130/30 Funds: What is Behind the Commercial Offensive?
Walter Géhin, Research Associate with the EDHEC Risk and Asset Management Research Centre, and Business Analyst with Atos Euronext Market Solutions
High-conviction funds, beta-one funds, short extension funds, limited-shorting funds, long-enhanced funds, active extension funds, hedge-fund lite: there is a wide range of terms for what is most frequently called 130/30. Broadly, this strategy initially invests 100% in an index, sells short 30% and uses the proceeds from the shorts to buy an additional 30% likely to beat the benchmark.
In the United States, these funds, having been created only three years earlier, had already pulled in US$50 million or more as of late March 2007 and US$100 billion by August of the same year1. According to Morgan Stanley2, assets under management in Europe amounted to US$5 billion in May 2007, and it forecasts an increase to US$40 billion by the end of 2007.
Wilshire Consulting notes that of a list of 80 funds “only about a third of these strategies have a track record with money in the product” and that “only one strategy has a track record dating before 2004”3. In spite of recent launches and track records too short to draw conclusions from, the 130/30 industry is attracting strikingly large inflows. The last several months have seen one 130/30 launch after another (Table 1).
In the early days, 130/30 funds were aimed mainly at pension funds. State Street Global Advisors’ Global Alpha Edge, one of the early funds on the 130/30 market, has cemented its leadership by pulling in money from pension funds – most recently, for example, from the Asda Group pension fund. In the same vein, the Teachers’ Retirement System of the State of Illinois “plans to allocate 10% of its domestic equity portfolio to a 130/30 large-cap core strategy within the next two years”4.
The introduction of UCITS III rules and shorting options using synthetic products have made it possible for 130/30 fund marketers to target retail investors. This is the case, for example, of UBS Key Selection Sicav - US Equities 130/30 B, with a minimum investment of one unit (priced under US$10 in August 2007), ThomasLloyd’s Long-Short Equity Fund, Robeco’s 130/30 European Equities Fund and MainStay’s 130/30 Core and 130/30 Growth funds.
In this paper, we examine some crucial points related to these funds: their theoretical foundation, the optimal level of shorting, the distinction between the quantitative and fundamental approaches, whether these funds are natural extensions of long-only funds, and finally the risk of neglecting their risk.
Theoretical Framework and Simulations
By relaxing the limit on shorting and extending the long side, 130/30 managers can take larger underweights and overweights – with respect to a benchmark – than long-only managers can; in other words they benefit from enhanced active weights. They can then better exploit their positive and negative views. Shorting is of particular interest when the manager wants to significantly underweight a stock that has a low weight in the benchmark. For this reason, 130/30 funds are sometimes called high-conviction or active extension funds.
Beyond this first development, it is interesting to consider the theory behind the creation of 130/30 funds. In terms of performance measurement, the increase in the information ratio is often highlighted by 130/30 promoters. The source of this focus on the information ratio can be found in academic papers published by Clarke, da Silva, and Thorley (2002)5 and by Clarke, da Silva, and Sapra (2004)6.
The information ratio IR is defined by Grinold and Kahn (2000)7 as follows:
E(RA) is the return of active management,
is the active risk of the managed portfolio, IC is the information coefficient corresponding to the correlation of security returns and actual outcomes, and measuring the forecasting skills,
and N is the number of securities in the universe of investment.
Clarke et al (2002) add a transfer coefficient TC:
That the information ratio as defined by Grinold and Kahn (2000) does not exactly match the reality, where there are constraints on active weighting, justifies their addition of the transfer coefficient. It measures the proportion of the signals transmitted by security rankings that are transferred into active portfolio weights, thus allowing the constraint on short positions, among other constraints, to be taken into account.
In various tests, Clarke et al (2004) discover that the removal of restrictions on shorting generates an increase in the transfer coefficient from 0.33 to 0.68. If we apply this finding to comparisons of 130/30 and long-only managers, we can see that 130/30 managers, generating a higher information ratio, can better exploit ranking signals.
In a Morgan Stanley research paper, Leibowitz and Bova (2006)8 examine the impact of long and short extensions on alpha, through the move from a long-only portfolio (with 25 long positions) to a 120/20 portfolio (with 42 long positions and 8 short positions). Securities are included in the portfolio on the basis of their alpha ranking.
It is noteworthy that the addition of the eight new short positions (the short extension) generates an alpha increase higher than that generated by the 17 new long positions (the long extension). While the short extension increases the alpha from 1.85% (alpha of the long-only portfolio) to about 2.75%, the 120/20 portfolio generates an alpha of 3.08%. The ratio of alpha to tracking error for the 120/20 portfolio is 0.63, as opposed to 0.54 for the long-only portfolio. These simulations indicate that alpha increases more quickly than tracking error when active extensions are added.
1X0/X0 Rather Than 130/30
In the US, the Regulation T limits gross exposure to no more than twice the investment capital. European Union regulations likewise restrict market exposure to 200%. So shorting is limited to 50%, corresponding to a 150/50 fund.
Managers thus have a range of configurations at their disposal, from 110/10 to 150/50. 130/30 is restrictive and is more appropriately referred to as 1X0/X09. The next question, logically enough, is what determines the optimal level of shorting.
Liodakis (2007)10 enumerates the different determinants of the shorting level, and examines their mechanisms. He identifies variables that are within the control of the fund manager, called endogenous variables, and those that are not, called exogenous variables. Of the endogenous variables he studies, there are, in particular, the tracking error target, the cap size of the stocks on which the manager has negative views and the number of stocks in the fund. Exogenous variables are the manager’s skill at selecting winners and losers, market conditions and costs, and turnover.
A series of simulations suggests that a high level of tracking error is associated with a high level of shorting. In other words, a fund that targets high tracking error must implement high shorting. But the results of these simulations may also require funds with high shorting levels to win investor approval for high levels of tracking error. In the same vein, Alford (2006)11 reports results about the relation between alpha and tracking error. For a given level of tracking error, “the benefit of going from long-only to 120/20 is larger than that of going from 120/20 to 140/40”. Worded differently, as the level of shorting increases from 0 to 40%, the alpha continues to grow, but less and less. He also notes that “higher levels of risk and return require greater amounts of shorting”12.
A manager who focuses on small-cap securities to find shorting opportunities will benefit from a high level of shorting to a larger extent than will a manager focusing on large-cap securities. The reason is that small-cap securities have low weights in the benchmark, and shorting small caps magnifies the negative views expressed by the manager.
Correspondence with skills at selecting winners and losers is not insignificant. Keeping in mind the alpha generated at different levels of shorting, simulations show that the optimal level of shorting is 44% for a manager better at picking sells; 40% for a manager symmetrically skilled; 26% for a manager better at picking buys; and 19% for a manager with no skill in picking sells. Not surprisingly, the highest alpha is obtained by managers with symmetrical skills who short at their optimal levels.
Market conditions can, of course, limit the positive effects of using short extensions.
Finally, the positive effects of shorting can be wiped out either by the turnover costs that long and short extensions imply or by the borrowing costs of shorting.
With such a wide range of variables, some of them beyond the manager’s control, affecting the required level of shorting, the supremacy of one level of shorting cannot be established. It is not surprising that some 130/30 funds have a flexible level of shorting. New York Life Investment Management LLC notes13 about its Mainstay 130/30 Growth fund: “The Fund will generally hold long positions equal to approximately 130% of the Fund's net assets and short positions equal to approximately 30% of the Fund's net assets. However, the long and short positions held by the Fund may vary over time as market opportunities develop. The Fund's long positions may range from 120% to 140% and its short positions may range from 20% to 40%.”
Quantitative or Fundamental Approach?
For stock selection, 130/30 fund managers can take the same approaches taken by long-only managers: the quantitative and the fundamental approaches.
The quantitative approach makes use of proprietary computer programmes that analyse prices and volumes. Running on a pre-selected universe, by capitalisation size, say, these programmes rank stocks in order of their attractiveness. In the context of 130/30 funds, the quantitative approach has the advantage can be applied easily to an extensive range of stocks, including those to be shorted, without requiring modification of the programme used by long-only teams. In other words, this method is scaleable. Stocks that are ranked at the bottom by the programme, while they are ignored by long-only managers, are likely to be sold by 130/30 fund managers. For example, the 130/30 Limited-Shorting Strategy commercialised by Axa Rosenberg uses a proprietary model that scans and evaluates a universe of more than 19,000 stocks every three minutes. SSgA’s Global Alpha Edge works on 2,000 stocks14. Goldman Sachs Asset Management ranks about 6,500 stocks.
The aim of the fundamental approach is to analyse indicators such as market share, accountancy results, product lines and human resources to calculate intrinsic values. This approach cannot be systematically applied to each stock of a given universe. Consequently, it is less scaleable than the quantitative approach. Moreover, in the context of 130/30 funds, short signals are not the same as long signals. Consequently, to identify shorting opportunities, it requires managers with fundamental expertise.
The purportedly easier adaptation of the quantitative method to the short side undoubtedly explains why the vast majority of 130/30 funds are driven by quantitative principles. According to the executive director for Morgan Stanley Prime Brokerage, New York, quoted in a Pensions & Investments article, “quantitative managers have won about 80% of the mandates for active extension strategies”15.
Some large 130/30 funds, however, take a fundamental approach. JP Morgan Asset Management, for example, has amassed US$1.5 billion in two fundamental 130/30 funds and was the ninth largest manager in terms of assets under management according to a Pensions & Investments ranking for late March 2007. One of their managers declares in European Pensions & Investment News: “If you have a structured ranking process, which we do on our fundamental research side, it is possible to create a 130/30 product without a quantitative process. It is more natural in a way, because they automatically have stocks that will go into the portfolio and stocks that will not”16.
Other funds have opted for a combination of fundamental and quantitative methods. Quoted in an Investment Advisor article, the manager of the Maverick Capital 150/50 fund says that “although quantitative tools play an important role in our process, our investment decisions are driven by bottom-up, fundamental research”17.
Are 1X0/X0 Funds a Natural Extension of Long-only Funds?
1X0/X0 funds have some characteristics in common with long-only funds. Alford (2006) states their active weights and their full market exposure. “Active weighting”, as mentioned above, refers to the practice, frequent among long-only and long/short managers, of underweighting and overweighting stocks relative to a benchmark. Like long-only funds, 1X0/X0 funds are fully exposed to market risk; in other words, they display a beta of one. As a matter of fact, the net market exposure – the long positions less the short positions – equals the initial investment.
Moreover, 1X0/X0 funds, like long-only funds, take a relative return strategy. 1X0/X0 funds are subject to benchmarked management, where indicators such as tracking error are monitored.
Are these traits in common enough to make 1X0/X0 funds natural extensions of long-only funds? The use of shorting in 1X0/X0 funds – even limited – suggests that they are not.
First, the skills required to identify shorting opportunities are purported to be different from those for identifying long opportunities. Some commentators argue that the signals given off by securities on the way down are different from those given off by securities on the way up. Thompson, Siegel & Walmsley LLC (2007) notes that “the time horizon with short selling is much shorter”18. The quantitative method may be a solution to this skill problem. Again, stocks ranked at the bottom by the quantitative programme – while ignored by long-only managers – are likely to be sold by 1X0/X0 fund managers.
In theory, the potential loss from a short position is unlimited. As a result, investors must weigh their risk tolerance when including this parameter in their decisions. The risk profile of a 1X0/X0 fund also depends on the neutrality of its X0/X0 segment. Pairs of long and short positions that are neutral in terms of sector, currency and/or capitalisation diminish the risk of the long-short extension.
Central to the requirements on the use of shorting is the mandatory selection of a prime broker and the negotiation of a contract where lending rates and fees are stipulated. According to a SSgA’s note, “Positions must be marked to market on a daily basis taking into account any dividend. Changes in status, such as corporate actions, mergers or spin-offs, must be properly handled”19. Counterparty risk is also introduced, because the prime broker could become insolvent.
Thompson, Siegel & Walmsley LLC (2007) highlights administrative costs and larger transactions costs as a result of higher turnover. Leibowitz and Bova (2006) remark that “if shorting costs become too high, the resulting alpha degradation would eliminate any benefits from short extension”. In a nutshell, the management of costs generated by shorting is critical to the success of 1X0/X0 funds.
It is well established that before any investment in a 130/30 fund investors should verify that the 1X0/X0 manager has experience in shorting and in working with prime brokers. Managerial experience is even vaunted by 1X0/X0 promoters in their marketing arguments. For example, AXA Rosenberg emphasises its 17 years of experience in shorting stocks in the US and worldwide. Maverick argues that its “investment team has over 230 years of experience in shorting stocks, while many 130/30 managers come from a long-only background and will be shorting for the first time”20. In a press release announcing the launch of three 130/30 funds, Invesco boasts of its 14 years in shorting, beginning with its first market-neutral product in 1992.
The Risk of Neglecting 130/30 Risk
A 130/30 fund manager, quoted by Financial News, declares: “I have a little worry that people investing in 130/30 thought they were hedged portfolios”21. Here, the risk that investors underestimate the risk presented by 1X0/X0 funds is underlined.
Among possible sources of confusion is the use of the label “hedge fund lite” and of such expressions as “a bridge between our long-only portfolios and our market neutral products”22. A consultant argues that “these strategies offer access to the general concept of hedge fund investing under the watchful eye of the large, name-brand asset managers”23.
Considering their intermediate level of shorting, it is not erroneous to say that 1X0/X0 funds fall somewhere between long-only and market neutral funds. Unfortunately, these claims and these labels may – for less informed investors – blur the boundaries between 1X0/X0 and market neutral funds. In practice, the goals pursued by 1X0/X0 funds and market neutral funds when they use shorting are fundamentally different. While most hedge funds use shorting as a means to hedge their bets or limit their exposures, 1X0/X0 funds use shorting to improve their performance over a long-only benchmark. The possible confusion about neutrality certainly stems from the fact that 1X0/X0 funds have a X0/X0 component. This 1X0/X0 component is, as we have said, neutral under a range of conditions (sector, currency, capitalisation). But in the end, even if the X0/X0 component is fully neutral, 1X0/X0 funds, like long-only funds, have a 100% net market exposure. When examined for neutrality, it turns out that 1X0/X0 funds do not fall somewhere between long-only funds and market neutral hedge funds; they are at the same point as long-only funds; they are beta-one funds.
Returns posted by several major 1X0/X0 funds in the wake of the recent subprime crisis give some insights on the reality of their risks. As reported by Pensions & Investments, Independence Investments’ “130/30 strategy underperformed its benchmark, the Russell 1000 index, by 237 basis points in July, and Deutsche Asset Management’s large-cap growth 130/30 strategy underperformed its benchmark, the Russell 1000 Growth index, by 217 basis points”24. Financial News25 reports that “Goldman Sachs Asset Management’s US large-cap 130/30 fund lost 4.4% last month26 and is understood to have been down by as much as 8% for the month to August 9”, and that “Barclays Global Investors’ US equity 130/30 product was down 4% for the first seven trading days of August. The S&P 500 was down 0.9% over the same period”.
Broadly speaking, the label “hedge fund lite”, is, because of the heterogeneity of hedge fund strategies, too vague. In the hedge fund galaxy, the long/short equity strategy is certainly the strategy that most closely approaches that of 1X0/X0 funds. As a reminder, long/short equity funds invest in both long and short equity portfolios. They do not seek a permanently neutral position in terms of market risk and they are generally characterised by a significant correlation with major stock indices.
In FT Fund Management, Bluerock Consulting’s chairman declares that 1X0/X0 funds are a “wonderful product for the product provider”27. These remarks are certainly too critical. Their merit, however, is that they emphasise that 1X0/X0 funds are by no means miraculous products generating returns close to those of hedge funds while running risks closer to those of long-only funds.
The skyrocketing amount of assets under management and numerous new launches show that an increasing number of investors are placing their hopes in this strategy. But blind trust is inadvisable, and it behoves the investor not to leave certain points unexamined, in particular, the experience of managers and institutions in shorting, as well as their ability to adjust shorting levels to cope with market fluctuations.
1X0/X0 funds are clearly not natural extensions of long-only funds. So investors who traditionally invest in long-only funds may be disappointed by increased volatility and periods of larger losses.
These funds are no more or less attractive than long-only funds or hedge funds. Their appeal depends simply on the investor’s appetite for risk. The attributes of 1X0/X0 funds make them medium-risk funds between long-only funds and long/short equity hedge funds. Longer track records and observation of their behaviour in different market environments will ultimately decide exactly where along the continuum between traditional and alternative funds these funds will fall.