No reason to continue the long-only route

Directional long-short strategies – also known as ‘active extension’ or ‘130/30’ strategies – enable investors to hold short positions in moderation in traditionally long-only mandates. In case of a 130/30 strategy, for example, the manager invests up to 130 per cent of available capital in long positions while taking short positions of as much as 30 per cent, so that the net equity exposure equals 100 per cent of available capital.

The rationale behind directional long-short strategies is simple and intuitive. Investors with proven skills in stock selection often have a symmetrical skill in understanding which stocks to avoid or sell short.

Figure one shows a theoretical investment model’s historical stock selection skill. Based on this model, the most favoured quintile of stocks outperformed the benchmark by 3 per cent while the least favoured stocks trailed by 3 per cent. In an unconstrained long-only portfolio, the manager may significantly overweight the top quintile stocks by, for example, 200 per cent or 300 per cent or more of the benchmark weight but can only underweight the lowest quintile by a maximum of 100 per cent. Therefore, long-only portfolios may limit a manager’s ability to take full advantage of the information in their investment process – whether quantitative or based on experience and judgment – by limiting the size of bets against expected poor performers. This is commonly referred to as the ‘long-only constraint’ and applies to most institutional equity mandates.

In a portfolio that allows short selling, the manager may take equally significant bets in favour of expected outperformers or against expected underperformers.

For example, within the MSCI Europe only 53 stocks have a weight above 0.5 per cent (as per end of June 2007). In this way, long-only portfolios can prevent managers from taking full advantage of the information available to them. While the long-only constraint forces some ‘dead weight’ in the portfolio, relaxing this constraint leads to a better utilisation of risk through the more efficient use of information. In the long run, this should lead to a superior risk-return pattern in comparison to traditional strategies.

Directional long-short and equitised market neutral

Directional long-short strategies, while relatively new to the institutional investment scene, are very similar to the market neutral strategies that institutional investors have used successfully for well over a decade. In a pure market neutral strategy, conceptually, the manager may invest up to 100 per cent of the portfolio in favoured long positions, then sell short as much as 100 per cent of the portfolio in stocks expected to underperform. The net exposure of the account is therefore ‘neutral’ to market movements because any upward or downward movement of the market will have, for the most part, offsetting impacts on each side of the portfolio. If the market rises, for example, normally the performance of long holdings will be positive while that of shorts will be negative. The performance of the account (net of longs and shorts) will be attributable to net exposures or ‘bets’ made by the manager. Depending on the manager’s style, these bets may be related to a factor tilt such as a value bias. Alternatively, they may be attributable only to expected return differences of the long portfolio versus the short portfolio based on a manager’s investment process with the elimination of all extraneous factor exposures.

To gain market exposure in market neutral strategies – or to ‘equitise’ the market neutral portfolio – investors typically overlay their market neutral portfolio with a 100 per cent portfolio exposure in long equity futures contracts of the benchmark. For US equities, this is most commonly the S&P 500 Index. With futures exposure, the equitised market neutral strategy consists of two key components: a 100 per cent index exposure obtained from futures plus the manager’s value added or alpha (the net performance of the longs and shorts). In the case of directional long-short strategies, the net long exposure arises from the significant portfolio bias to long holdings (130 per cent) versus short exposures (30 per cent). These strategies, therefore, do not use futures contracts in their process. Figure two illustrates the basic structure of an equitised market neutral strategy versus a 130/30 directional long-short strategy.

From 110/10 to 150/50

Analysts have been studying several variants of directional long-short strategies that generally range from a 10 per cent to 50 per cent short exposure (respectively called ‘110/10’ and ‘150/50’ strategies). A ‘best’ strategy among the possible combinations of long and short positions does not exist. Rather, there is usually a preferred level of shorting or leverage that works best for a particular set of needs. When deciding on the most appropriate degree of short selling, one should consider the client’s tracking error tolerance, his benchmark and the inherent investment approach. Other factors such as limitations on holdings, turnover, or active sector, stock, or industry constraints may also play a role.

Quantitatively oriented managers can offer inherent benefits in directional long-short mandates. First, they are usually experienced in measuring portfolio risks and in constructing portfolios that optimally balance return and risk. Second, as many are experienced at managing market neutral strategies, they already have the necessary skills to manage directional long-short strategies.

Traditional long-only managers may find this environment challenging. While being able to identify short sale candidates, few have the tools to ‘test’ their process historically or illustrate examples of actual, relevant past performance. Being used to selecting purchase candidates only, traditional managers may find it difficult to draw up a substantial list of shorts, a task which becomes even more difficult if they try to intuitively balance the risks of long and short positions.

Systematic risk controls may be a relatively new concept to these managers, and those that invest the portfolio in deep value stocks while shorting a list of richly priced growth stocks may encounter rocky performance until their strategy successfully plays out. This particular approach would have been painful to investors during the tech bubble of the late 1990s, and although it eventually worked, one wonders how many investors would have replaced the manager before the strategy had been vindicated.

Volume and scale may also present a real obstacle to traditional managers. Many hedge funds managed by traditional ‘stock pickers’ create short exposures by selling short sector-based exchanged-traded funds (ETFs) against their fundamentally selected long portfolio. This type of short structure is very different from the directional long-short construct explained earlier. We noted that a key benefit of directional long-short strategies was the ability to take advantage of ‘negative expected alpha’ information gathered from the investment process and shorting sector ETFs does not actually achieve this objective.

Finally, directional long-short strategies introduce a host of operational complexities for managers. Compared to building long-only portfolios, the construction of a long-short portfolio is much more involved from an operational, legal and trading point of view. Experience and skill in this area are crucial, as is scale expertise. The ability to borrow the necessary securities for a short sale, along with the increased complexity in reporting, custodial, and clearing functions makes a $500m (e372.5m) short portfolio significantly more complex than a USD 50 million short portfolio.

The bottom line – performance expectations and the unexpected

Based on our simulations and those of other firms, we expect that a typical 130/30 strategy may add anything from 50 to 150 basis points of performance relative to its long-only counterparts. That said, one should keep in mind that these are rough estimates that may be considered either too conservative or too aggressive. Nevertheless, while admittedly the industry has limited experience in this area, many academics and practitioners agree that relaxing the long-only constraint should benefit investors in the long run with regard to enhancing returns, risk reduction and improvement in risk-adjusted performance.

The issue of what can go wrong gets little coverage in new product discussions. When looking at strategies with complex operational infrastructures, the initial risk-related concerns often focus on a systemic or localised operational breakdown. While such issues should not be discounted, we believe that a more common – indeed highly likely – risk these strategies present is periods of underperformance. Directional long-short strategies, like any other active strategy, are managed to an investment process designed to add value. However, like all investment strategies they can also experience weak relative results, and, at times, extended underperformance. Since these strategies are, in effect, leveraged to a given approach, their periods and intensity of underperformance can be difficult to endure. Another concern is that the active return of the longs and shorts may or may not be directionally the same. The longs may add value while the shorts may subtract value or vice versa. Alternatively, both longs and shorts may add or subtract value at the same time. We have witnessed periods in our own live process for market neutral that resemble each of these scenarios.

We did not mention these potential pitfalls in an attempt to dissuade investors from directional long-short investing. Rather, we believe such strategies are a useful and more efficient approach to portfolio management over the long term. Yet we caution that shorter term results could be unattractive compared to their long-only equivalents if a manager’s process is temporarily out of style.

When market neutral strategies were being introduced to the institutional marketplace, many investors believed – or were led to believe – that these strategies were designed to work in both up or down markets. While theoretically true, market neutral strategies could not work if the manager’s underlying stock selection process were out of favour. Therefore, when the inevitable underperformance occurred, investors were prone to dismiss the strategies in the short run as “broken” instead of staying with them for their potential long-term benefits.

The recent turmoil in financial markets delivered a good example of what can go wrong in directional long-short strategies as most quantitative fund managers have experienced abnormal performance in their returns. Many quantitative models behaved just the opposite way they were expected to. Factor returns turned negative over the past few weeks and correlations between the returns of quant managers increased.

One of the possible catalysts of this development was the losses suffered by many multi-strategy managers. As it was almost impossible for them to trade their credit portfolios in virtually illiquid markets, these managers sought to raise cash in other, more liquid markets. One of the most liquid markets is the equity market. As a consequence, short names began to rally as investors were forced to deleverage their short positions, while the more attractive stocks were selling off.

Based on past experience with such developments this will reverse sharply. A similar situation occurred in 1998 when Long Term Capital Management was in trouble. At that time, quant strategies also underperformed due to irrational market behaviour. Such situations create opportunities as stocks with exceptional fundamentals trade at substantial discounts. Even though it is impossible to predict how long it will take until the market refocuses on the fundamentals, stability has proven to win in the long run. Investors should be aware of the fact that, even though directional long-short strategies offer an attractive risk-return pattern, there is still a risk of underperformance under certain market conditions. The predictive ability of the underlying stock selection model is of vital importance for the success of both the long-only and the directional long-short strategy.

Directional long-short strategies represent a new approach and opportunity for traditional long-only investors. In today’s market environment, the ability to add value on both long and short exposures in a single portfolio is an attractive option that investors should consider. Simulations indicate that a typical 130/30 strategy may result in an outperformance of 50 to 150 basis points per annum compared to long-only portfolios without increasing the risk. Even if these are rough estimates and the industry has limited experience in this area, relaxing the long-only constraint should benefit investors over the long run by boosting returns, lowering risk and improving risk-adjusted performance.

The expressed views and opinions are those of the authors and are subject to change without notice.