Risk budgeting

BY END-2008 TRUSTEES of pension funds and investors in other long-term investment strategies could be forgiven for concluding nothing they learned about how to optimally manage those assets held true anymore.

• Diversification, or the law of not putting all your eggs in one basket, was one of the first principles to fail. Worse, in the global arena the more diversified your holdings the more devastating your results were. Investors who opted for alternative asset classes in the quest for uncorrelated sources of alpha were caught up in painful liquidity squeezes. And the traditionally low-risk and ostensibly risk-free asset classes reflected in bonds or money market funds were where some of the worst damage occurred.
• The concept of risk premiums for asset classes took on a whole new meaning as investors seriously questioned whether they could still believe in the principle of equities for the long term as the optimal inflation-beating asset class.
• The principle of setting strategic asset allocations (SAA) and designing solutions to target those mixes over the long term (on the academic argument that active market timing typically had a lower probability of success than a strategy that stuck to narrow bands around the SAA) appeared, in September, to represent the height of folly. (That said, by December performance outcomes suggested that indeed asset managers who looked smart getting out of the market before the fall were often the same managers who failed to get back in when the market rose ‒ giving the SAA a bit of a new lease on life.
• Active management looked like an exercise in futility, as managers globally appeared unable to beat their benchmarks in the market crisis ‒ a time when active managers traditionally outperform. As such, the viability of active strategies had been thrown into question.

The behavioural finance specialists tell us that when things become that uncertain and the old heuristics fail to hold, what investors (and even their advisers) typically do is abdicate ‒ they abandon their long-term investment strategy, determine they simply don’t know anymore what is or isn’t right for them and hand their money over to that full service balanced manager who weathered the crisis the best.

Put another way: they quickly forget the small print under every performance survey that said past performance is no assurance of future success. Why is that cautionary so critical? Every historical analysis of performance is contaminated by the “start/end bias” phenomenon. Shift the start or end point of the data analysis by one year and your insights can be completely contrary to what another data period would suggest was true.

While the value destruction typically unleashed by that abdication is the subject of numerous studies on pension fund investing, the greatest loss to trustees and their fund’s members is any clear insight as to how a specific investment policy or strategy may link up to the specific requirements of their members. In a sense, by chasing after the successful solutions for last year’s problems they jettison the only clear road map that they have to address next year’s (or the next 10 years’) problems. As such, they leave the fund with no meaningful basis against which to judge any future outcomes.

Markets do return to normal ‒ even after a crisis as big as this one. So what can we do to get trustees comfortable with the imperative of taking back the reins of control on their funds and re-establishing a meaningful game plan?

Turning the decision framework on its head establishes the right behaviour

An honest assessment of current decision-making processes show that, more often than not, trustees are sold a solution or a specific fund manager, typically on the strength of that portfolio’s past returns. Or, as we said earlier, they’re drawn to the victor of the last war on the mistaken assumption that the next war will be the same.

The only way to break that suboptimal behaviour is to flip the decision-making process on its head.

A far more effective approach to decision-making in long-term investing is one with a starting premise that asks how much you’re prepared to lose to maximise the likelihood of achieving a realistic return within a defined timeframe. In the real world of investing here are the hard facts: returns in excess of cash can only be achieved by accepting risk. Risk means knowing not just that equities are more risky than bonds but that certain active strategies are more risky than others in terms of their probability for success and the value destruction they can generate if they fail. More importantly, every investment strategy has an embedded cost ‒ even so-called passive strategies. So the real challenge on the trustees’ plate is to develop a framework for weighing four different variables on an on-going basis:

1. The probability for success of a given strategy.
2. The quantum of potential loss should it fail.
3. The cost of employing the strategy, both at the explicit charge level and implicit in the time and resources required for monitoring the strategy.
4. The quantum of potential monetary value added should the strategy succeed.

Introducing risk budgeting

What we’ve just described is a conceptual framework known as “Risk budgeting”. Risk budgeting has been around for more than a decade and yet ‒ for a number of reasons ‒ it’s simply failed to become an integral factor in the consulting process or in the investment policy-setting process. To some extent that may be a function of the fact many people simply aren’t sure how to go about it. Additionally, the complex interplay of factors has meant the process requires its own model or decisionmaking tool to get to quantifiable answers.

Whatever the reasons in the past, the reality is that such barriers are increasingly being overcome: the tools are being created and the “guidebooks” for consultants and trustees on how to tackle the problem have been written.

So how does it actually work in practice? What follows is a brief summary of exactly what a board of trustees needs to think about to formulate a risk budget for their fund.

From a pension and provident fund perspective, risks can be categorised as follows¹:

• Strategic/policy risk: Trustees rely on an asset and liability modelling exercise that determines the appropriate asset mix for the pension or provident fund serving as the strategic benchmark. The strategic risk of the fund is the risk of poor performance of the strategic benchmark relative to the value of the liabilities ‒ that is, the potential loss that arises when liabilities can’t be met due to underperformance.
• Structural risk: This arises from any mismatch between the aggregate of the portfolio benchmarks and the total fund benchmark. For example, if the actuary modelling the optimal strategic asset allocation for a fund used the ALSI benchmark, and the benchmark for the equity solution in the fund is the median manager in the Alexander Forbes Large Manager Watch survey, there could easily be significant performance variations between the two measures.
• Active risk: The risk the active managers take to achieve their target levels of benchmark outperformance. The underlying style or strategy utilised by the active managers could introduce risk if it leads to underperformance relative to the benchmark. It’s therefore of utmost importance to choose those types of active manager strategies that will, in aggregate, lead to a portfolio robust enough to meet the benchmark over relevant timeframes. Ultimately, the aim is to maximise return per unit of risk from each strategy (eg, value, small caps, growth).

The overall risk for a pension or provident fund is thus the “sum” of the active, structural and strategic risks described and the risk budgeting process entails the efficient allocation of that total fund risk between different sources of value-add.

But how do you know where to take risk and how much should you take? For a pension or provident fund those questions are shaped by an underlying understanding of what drives investment performance, with specific guidance from advisers or actuaries utilising risk measurement techniques. Watson Wyatt² proposes a simple approach to help trustees by breaking down the total risk budget across different asset classes or fund managers in the same way a business allocates costs between its different divisions while adhering to some basic principles:

Principle 1: Investors should diversify their sources of risk as much as possible. Yes ‒ the rule of diversification still holds, despite all 2008’s evidence to the contrary. Spending risk in one place is unwise and costly. Since markets are volatile and correlations between investments change, it makes sense to spend risk in different places to ensure some decisions will be rewarded at all times.

Principle 2: Investors should only take risk where it’s likely to be rewarded. Although an obvious observation, this requires a clear framework of risk and return measurement.

Principle 3: If an investor must take a given amount of risk then he should take it where the expected return is highest. Investors are always looking to maximise expected return for a given amount of risk, as measured by a tracking error or value at risk number.

The ultimate decision on how much risk to take should be based on a qualitative and quantitative decision: that is, considering qualitatively the security of members’ benefits as well as the fund’s risk tolerance, which may then be quantified by the fund’s consultant or actuary.

The chart below provides one template for a decision-making discussion between trustees and their consultants. While the template is far from exhaustive it does provide a feel for how decisionmakers can systematically work their way through the different levels of decision-making, determine their probability of success and potential value-add to the fund and determine either the extent to which they’re currently taking on those risks in their fund or, more importantly, whether they want to take on those risks to get better performance in the future.

By examining each level of potential value creation for the fund ‒ and assessing the quantum of potential value to be gained at each level ‒ the fiduciaries now have a much clearer (and more realistic) set of expectations for performance outcomes.

What’s missing from the equation at this point is how the various trade-offs can be quantified between the contribution of a strategy to potential return and the quantum of potential loss that might be realised. That’s where a solutions provider or a consultant with an appropriate risk budgeting model plays a key role.

Why bother?

What risk budgeting offers to trustees and their consultants is the opportunity to evaluate every dimension contributing to the final performance of a fund (while looking to the future, not simply retreating from the past!). Trustees and their consultants may then take decisions from a combined cost/benefit/probability/risk of loss perspective and in so doing determine whether the fund should be exposed to a given strategy and why.

And, finally, they can formulate clear guidelines for their fund managers or solutions provider to follow, ensuring there’s as close a match as possible between the trustees’ expectations and what the manager feels he can deliver.

As Daniel Polakow has said in his own research on risk budgeting: “Understand risk-budgeting and you’ll comprehend both the merits and shortcomings of active management but also appreciate what’s actually possible within active management.”

Those insights alone are worth their weight in gold.

References

1 The Practicalities of Budgeting, Managing and Monitoring Investment Risk for Pension Funds. 2001. Brooks, Bowie, Cumberworth, Haig & Nelson.
2 Risk Budgeting and the Art of Good Risk Taking. http://www. watsonwyatt.com.