Allocation based on risk factors rather than traditional asset categories is gaining traction, says Miranda Schoutsen, but some investors are wondering whether the old ways are not better
The financial crisis revealed that diversification across asset classes does not necessarily mean diversification of risk, as institutional investors the world over found to their detriment. Dutch investors were no exception, prompting a drive by the trade union FNV Bondgenoten to help pension funds get a better grip on risk exposure in asset allocation, according to Jose Suarez Menendez, a certified financial analyst who serves as pensions adviser to FNV Bondgenoten. Suarez Menendez sits on the investment committee at the €45.8bn industry-wide metalworkers fund PMT, as well as on the trustee board of the €12bn industry-wide retail sector scheme.
Dutch trustee board seats have been divided more or less equally between employer and employee representatives, and unions such as FNV Bondgenoten supply many of the employee representatives to various trustee boards.
Two years ago, after attending a meeting at which APG’s Pieter van Foreest presented an integrated bottom-up risk management approach, Suarez Menendez began to wonder if these same principles might not be applied to asset allocation as well. He teamed up with Foreest and representatives of Dutch custodian Kas Bank en consultant Ortec to explore this question, resulting in a recent report published by FNV Bondgenoten entitled ‘Investment risk – an approach aimed at controlling risks in pension fund investment policies’.
“With the current approach, pension funds cannot get a firm enough grip either on underlying risks or on the presumed diversification benefits,” Suarez Menendez says. “And moreover the extent to which various risk sources impact on asset prices is not constant.”
To control exposure to underlying risks, the allocation process should start from risk sources not assets, the report notes. “With a fixed asset allocation, you’re more or less at the mercy of the financial markets,” says Suarez Menendez. Because even though pension fund trustees are aware of underlying risks in the portfolio, the traditional approach simply does not provide sufficient information to adequately manage the portfolio based on those risks. Since investments can be regarded as the structured products of risk sources, the sources of risk themselves remain hidden from view, the report explains.
The authors present an alternative method starting from a risk budget which is then allocated to selected risk sources, including interest rate risk, inflation risk, country risk, credit risk, equity risk and currency risk. Instead of a fixed asset mix where the underlying risk sources will vary, this approach is based on a fixed mix of risk factors, where the asset mix varies over time.
It is a matter of perspective, says Pieter van Foreest: “These are two sides of the same coin. The question is, what do you look at first: risk factors, or asset categories?” Above all, the risk-based approach offers more insight, he adds. “I can show trustees a host of different investment instruments, but in the end what they’re really buying is specific sources of risk. By looking at the underlying risk factors they gain a better understanding, which enables them to make better, more informed choices.”
Interest and doubts
The report published by FNV Bondgenoten and its partners has attracted a lot of interest in the Dutch pension industry, and a seminar presenting the method last October was sold out. But not everyone is convinced.
Bruno de Haas, head of policy and research at Media Pensioen Dienst (MPD), asset manager of the €4bn media sector pension fund PNO Media, says: “We try to always examine new perspectives to see if they have something to offer. Lately a lot of new approaches have emerged proposing different ways to determine the risk profile of the asset mix, and the ‘risk source method’ proposed in the FNV Bondgenoten report is one of them.”
He believes it makes sense to look beyond portfolio and asset categories at the sensitivity to risk factors and ask the question: for which risks am I being rewarded? “But that said, looking at a risk distribution pie chart still doesn’t tell me whether my portfolio is optimally diversified. It’s all well and good, for instance, to talk about country risk, but which country’s default risk are we talking about – The Netherlands? Italy? Or Ghana?” he says.
“Every asset category depends on economic growth so I’m not surprised that correlations may be high. But how does that help me to organise the investment mix? What is important is that the rewards for investing are commensurate to the risks. So share price is an important consideration to decide which risks are justified.”
John van Markwijk, CIO of €31bn metal workers fund PME, says that his pension fund keeps a close eye on its exposure to country, credit, inflation and interest rate risk. “PME also takes exposure to credit risk, liquidity, leverage and complexity into account in its portfolio and within asset categories,” he says, adding that he has read the FNV Bondgenoten report with interest.
“Unfortunately many asset categories can’t be cut up along those lines, which may lead to the conclusion that the old, traditional categorisation into equities and fixed income wasn’t so bad after all,” Van Markwijk says. “And in the end even many risk-return ratios can be reduced to the basic categories of stocks, bonds, property and cash.”
He adds: “To benefit from the risk source method and related approaches you either have to know exactly what you’re doing or be the first to implement the method.”
MPD makes a point of considering all alternative methods, as De Haas explainss, and so has also explored risk parity, which aims to have various components of a portfolio contribute equally to the overall portfolio risk. De Haas: “For pension funds this is a rather complex method to implement, because of the use of leverage among other reasons, but risk parity does offer a way to diversify and control risks while earning an attractive premium.”
Despite MPD’s open-mindedness when it comes to alternative risk-based allocation perspectives, PNO Media did not change the risk profile of the asset mix after the 2008 credit crisis. “The most important decision a pension fund has to make is still the extent to which it will hedge its interest rate risk, for instance buying interest rate swaps to protect against falling interest rates. Tinkering with the weightings in asset categories is fairly inconsequential by comparison,” says De Haas.
Pension funds that did not hedge the majority of their interest rate risk have taken the worst hit. “Hedging is the way to insure against bad weather scenarios. In any case, buying interest rate swaps is the same thing as borrowing money to invest in bonds. So pension funds with swaps have in fact put into practice the risk parity approach even if they weren’t aware of doing so.”
Three pension funds – the industry-wide grocery trade fund, the corporate scheme of Kas Bank and the scheme for Cordares employees – are set to conduct a pilot study in 2013, maintaining a ‘shadow allocation’ organised along the lines as set out in the report. Suarez Menendez says: “Over the course of the year we hope to gain sufficient implementation experience and information to publish a follow-up report.” In addition, Van Foreest, Suarez Menendez and others are now exploring practical questions, such as how best to explain risk sources and various allocation choices.
The regulator (DNB) was informed of the report but has not been involved in its production. “We have conferred with DNB as we went along and they were enthusiastic about our initiative but the regulator doesn’t give a stamp of approval,” says Suarez Menendez. “It was important to us to be able to show that the industry is capable of initiating improvements without being ordered to do so by the regulator.”