Real risk is not about Greek symbols; it is about losing money. It is not about tracking error or weighting reports – it is something that is offensive and hurts
Writing about any subject that has been in fashion for the past 18 months is always fraught with danger. Indeed, too much has probably already been written on long-short investing.
Advisers have been left with an array of mixed messages from different interest groups. However, far too much of the debate has been taken up with talk of structure, domicile and legislation and far too little about the important things that affect clients.
As a student of risk measures I learnt to cover a page in Greek symbols, analyse portfolios to three decimal places of accuracy and optimistically feel I understood “risk”. Well before 2008 it was very clear much of this output was pure garbage – Long Term Capital Management in
1998 and the tech bubble in 1999-2000 were perfect case studies demonstrating the assumptions that sit at the heart of the vast majority of risk models are just plain wrong.
For those that had not noticed, 2008 put this on the front page of every newspaper. Markets are not efficient, future volatility is not best predicted by measures of past volatility, and correlations have a horrible way of humiliating us all. The risk of extreme outcomes is far higher than assumed by risk models.
Macroeconomic focus
Diversification is a fabulous thing, but in times of stress, correlations rise sharply leaving no place to hide. As the world is currently obsessed by a further economic slowdown, it is perhaps not surprising macroeconomic indicators have become a focal point.
The dispersion of return between stocks has fallen sharply, leaving slim pickings for the stockpicker. Worse still, those that believe they can manage an absolute return fund from a portfolio of long equities have some explaining to do – on many days in the past quarter, more than 95% of stocks fell. On some days, the number of stocks falling actually hit 99%, hardly a chance for those looking to find the stocks that will rise. Correlation is among the single most important factor in building portfolios (either individual stock portfolios or portfolios of different funds or asset classes). Conversely, there is so little good analysis on it and very little convincing data.
Debate
I recall chairing the Asset Allocation Committee for a major global asset manager. The debate on the outlook for different asset classes would rage for hours with sophisticated, nuanced discussion; the answers may well have been wrong but there was no shortage of well-argued debate. Then came the second part of the discussion: ‘What are the likely correlations between these asset classes?’ This same group of thinking individuals fell strangely silent at this time and simply turned to recent history.
Perhaps unsurprisingly, correlation data simply does not raise the same passion in people, but it is the thing which, during every major market setback, has been a most critical variable. To put this in context, how many people are aware US equities (S&P 500) are currently 68% more highly correlated than the five-year median.
Sadly, there are no silver bullets here. The only defence is a deep cynicism of traditional risk models and to always ask the question: ‘What would happen if all these assets suddenly became very highly correlated, if their prices moved as one?’
Skill and luck
Real risk is not about Greek symbols; it is about losing money. It is not about tracking error or weighting reports – it is something that is offensive and hurts.
Skill and luck are often confused. Lucky managers do well for one or two years. Skilful managers have done well for many years. At certain times, the markets may move against skilful managers’ approach but they seldom lose their mojo for long. It continues to stagger me how little the market differentiates between ‘skilful’ and ‘lucky’ and how many long-term, average or below-average managers survive and get paid exceptionally well. The industry needs ballast, but why do investors still pay so much for it? In most other industries, the difference in price between a low- and high-quality product is huge; in our industry it is often a few basis points.
Innate understanding
To me a skilful manager is a very different animal from the average. They are characterised by an obsessive slavishness to their art, have an almost scary belief in their philosophical underpinnings and a work ethic that should embrace us all; they feel fully accountable for the outcome of their decisions and, critically, have an innate understanding of the nature of risk.
It is no coincidence many of the skilful equity managers have been drawn away from long-only investing. Their innate feel for the risks embedded in stocks allows them to run ‘flexible equity’ long-short portfolios. In identifying skill, I strongly believe there is no substitute for track record. It is not easy to become a good long-short manager and many fail. Long-term track records are the only true test. There is also no substitute for ensuring they operate in a structure that makes portfolio managers entirely accountable – discretionary bonuses and vague responsibilities are a nonsense. To be successful takes 100% of your time, with no distraction.
Flexibility
The absolute return long/short sector can boast many of the best long-term equity managers – firms like Sloane Robinson or Egerton are not just lucky. Skill is critical, but it is the flexibility afforded to absolute return managers that is also fundamental to ensuring their returns are not dragged inextricably back to the index. In other words, it is about ‘alpha’ and ‘exposure management’. Having the flexibility not to own a long-only portfolio when 99% of stocks are down is critical.
Compound arithmetic is the eigth wonder of the world – it is remarkable. The impact on long-term returns is extraordinary. A manager who suffers a 50% decline in NAV, as many long-only managers did, needs 100% return just to get back to break even. The long-short manager who was down 10% needs just 11%. If these returns are not achieved, the differential continues to widen inextricably over time. Absolute return managers have a place in portfolios in part because of their power of compounding – avoiding major collapses is critical.
Timing
But the other critical component is timing. Absolute return funds are generally far less volatile than long-only funds. Volatility is the thing that undermines returns most – a fund that returns a steady 0.5% per month is arguably far more likely to give an adviser a positive return than a volatile fund giving 10% a year. More often than not, the timing of the purchase or sale of the assets will coincide with a period of extremes and reduce the return. While I am sure each of us believes we can be contra-cyclical, the evidence is sadly the opposite – look at unit trust sales data.
The debate about long/short investing will continue to persist, particularly at times of low equity return dispersion as currently. However, the power of compounding is enduring – I recently did an analysis of RWC’s bench of managers over their lifetime of running long-short funds. Each had returned between 1.5 and three times the market return with approximately half the volatility. The low volatility ensures the returns are more available to investors.
Peter Harrison, chairman and CEO, RWC Partners
