Managers are being rewarded for skill like never before. Constraints are being relaxed on their mandates and performance fees introduced to better incentivise the skilled managers.
It looks, on the surface, like the cliched win-win situation. The trick, though, is to identify that skill. The one constraint being increasingly unburdened from equity managers and about which there is no argument from the academics is the long-only constraint. Put simply, the argument is intuitive: allowing a manager to also short stocks provides for greater choice; it’s a breadth-of-market argument. A skilled manager should outperform more consistently, all other things being equal.
The empirical evidence to date supports the theory, however, some will inevitably argue that the universe of long/short managers is relatively small and time periods for study relatively short. The latest Mercer Australian shares survey, for periods to April 2008, shows that the median equitised long/short fund – mostly 130:30 funds but including some with more leverage than that – would have been in the upper quartile of performance for Australian long-only funds for most periods in the past five years (one, three and five years).
The common elements of these funds is that they allow a certain level of shorting and they have a beta of one. This means that, in the case of 130:30 funds, with 130 per cent long and 30 per cent short, the investor effectively has 160 per cent of their money at work.
A customised Mercer survey (see table) indicates specifically how the long/short managers fared since the credit crunch hit the markets last year, when November was the first full negative month post-crisis. This still shows outperformance by the long/shorts, although to a lesser degree than one, three and five-year periods, relative to a long-only universe of Aussie equities managers.
The point about the credit crisis is that many quant-orientated managers tended to underperform their fundamental peers for several months, prompting a lot of debate as to whether they had lost their magic altogether. In Australia, Watson Wyatt produced a client note suggesting that because of the weight of money in largely similar quant strategies, these managers may not outperform by as much in the future as they have tended to do in the past.
The two culprits in Aussie equities – Barclays Global Investors (BGI) and State Street Global Advisors (SSgA) – happen to be the largest and arguably most sophisticated of the quants in the world. They haven’t taken too kindly to suggestions their underperformance may be anything other than cyclical. Richard Lacaille, head of global active equities strategies for SSgA, points out that performance has improved since January. “We’ve lived through various market conditions for a long long time,” he says. “When you have a change of leadership, the quants tend to have a period of underperformance and then they come back.”
Rob Goodlad, the chief executive of SSgA in Australia says bluntly: “It’s not over for quants.” Morry Waked, BGI’s chief executive in Australia, says that if many people are doing the same thing with investments, irrespective of whether they are quants or fundamental managers, it will be hard for them to add value. In the US in particular, he says, there are a lot of ‘plain vanilla’ quant funds. You need proprietary research to stay ahead of the game.
The three standard factors used by many quant firms are value, momentum and revisions (which some include in momentum). Watson Wyatt estimates that between 2001 and 2006, the global quant money managed by the largest eight quant managers increased five fold. This does not include the money managed using quant tools by some large fundamental managers, let alone hedge funds which use quant techniques.
The two main elements which emerged in the US and elsewhere in November were the de-leveraging theme and the looming macroeconomic downturn. Quant analysts were slower to react than their ‘real money’ counterparts, Lacaille admits. But this does not reflect a backward-looking bias. “While we rely on patterns, we still use a lot of forward-looking measures,” he says.
Roz Amos, the head of manager research for Watson Wyatt, Australia, is measured in her comments, well aware of the sensitivities surrounding the issue. “It’s not an absolute ‘no’ from us, it’s more a ‘be cautious’,” she says. “Investors need to check the managers’ experience and see what tools they’re using. Will they be ahead of the pack in an increasingly crowded space? What will they do in a crisis? Managers need to show their risk management side is sewn up if they are going to be doing a lot of shorts.”
The manager which forged ahead of the long/short pack in Australia over the past 12 months uses both quant tools and fundamental research. Tribeca Investment Partners, which last month aligned with Grant Samuel Funds Management, has a two-year-old active extension fund (maximum 150:50) run by Sean Fenton. Fenton has a quant background with AMP Capital Investors but employs fundamental analysis as well on both the long/short fund and the broad market long-only fund which is also his responsibility for the past four years. “You often see quants underperform in periods of volatility,” he says. “But that doesn’t mean that their models are flawed. You see the comments about there being too much money in the same strategies that pop up whenever you get that volatility. But quants trade off behavioural biases that are ingrained in the human psyche. They’re unlikely to go away any time soon.”
He says that since the credit crunch, however, investors are taking more notice of the managers which use fundamental analysis. “You could argue that some quant signals are becoming self perpetuating, that the markets are becoming pre-emptive in their predictions.” Usually about two-thirds of the alpha for the Tribeca active extension fund will come from the long side, so Fenton sees the long/short fund as being “simply more efficient” than long-only.
In the selection of a fundamental approach to long/short, investors need to be wary that the manager truly has the skill set to short stocks. Analysts take time to adjust from a long-only environment and fundamental managers generally need more resources to analyse potential shorts. Van Eyk showed where its preferences lay in its latest Australian equity review, which included active extension managers for the first time. Of the five 130:30 type offerings it deemed good enough to be rated, four were quantitative-based and just one was fundamental.
While critics may question the relatively short time periods that 130:30 funds have been operating, a couple of post graduate students at Stockholm School of Economics, Carl Armfelt and Daniel Somos, set out to show, in theory at least, how these funds will tend to outperform in all market conditions over a very long time.
Armfelt and Somos used a statistical technique to analyse the performance of active extension strategies using 25 Fama-French portfolios formed on size and book-to-market valuations. The portfolios all had a beta of one and gross exposure above 100 per cent of net asset value (which means they had some shorting), back-tested to just before the Great Crash of 1929. “Our study shows that a pool of active extension 130:30 portfolios outperform a pool of long-only portfolios over the full sample period from 1927-2007,” the authors say. “Looking at shorter 10-year periods, the active extension portfolios predominantly outperform the long-only portfolios (that is, have better risk-adjusted returns). The results are especially robust for the second half of the 1927-2007 period.”
The 10-year average annual return for 130:30 funds, over the entire 80 years, was 15.9 per cent, against 14.1 per cent for long-only using the same alpha sources and 10.1 per cent for the benchmark. In the most recent 10-year period the gap was bigger: 20.0 per cent a year for long/short against 16.8 per cent for long-only between 1997 and 2007.
The argument in the academic world, almost since Richard Grinold and Ronald Kahn further defined the “Fundamental Law of Active Management” in 1994, is not whether relaxing the long-only constraint makes for more efficient portfolios but rather what the optimum level should be. In Australia, 130:30 is the norm whereas in the US, 120:20 funds are just as popular.
Tribeca’s Fenton points out that the Australian share market has a greater skew towards large-cap stocks than the US, so more shorting is required for optimum efficiency. But once you get past about 130:30 or 140:40 the increase due to portfolio efficiency starts to diminish and the increase due to leverage is more dominant. JANA Investment Advisers was an early supporter of active extension strategies and has its own long/short fund of funds for retail investors, on the MLC platform, for both Australian equities and global.
The JANA Australian equities long/short fund has three quant managers – SSgA, Perpetual’s Q1 team and Acadian – but John Coombe, JANA executive director and head of consulting, Sydney, says the firm is looking to diversify with the addition of fundamental managers. JANA has put about 10 per cent of its Australian and global equities for which it has discretion, in its implemented consulting service, into the 130:30 fund of funds. JANA’s first experience in Australia was with BGI’s market neutral fund, launched in 2001. Market neutral funds have equal longs and shorts, giving them a beta of zero.
Coombe says JANA wanted the fund to be equitised to make it more palatable to clients, which BGI did. The equitised long/short fund is the equivalent of a 180:80 fund, with the beta coming from a futures overlay for clients who want that. The fund has outperformed all others over longer periods but has been closed to new money for some time. Last year BGI launched a detuned version, a proper 130:30 fund, for existing clients. This fund has some capacity still available. There is also a version for tax-exempt clients.
The SSgA experience is more recent but no less successful in business terms. SSgA launched its Australian ‘Alpha Edge’ fund in late 2004. Goodlad says the available capacity of $1 billion was oversubscribed about two-and-a-half times in just a few weeks.
The strategy was then exported to other SSgA regions and a global version launched in 2005. The Global Alpha Edge fund has raced away to $US12 billion in funds under management, increasing three fold last year alone. Coombe says JANA did a lot of work on equitised long/short funds, after “backing” the BGI fund in 2001. “AustralianSuper (then ARF) and some others started to look around for other managers offering leverage in the Australian market. This spawned discussions with overseas managers, such as LSV, Acadian, AXA Rosenberg, Goldman Sachs and BNY Mellon, as well as SSgA,” he says. “We had a taste of it and it was good. It was like wine; if it’s good you’ll want another bottle.”
Coombe does not like the distinction of 130:30 managers being split between quants and fundamentals. “Managers are a mixed bag,” he says. “When we started we just had quants but we decided to look across the whole market. We have moved to fundame-ntal too.”
JANA put ARIA into the old Suncorp long/short fund and has other clients in the Paradice portfolio which allows up to 115:15 long/short. Coombe says that the most difficult part for a traditional long-only manager is to get the analysts to think about short stock ideas. “You have to incentivise and encourage them in different ways. Suncorp shifted the portfolio manager to the trading desk.”
JANA had hoped its exposure to long/short funds would protect it in a down market, the first test being last November. The protection did not eventuate because of the poor returns for the big quants. “We think this was because it was a liquidity event,” Coombe says. “A lot of hedge funds used quant techniques to decide how to run money. We had huge de-leveraging and as the balloon got smaller they had to sell their positions.” He says that quant strategies are not dead; they have just been through a bad period. Performance has picked up since the end of March. “If quants are dead then so is active management.”
The “mixed bag” of managers includes some quants which do not use factor-based signals, such as AXA Rosenberg, and some fundamental managers which have relatively concentrated portfolios, such as Fortis Investments (formerly ABN AMRO). Concentrated portfolios are generally considered the antithesis of long/short, certainly in the academic literature.
While purists may prefer long/short over concentrated, for funds it may come down to the practicality of the costs of generating the alpha. Coombe says they may ask: ‘is the alpha generated sufficiently high that we’re not better off with a concentrated portfolio net of fees?’ However, I haven’t met many trustees who can stomach the volatility of a concentrated fund.”
The most recent additions to the long/short universe in Australian equites are from fundamental managers: BT Investment Management, Challenger International, Colonial First State and, the latest, from UBS Global Asset Management. These, together with quant funds from Acadian and Perpetual are not included in the Mercer survey published with this report. UBS expects to launch a 130:30 fund in October, to be run by John Campbell, who works with Stephen Wood, the portfolio manager of the firm’s Emerging Companies fund. He has been running a paper portfolio which will be seeded with internal money in July.
Campbell, a former proprietary trader at BT in the 1990s, says he is well aware of the questions investors have over fundamental managers which move into the shorting space. “Our style makes us well positioned for this fund,” he says. “Our approach is to cover about 95 per cent of the index through company research and our own valuations. At any one time we have stocks which look cheap and stocks which look expensive. Conceptually, we’re not doing anything new. It’s very easy to extend our methodology.”
Risk control measures include a 3 per cent cap on absolute short positions in any one stock and the shorts will be smaller bets than the longs. While this is UBS’s first long/short fund in Australia, the firm has been running similar strategies overseas for several years, including an Asia ex-Japan 130:30 fund from Singapore, where UBS has its regional execution hub. In the US, a survey by Pensions & Investments in April showed that 53 managers ran a total of $US66 billion in active extension strategies as at March this year, which was 22 per cent up from a similar survey last September.
The survey revealed the fastest growing of these strategies was that of fundamental manager JP Morgan Asset Management which had $US8.6 billion in 130:30 assets. Doug Burton, managing director of AXA Rosenberg in Australia, says that quant investing is not its own style; it is a way of implementing the insights. AXA Rosenberg’s models are not return-based factor analyses but rather look at current prices and relate these to a company’s current fundamentals, with a second model forecasting next year’s earnings. The firm focuses on 170 different business lines, or types of business, across about 21,000 companies for its global fund.
Over calendar 2007, AXA Rosenberg’s US 130:30 fund provided a positive 7.3 per cent above benchmark. “We were in a totally different performance space to the other 130:30 managers (for US equities),” Burton says. “This reflects the fact that our insights are also very different.” The quants have to be good at implementation because they tend to have a larger number of positions and higher turnover. AXA Rosenberg’s process, for instance, results in about 500 longs and 300 shorts. “We have a lot of ideas and there are easy substitutes for them,” Burton says.
Ian Manton-Hall, chief executive of Fortis in Australia, says investors liked the firm’s equities capability (the core fund is about 150 bps above benchmark for three and five-year periods to March) but because the 130:30 fund was to include shorting it was decided to make it more concentrated. “We thought that on the short side there might be a limit to what we could do,” Manton-Hall says. “So, we thought we’d go concentrated on the long side.” The fund, seeded with internal money in 2006, currently has 25 long positions and nine short. Four out of the top 10 ‘wins’ for the manager in the past year were from shorts, the biggest being Centro.
Fortis has stop-loss positions in place, so the shorting “doesn’t get out of control”. At a 40 bps loss over a week, the stock is reviewed, and at 70 bps on the same test it has to be sold.
According to Rick Roberts, a partner at US-based quant manager First Quadrant, when fundamental managers are running 1ong/short portfolios, they need to “cede control” at the implementation stage to the quant team, for risk control and implementing the short ideas. “At the centre of the debate is the question of whether all successful managers should have active extension strategies. The answer is patently ‘no’.”
First Quadrant, one of the managers with Affiliated Managers Group of the US, has been running long/short market neutral and 130:30 funds since 1991. In Australia, where the firm has had client funds for more than 10 years, First Quadrant is well known for its global macro capabilities.
Roberts says the performance of some of the quant managers last year shows that risk management was paramount. “Risk management is often misunderstood,” he says. “Our (quants’) underperformance in August was not outside expectations.” Market neutral funds, which represent the most efficient portfolios in a world that follows the Fundamental Law of Active Management, have not yet attracted super fund interest to the extent the alpha extension strategies have.
Some believe that market neutral funds, which offer the flexibility of being able to be applied over any benchmark, will inevitably gain their rightful place and 130:30 funds are providing a stepping stone for super funds to get used to the idea of shorting. Watson Wyatt’s Amos says that 130:30 funds should really be compared with hedge funds; not with long-only funds and not with concentrated funds. “They are nicely packaged easy-to-use constrained hedge funds,” she says. “The name 130:30 is misleading. There is a vast range of funds lumped in. They have the same net exposure (one) but varying gross exposures.”
She believes that over the long term, most long-only managers will be doing some element of shorting. “At least it will be on the agenda for everyone. The theoretical case (for shorting) is strong. Pricing can also look good. Some managers are very experienced and doing (130:30 funds) well. “Once you start thinking about long/short you ask yourself ‘why be limited to something that has a forced level of shorting and a fixed level of beta?’” Pricing on 130:30 funds is usually between the long-only fees and hedge fund fees. Managers will typically charge the same base as long-only and then with a healthy performance fee (say, 20 per cent) above the benchmark or benchmark plus a bit.
Their transaction costs will also tend to be higher than long-only, reflecting increased governance for risk, higher turnover and stock borrowing costs. Managers are defensive about charging higher fees for the shorting component, however, they rightly point out that they have to manage their capacity across their range of higher-alpha funds, of which 130:30 may be just one part.
Rogers Casey, the US-based asset consultant, which is Intech Investment Consulting’s offshore information partner, produced an interesting paper on 120:20 funds last year, in which the firm suggested a benchmark which was not the broader market index but rather the long/short manager’s own long-only portfolio.
While not all managers would be able to comply with this, it would perhaps be a fairer representation of the added value through the portfolio with shorts. However, some long/short managers suggest that the long-only component of the portfolio would necessarily be different from the long-only fund portfolio because of the additional risk being taken on in long/short.
Watson Wyatt has developed a reputation among managers, at least, for being somewhat aggressive in its approach to fees, more so than other asset consultants. At last year’s annual meeting with managers, in Sydney, there was some vigorous exchanges about fees and then the firm produced a global note for clients in February where it said: “Fee structures in the asset management industry are too high for the value they offer”.
The consulting firm said that, on a global basis, total annual investment costs for pension funds increased on average by 50 per cent in the past five years – averaging about 110bps compared with 65bps in 2002. “A key reason for this is the rise in investors’ focus on alpha, which has increased their appetite for alternatives, such as hedge funds, private equity and real estate.
“Investors naturally assume that they are paying these high fees to reward manager skill, or alpha. But in most cases they are wrong. Instead they are paying alpha fees for beta performance, because the main driver of returns in recent years has been the strength of the markets.”
Watson Wyatt provided some typical examples of manager fee structures, including this long/short fund: “Take a long/short equity fund that is 100 per cent long and 30 per cent short, charging 1.5 per cent a year base fee on net asset value, plus 20 per cent of absolute performance (a fairly typical situation). Assume the manager can add an impressive 5 per cent a year alpha on both the long side and short side for every 100 per cent gross exposure (it should be noted that over the long term this would be a remarkable result), on top of a long-term annualised market return of 10 per cent and an annualised cash return of 5 per cent. On this basis, the gross annualised return would be 15 per cent, so investors pay an annual fee of 4.2 per cent, or 65 per cent of the alpha produced by the manager (4.2 per cent/6.5 per cent). In order to pay the manager only 50 per cent of the alpha in fees (the client is taking all the risk after all) the manager must generate about 7.5 per cent a year alpha on each side, and that is Warren Buffett territory!” Emotive stuff.
The Watson Wyatt note on fees also described the skewed nature of fees whereby a base is augmented by a performance fee, because managers do not repay the base if they underperform. The consultants say that performance fees should not be calculated on an annual basis but rather on a rolling three-year basis or longer.
The fee debate, of course, is not confined to long/short equity funds, let alone 130:30 funds. But it does provide another example of where super funds need to more closely scrutinise the offering from managers and, in particular, question the definition of skill implied in their fee structures.
The best-case scenario for adding limited-shorting funds to a super fund’s portfolio is that they will provide a cost-effective lift to the overall alpha generated. At worst, this lesson may be a little more expensive than other equities strategies.
We’ll give the last word to Tribeca’s Sean Fenton: “If a super fund doesn’t include shorting strategies in its portfolio, it might as well go indexed.”