Short-selling In A Long-only Portfolio
By: James Clark
Is short-selling risky, something that belongs only in hedge funds? James Clark, of Aurion Capital Management Inc., makes the case for short-selling in long-only portfolios.
To some, short selling is ʻbad.ʼ Short-selling is risky. And shortselling is only for hedge funds. However, the time has come to consider the role of short-selling in a longonly portfolio.
Over time, short-selling has been seen as risky. There are many issues for the shortseller to address including:
- financing costs
- the potential for a short squeeze
- the natural long-term upward trend of the market
- the theoretical unlimited downside risk
Because of these risks, short-selling has been the domain of specialty hedge funds. Long/short strategies, short only funds, and market neutral funds dominate this territory of investment management. Generally, people understand the concept, but they have been reluctant/slow to embrace it in regards to the management of traditional portfolios in their pension funds. It is something that ʻother people doʼ to make returns, but not ʻlong term investors.ʼ
Short-selling within funds is not a foreign concept in Canada. In the recent past, there have been respectable short transactions. In the heady days of the late ʼ90s, when BCEʼs large stake in Nortel was overshadowing the rest of the company, there was a company called Teleclone. Teleclone was essentially a fund which was created to allow investors to buy a pure BCE investment, excluding its 52 per cent stake in Nortel. Telecloneʼs sole purpose was to unlock the value of all of BCEʼs other companies excluding Nortel. To do this, Teleclone held a long position in BCE and a short in Nortel, leaving a net play in BCE. Investors were effectively ʻshortʼ Nortel within a structured investment vehicle.
So what is the role of short-selling in a long-only portfolio?
The management of a traditional equity portfolio starts with idea generation by the portfolio management team. Depending on the investment style of the manager, idea generation can include finding investments with growing sales forecasts, earnings momentum, above average profitability, a low P/E ratio, or a low P/B. When a good investment idea is generated, it is passed through a screening process which includes a number of portfolio constraints. The investment idea can be excluded from the portfolio if it does not exceed the hurdle that the constraints impose. These constraints could include consistency with the managerʼs investment style, eligible securities, cap limits, sector exposures, or a limit on the number of holdings. If the idea survives this process, it is added to the portfolio. It is taken for granted during this process that short-selling is not allowed in the portfolio; therefore, all stock selections must be viewed to be moving in a positive direction based on the portfolio managerʼs research and experience. What is also generated, however, are investment ideas that the portfolio manager views as overvalued. These ideas are not included in the portfolio, but they can still be profitable to the portfolio if they are shorted.
Idea generation can, among other things, involve a process of relative valuation. Stock A is attractive and Stock B is not. This type of conclusion is usually reflected in the portfolio by making an allocation to Stock A – investing say three per cent (assuming this is the maximum allowed by the investment constraints) and by allocating a zero per cent weighting to Stock B (assuming no allowance for short-selling in the investment constraints). Therefore, assuming both Stock A and B have market index weights of 0.5 per cent, this leads to Stock A being overweighted by 500 per cent and B is underweighted by 100 per cent. In fact, any security in the market index that is not included in the portfolio is underweighted by 100 per cent. However, favoured securities held in the portfolio can be overweighted from 300 per cent to 500 per cent, depending on the investment constraints.
Forced To Buy
To develop this further, if the portfolio manager had an expected return for Stock A of 10 per cent and an expected return for Stock B of negative 10 per cent, the most the portfolio manager would be able to gain for a long-only portfolio would be 0.3 per cent (10 per cent applied to a three per cent allocation). If, however, the portfolio manager was able to short Stock B in an equal weight, they would be able to gain 0.6 per cent for the portfolio. Under the long-only constraint, the manager may be forced to buy an alternative long position in a security which may not provide the portfolio as great a return as shorting Stock B. The managerʼs investment decisions and investment returns, and that of the investor, are being limited by having to invest in less attractive stocks. The investment decision is being controlled more by perceived risk controls than by seeking the best investment for the portfolio.
Therefore, if a manager is allowed a range of overweight positions, say 10 per cent to 500 per cent, why not allow the manager to underweight a position with a similar range? In the previous scenario, the portfolio manager would be allowed to short the investment in Stock B and thus create a greater potential positive return for the portfolio, rather than taking a less attractive long alternative.
This strategy would make a broader use of the portfolio managerʼs idea generation capabilities. As was mentioned earlier, as the manager generates ideas based on his investment process, (growth, value, momentum), they are also identifying investments which might be viewed as overvalued due to slowing sales, loss of market share, new competition, lack of technological breakthroughs, and potential lawsuits (tobacco, patent infringements, pharmaceuticals). ʻOvervaluedʼ opportunities would not make it through the managerʼs decision-making process for a long-only portfolio, but it could be included if shorting was allowed and the manager thought that it would bring a positive return to the portfolio.
Shorting A Stock
By allowing the manager the opportunity to short a stock, one must not ignore the risk management process. Just as there are risk controls with any long investment, there should be an equal emphasis on the risk controls for a short-sell within the portfolio. The risk controls for managing a short position are similar to those used to manage a long position. Short-selling risk controls can include:
- limiting the number of shorts per each sector or industry
- restricting the size of each short position
- using stop-loss limits
- shorting only highly liquid stocks
So what could short-selling in a longonly portfolio look like? If short-selling is viewed as a way to take underweight positions of more than 100 per cent, the number and size of the short positions should look similar to the number and size of over-weight positions of more than 100 per cent. If overweight positions of more than 100 per cent are less than 10 per cent of the portfolio, short positions should also be less than 10 per cent of the portfolio – symmetry of returns opportunities. In summary, a portfolio that allows the use of short-selling would look a lot like a long-only portfolio, but it would have better balance by including tactical and strategic long and short positions.
With that issue resolved, we will next consider the role of foreign equities in a Canadian-only equity portfolio!
James Clark is vice-president, business development, at Aurion Capital Management Inc..
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -