviation, a measure of dispersion, of 32 per cent, lower than both the 35.6 per cent for low-active funds and 33 per cent for the ETF, meaning that the punchier funds also have higher returns in risk-adjusted terms.
“When you look at annual volatility, the highly active and passive are almost identical and low active are more volatile. Are [highly active managers] taking higher risk? It seems not,” he says.
One quibble raised by Prof Clare is that the data may suffer from survivorship bias, with some of the poorly performing funds that are more likely to be closed down omitted from the figures. However, any impact from this is likely to be small, given that “funds don’t tend to shut down too often”.
Nor would there appear to be any selection bias, as the 124 GEM funds in Copley’s database are chosen on the basis of how regularly they file their data, rather than on performance grounds.
This leaves two questions. Firstly, if the sustained outperformance by highly active managers is real, how is it being achieved? A fund with high active share clearly has more scope to significantly outperform the benchmark, but of course it also has more scope to underperform it as well. A positive skew is far from guaranteed.
Data from other markets suggest fund managers are no more skilled at picking stocks than the average market participant — whose average return will of course be the index return.
However, one systematic bias that highly active managers may be able to exploit is the widespread presence of state-owned (or in reality state-controlled) enterprises in the MSCI EM index.
Previous research from Copley revealed that highly active funds largely shun Chinese SOEs such as Bank of China, Agricultural Bank of China, PetroChina, China Life Insurance and China Petroleum & Chemical, better known as Sinopec, as well as Russia’s Gazprom.
This is likely to have been a winning strategy, in recent years at least. Data from Geoff Dennis, head of global EM equity strategy at UBS, suggest that in the five years to the end of 2015, the SOEs in the MSCI index on average lost 17 per cent in total return terms, while the non-SOEs only lost 4 per cent.
Moreover, between the market trough in March 2009 and the end of August this year, the non-SOEs’ return of 154 per cent was more than twice the 74 per cent generated by the leaden-footed SOEs, UBS found.
The relative absence of SOEs, particularly Chinese financials, is “definitely a factor,” says Mr Holden. Both Prof Clare and Mr Dennis back this argument.
Active fund managers may also have some discretion to buy stocks that are too small to qualify for the MSCI index, a potential advantage given the widely noted tendency for smaller companies to outperform larger ones in the long term.
The second question is, why has this outperformance not been more widely appreciated?
Mr Holden’s view is that the outperformance has to a large degree been obscured by the pedestrian record of the closet trackers, his low-active funds. Moreover, even some of the more recent analysis that has attempted to shine a light on this conundrum has erred, he argues.
Funds with an active share of between 60 and 75 per cent are often categorised as genuinely active, but Mr Holden, whose database includes the full holdings of these vehicles, prefers to lump them in the closet tracker camp.
“I always thought 60 per cent was too low a barrier,” he says. “When you look at funds in the 60s and low 70s you can tell they follow the index. They all have the same names in them.
“It’s the middle tier of funds, with active share in the 60s and 70s, that people think are active funds but are in fact not, which drag down the performance of active funds.”
So can we expect more funds to become unconstrained in an attempt to improve their performance? Perhaps not. Mr Holden says that when he talks to portfolio managers, he often gets the feeling they would prefer to be let off the leash, but their risk managers simply won’t let them go.
Follow us on Twitter @em_sqrd
Copyright The Financial Times Limited 2016. You may share using our article tools.
Please don’t cut articles from FT.com and redistribute by email or post to the web.