Sections

Spending the risk budget

Risk budgeting is a multi-faceted problem, and there are many kinds of interpretations. These arise from the different ways we can define risk and the ‘budget’. When we are analysing the risks within a pension fund, it is best to view the pension fund as a part of the sponsor’s enterprise. This is because in a typical defined benefit pension plan, the performance of the pension fund assets has direct consequences on the revenue accounts and balance sheet items of the sponsor.
The main fiduciary responsibility of the trustees is to ensure that the benefits of the current beneficiaries will be paid when they fall due.
The risks in pension plan financing should therefore be analysed from the perspective of the various cash flows, both from the liabilities and assets.
The key features of pension plan liability cash flows are:
o they are of very long-term nature, often spanning 60 years and beyond;
o they are uncertain as to their size because of their links to salary and price inflation; and
o they are uncertain as to their timing because of demographic factors and their long term trends, eg, mortality, leaving service, early retirements etc.
The pension plan asset cash flows are the same as those of any other financial institution and should require no special treatment.
The problem with cash flow analysis is that it is generally difficult to compare cash flow streams. Cash flows are multi-period entities, and in order to compare them, we need to have a rule that places a present day value (or price) on any cash flow in the future. This process is called discounting or pricing, and it reflects the intuitive notion that £1,000 available to us immediately is more valuable than £1,000 available to us in, say, 20 years’ time. So every cash flow at future date has a value that is a function of the expected amount and the discounting process.
By adding together the present day values for all cash flows we get two simple numbers, PVA (value for the assets) and PVL (value for the liabilities) to compare. The risk can be defined as the uncertainty of the PVA – PVL (surplus) or PVA/PVL (the asset/liability ratio, ALR).
When there is no liquid market with openly quoted prices (eg, for general pension fund liabilities), there is no ‘natural’ pricing rule and it is left to the actuaries, regulators, accountants etc, to devise their own rules to calculate PVL. When there is a liquid market, such as the stock market, we have a ‘natural’ pricing rule, namely that PVA = ‘market price’.
Factors that may affect the assets and liabilities cash flows directly are:
o salary inflation
o price inflation
o demographics
o mortality
o withdrawals
o early and ill-health retirements
o actual pension fund asset returns (value-for-money guarantees)
o redemption yields on bonds (transfer value calculation and annuities)
o UK FT All Share equity yield (transfer value calculations)
Factors that may affect the pricing
(valuation) rules:
o expected returns from the pension fund assets (typically used for ongoing, funding valuations)
o redemption yields on AA bonds (FRS 17 valuation basis)
o redemption yields on Gilts and UK FT All Share equity yield (MFR valuation basis).
Other risk factors are:
o sponsor’s credit rating (default risk)
o regulatory changes
o taxation
o amortisation periods
Most risk budgeting models focus on the financial risk alone and what may appear risk-free in purely investment terms is in fact risky if taken in a wider context. In particular, the long term mortality improvement represents a real risk for long term pension fund funding status.
An important choice we have to make for risk analysis is the valuation basis for the cash flows; we normally use the FRS17 valuation basis. The financial metrics that we use is the FRS17 asset/liability ratio (ALR = PVA/PVL).
We will take as risk the variability of the change in the ALR over a year (A/L return), as measured by its standard deviation.
How do we set the overall risk budget?
The idealised portfolio consisting of assets whose returns would precisely mimic the liability return is sometimes called the Liability Benchmark.
The variability characteristics of the asset return on the portfolio will be largely determined by the strategic benchmark, the type of investment structure and investment manager mandates. The greater the mismatch (measured by the covariance) between the actual portfolio and the liability benchmark the greater will be the risk of adverse outcomes, represented by low asset/liability ratios and possible deficits.
In order that these adverse outcomes do not prejudice the ultimate solvency of the pension plan, we need to have an adequate amount of risk capital. This capital can be in the form of additional assets in the pension fund, provision in the sponsor’s balance sheet or the (implicit) agreement between the trustees and the sponsor to fund deficits when required.
The amount of risk the trustees and sponsor should be willing to take on board would be related to the amount of risk capital that they would be prepare to tie up, implicitly or explicitly, to cover these contingencies.
There are two important points to consider in setting the overall risk budget:
o should the trustees and sponsor seek any additional returns (ie, employ risk capital) in the running of a pension fund, which would typically not be a ‘core’ business activity of the sponsor? and
o what is the expected additional return that can be produced from the risk capital employed?
The quantification of the trade-off between the additional returns and the adverse outcomes can be assessed by a process commonly known as asset /liability modelling (ALM) study. Whilst ALM can provide the analysis, it cannot tell us how much risk can/should be taken. There are four key fund-specific factors that would influence this decision, as illustrated in table 1.
Each of these factors should be discussed with the trustees and sponsor, bearing in mind the various stakeholders in the fund and sponsor enterprise.
Once the size of the risk budget is determined (a governing board decision) the question of where it should be spent can be tackled (an executive decision). Two major areas where the risk can be taken are:
Policy: Strategic asset allocation (SAA) and benchmark. A matrix of major asset classes, percentage allocations and corresponding benchmark indices
Implementation: Investment manager structure and manager selection. A matrix of investment managers or manager types, percentage allocation to each and performance benchmarks/targets/controls for each investment mandate.
How do we allocate the total risk budget between strategic (benchmark) decisions and investment structure? The total risk budget can be viewed as comprising of two parts:
o the contribution due to the differences in SAA benchmark return - liability return the Policy risk budget;
o the contribution due to the differences in total fund return – SAA benchmark return the Implementation risk budget.
For a typical pension fund, the policy risk budget is much greater than the implementation risk budget, but some pension funds have opted for very low policy risk budget. It is largely driven by the decision as to how much should be invested in equities and alternative assets (real estate, private equity, hedge funds etc.) if we are in the FRS17 risk measurement framework, where corporate bonds are viewed as the risk-free or low risk asset.
The implementation of risk budget is largely driven by the overall investment structure and performance targets given to its asset managers rather than the actual investment manager selection. However, successful manager selection will be a key contributor to actually achieving the targeted outperformance within the risk budget framework. The schematic division can be shown graphically as shown in table 2.
The setting of the risk budget is of key importance to a pension fund and the sponsor. The Myners’ Report highlighted the strategic asset allocation as the most important part of the overall risk budget. Investment structure risk/reward needs careful monitoring and success will rely on the right choice of the investment managers and mandates.
Peter Ludvik is head of European asset allocation at Watson Wyatt LLP in Reigate

Have your say

You must sign in to make a comment

IPE QUEST

Your first step in manager selection...

IPE Quest is a manager search facility that connects institutional investors and asset managers.

  • IN-2412

    Closing date: 2018-02-28.

Begin Your Search Here