130/30: understanding the opportunity; mitigating the threat

We are uniquely well placed to explain the why and the how of 130/30 in terms which should make sense and be of practical use to any investment professional, so that they can harness the opportunity.

Go into a betting shop and place a bet on which team will win a football match. By betting on who will win you are also betting on who will lose. Now imagine you could still bet on winners, but the bookmaker would not pay out on the fact that, by predicting the winner, you’re also forecasting the loser, and so reduces the odds accordingly. You’d soon find a new bookmaker who’d give you odds on your ability to identify losers as well as on your ability to select winners. Yet, in the world of fund management, traditional houses have focussed all their efforts on finding “winners” and not made any money, for themselves or their clients, from their ability to identify losers. 130/30 addresses that.

Dramatic growth of 130/30 assets

A Merrill Lynch research report in March 2007 estimated that the strategy, only a few years old, had already garnered $50bn in assets. A subsequent paper in the 4th quarter of 2007, estimated that “the total 130/30 market could reach $1 trillion in 5 years.” There’s a widely held view that 130/30 is set to carve out a large share of the global equity investment space, possibly taking assets both from traditional long-only managers and alternative long-short hedge managers. It’s both a threat and an opportunity for all players in the asset management community.

The optimal mix of alpha and beta

Countless studies have shown that, over the long term, the broad equity markets, such as the S&P500, the TOPIX or the FTSE World Index, outperform bonds and cash. That is due to what is known as the “equity risk premium” and means that portfolios that offer “beta”, or market exposure, have the potential, over the long term, to offer better returns than cash or bonds. However, index trackers, which offer pure undiluted beta, lost appeal, especially in the 2000 to 2003 bear market, because investors wanted even better returns and less downside risk. This, a return in excess of beta without any incremental risk, is described as “alpha”. With ever greater demands on pension funds and personal savings, a strategy that has the potential to deliver “beta plus alpha” without taking on incremental risk is naturally attractive. That’s the appeal of 130/30.

The numbers in 130/30 represent the percentages, measured as a proportion of the money invested, which is deployed to gain long exposure to the market and short exposure to the market. So, a traditional long only fund could be described as 100/0, in that it’s 100% long and not short at all. In money terms, if $100m was invested, a 130/30 portfolio would hold $130m of long positions and would take $30m of shorts.

This approach, with a small proportion of shorts offset in money terms by an increase in the value of long positions held, means that a 130/30 strategy is always 100% net long, and will have a beta of around 1. This implies that the portfolio will generally participate in broad market moves, and so will benefit from equity market gains in the long term. However, in addition, profits can be made from the manager’s stock picking alpha, as the “winners” selected on the long side outperform the “losers” on the short.

The EU’s recently launched “UCITS III” rules, the gold standard for mutual funds, permit 130/30 funds to be sold to the mass public. With such endorsement, it’s probable that more restricted long only funds and less transparent long-short hedge funds will find a strong competitor in 130/30.

Increased ability to take meaningful positions

Many traditional managers would argue that they do not go short. That may be technically correct but only if their positions are measured relative to a cash holding. Since long only funds are benchmarked against a market index, the true measure of a fund manager’s decision making is the difference between the weight a stock represents in the portfolio and the weight in the benchmark index, what is known as the “active weight”. These active weights are often negative – i.e. the manager chooses to underweight a stock that is expected to underperform, so that the portfolio has a lower weight in a stock than the stock represents in the index, and the portfolio is effectively “short” when measured by comparison with the benchmark,.

In practice, the ability of long only managers to go short in this way is limited to a very small number of stocks. In graph 1, we see the distribution of stocks in the UK’s FTSE All Share index by the number of companies in categories representing different levels of weight in the index. Out of nearly 700 companies in the index, only 23 have a weight in excess of 1%, and only 10 a weight in excess of 2%.

This phenomenon is not specific to the UK. In Table 1 we see that the same pattern, though to varying degrees, applies to other leading long benchmarks.

In order to have a meaningful impact on relative performance an active position of reasonable magnitude must be taken. Yet even a modest negative 2% active position is, for a long only portfolio benchmarked against the FTSE All Share, possible for just 10 stocks. A 130/30 strategy, on the other hand, similarly benchmarked against the FTSE All Share, has a universe of many hundreds of stocks on which negative 2% active positions may be taken. Since long only managers are already comfortable with taking negative active positions, and believe that they can improve performance from so doing, it is clear that giving them more stocks on which they can take such bets has the potential to add significantly to incremental performance.

Is 130/30 the best ratio?

The explanation above makes it clear that 130/30 offers a better range of opportunities to add relative performance than 100/0. But why stop at 30% short? In practice, a 130/30 strategy will, from time to time, allow for a bigger or smaller weight in shorts, while staying approximately 100% net long. Nevertheless, there are two reasons why a balance of longs and shorts around 130/30 is likely to offer the best risk-return ratio over time.

Firstly, to go short it is necessary to have borrowed the stock. This has a cost in the form of a lending fee. Consequently, shorting is more expensive than going long, and so the expected relative performance of the stock must be greater to justify the trade. Assume the market is evenly split into stocks expected to outperform and underperform. The greater costs of shorting will lead to the set of short candidates – those with large enough expected underperformance to justify trading – being smaller than the set of long candidates. Secondly, while the manager has more potential to add alpha by being able to short stocks, his alpha may for a while be negative. This can be accommodated without excessive pain in a 130/30 because the returns are still dominated by the market exposure. Alternatively, in a 300/200 strategy, although the portfolio would still be net 100% long, the total exposure of 500% indicates a far greater proportion of returns expected to come from alpha. If alpha were negative, it could outweigh the long term positive of market exposure and lead to long-term losses.

Clarke, de Silva, and Sapra, in the Journal of Portfolio Management in 2004, tested the impact of relaxing various typical portfolio constraints, e.g. limits on stock positions, market capitalization, industry or sector and the long-only constraint. Their analysis indicated that the long-only constraint was the most significant in terms of inhibiting the impact of the manager’s alpha on portfolio returns. In addition, they found that 130/30 strategies achieve 90% of the benefit that could be obtained with a fully unconstrained 200/100 strategy. Thus, most of the benefits of relaxing the long-only constraint can be achieved with moderate levels of shorting.

Using shorting to enhance alpha without increasing risk

A simple arithmetic example shows how a limited amount of shorting can enhance return potential without increasing risk. All stock returns are assumed to be partly driven by a degree of market exposure which we will call “B”, as well as by a stock specific element which we will call “A”. In a portfolio, the B’s add up and create volatility, or risk while, because the A’s are stock specific, their risks do not add up and can be diversified to hold risk down, although their returns will still be additive.

An attractive stock will have beta and a positive alpha, which we could write as B + A
An unattractive stock will have beta and a negative alpha, hence B – A. If we hold both stocks in a long portfolio, with the one expected to outperform with the larger holding, say 70%, the portfolio adds up as follows: 70% of (B+A) plus 30% of (B-A) which equals B + 0.4*A. If, on the other hand, we hold the attractive stock long and the unattractive short we get: 130% of (B+A) minus 30% of (B-A) which gives B + 1.6*A. Hence, a long only portfolio is dominated by beta, while a 130/30 portfolio has the same beta but much more alpha and so gets a greater benefit from manager skill.

Here’s an imaginary example with two stocks, where we consider yield to be our measure of alpha, and we are seeking to increase it, but without raising our exposure to market risk, in this case represented by sector exposure to oil.

Quite simply, our sector exposure is exactly the same in both portfolios, but by being able to go short the stock with the inferior yield, we are able to significantly boost the average portfolio yield.

This can be further illustrated with an efficient frontier as shown in graph 2:

As a long only portfolio seeks to increase exposure to an attractive stock characteristic, yield in this example, risk soon increases as all the best candidates will probably share some common characteristic. For example, the best yielding stocks may all be financials, so a high yielding long only portfolio will take on a lot of financial sector risk. However, a long short portfolio has much greater capacity for increasing exposure to the attractive characteristic without increasing risk as it can keep sector weights closer to the benchmark while, within any sector, holding the high yield stocks (in this example) long and going short the low yield stocks.

Fat tails and unlimited liability

Historically, going short has been seen as riskier than going long because the most that could be lost with a long holding is the value of the investment, whereas with a short there is in theory no limit to how much could be lost. For example, a $100 long investment could go to zero, losing $100. But if one went short the same stock it could rise to $200, $500 or even $1000, losing $100, $400 or $900 – nine times the loss on the long side. Although theoretically correct, in a professionally managed portfolio this is not a realistic concern as it would be diversified so no position would represent a large part of the whole and because the manager would be alert to changes in stock prices and would close the position rather than allow the losses to grow to intolerable levels.

It would be more realistic to think in terms of one week as the “reaction period” i.e. the length of time that the manager may take to reach a decision to close out a trade. In practice, it is only losses over this period that an investor faces, in that they are happening too fast for the manager to react to. To illustrate this, Graph 3, below, shows the distribution of weekly returns for all stocks in the FTSE All Share over the last ten years.

Graph 3. Distribution of weekly returns for all stocks in FTSE All Share, 1997-2007

The graph shows that the vast majority of weekly returns lie between plus and minus 10%, and that the proportion of stocks showing positive returns is similar to the proportion showing negative returns, so that going long and going short should be equally risky. A closer examination of the graph shows that it has “fat tails” in that, in the regions outside plus and minus 20%, the line showing what has been observed in practice over the last ten years is much higher than the line representing a normal distribution. To the extent that there are more extreme positive observations than extreme negative observations it would be true that going short (when you lose money if the stock has a return of over plus 20%) is riskier than going long. However, the last ten years have been generally positive for the UK equity market, with no serious economic downturn in the whole period. Consequently, many stocks have been subject to takeovers, when one would expect an extreme upside move, and few have got into serious economic difficulties, when one would expect an extreme downside move. If economic conditions were to change, it is quite possible that, considering one week holding periods, being long would be riskier than being short.

Risk control issues

The efficient frontier above shows that a long short portfolio can offer the same returns as a long portfolio with less risk, or greater potential returns for the same risk. However, the portfolios on that frontier were built using a risk analysis system and, without such a system in place, there is no guarantee that the potential efficiencies of 130/30 can be extracted. Indeed, there is a risk that sector and style bets might be exaggerated in a long-short portfolio, leading to a large increase in risk.

Indeed, in the papers referred to above, Merrill argued that "Moving away from long-only investing requires more sophisticated systems, with a much heightened stress on risk measurement". The UCITS rules agree with this analysis as, in exchange for allowing managers increased investment powers, they impose upon them an obligation to develop and apply a formal risk management process.

Such a process will include a system that is able to consider the relationships between stocks, to measure the extent to which a particular stock’s performance is driven by market-wide, or systematic, factors and what can be attributed to stock specific effects. This is because, to keep risks under control while allowing, as in a 130/30, a greater total value of stock positions than the investment made, it is necessary to reduce the proportion of risk that is systematic and additive, while increasing the proportion that is specific and which benefits from diversification. This is shown in Graph 4 and table 2 below, where a characteristic associated with outperformance, the manager’s alpha, is increased in the 130/30 portfolio while keeping the expected risk level constant.

Regulation and operation

A hedge fund manager seeking to compete in a broader market with a UCITS III 130/30 fund will need to deal with unfamiliar regulatory issues, while a traditional manager will need to address the operational aspects of shorting. While neither firm will necessarily have the expertise in-house, the challenges are not significant and outside advice is readily available from risk consultants, prime brokers and administrators.

Stock borrowing has become such an established activity that many administrators and prime brokers offer turn-key solutions. After the short has been executed, there are no further material complications as pricing can be sourced in the same way as for a long position and liquidity is readily available. Unlike many derivative or exotic strategies, there are no “greeks” to worry about. If using a prime broker to execute the shorts, possibly in the form of “Contracts for Difference” (CFDs), account must be taken of counterparty risk, but this can easily be regulated by marking to market on a monthly or other basis.

Many managers may consider the hardest operational issue to be the generation of ideas for stocks to go short. Even this can be addressed with an appropriate risk system, as negative expected alphas can be implied from a manager’s long only bets and such a system can select shorts on a risk reduction basis alone.

The future is net long and partially short

“130/30” portfolios combine a traditional long-only portfolio with a 30% long / 30% short portfolio. They could be considered to be high-conviction long portfolios, which theory argues should produce alpha more efficiently, while still participating in the equity risk premium. US based quantitative managers have been early providers, but fundamental managers and hedge funds, in the US and elsewhere, are increasingly offering or considering the strategy. 130/30 funds in a regulated environment, such as UCITS III, have significant potential as they offer some of the investment efficiencies of a hedge fund combined with the transparency and risk control of a regulated fund.

Now that shorting has become an accepted investment strategy, it is likely that funds that deploy a degree of shorting will take an increasing share of assets under management, and the successful firms will be those that are able to manage the risks of shorting to achieve what theory promises. If you can choose winners you can choose losers, the secret is to get the odds on your side.

About David Androsoni

He is a Senior Consultant at EM Applications, where he advises investment firms around the world on the implementation of quantitative techniques into their investment processes. He has 7 years of experience in the field of Risk Management and Quantitative Finance, spanning many investment styles across several asset classes. David is a member of Inquire Europe.