Pension funds: understanding funding ratio risk better (part 2)

Post with image

Understanding the factors that support or weaken a pension fund’s funding ratio is important since it forms the basis of crucial policy decisions including the allocation to risk sources in the investment portfolio. In part 1, we argued that funding ratio risk was an important metric and could produce valuable insights. In part 2, we argue that a more detailed grasp of the sources of economic risk could help the portfolio manager adjust the allocation as and when required.

Looking at the economic sources of funding ratio risk

To gain a better understanding of the risks associated with a pension fund’s funding ratio, one could look at sources of risk such as changes in interest rates or inflation as an alternative for asset class-related risks. Anton Wouters (Head of Customised & Fiduciary Solutions), Raul Leote de Carvalho (deputy-head of Financial Engineering) and myself have developed a model for doing so[1].

In a study in 2012[2], Dutch trade union FNV Bondgenoten made a strong case for a risk management approach that is not based on asset classes, as applied by most pension funds, but on economic risk sources. A fascinating idea. With the aid of our in-house scenario generator and investment policy simulator, we have developed a novel method for gaining an understanding of the risk sources that affect a pension fund’s funding ratio. This can be used next to the customary metrics to determine funding ratio risks such as the magnitude of the tracking error of the return difference between assets and liabilities and the risk of a funding shortfall.

Economic risk sources: making it work

The basic idea is that particular risk sources affect the behaviour of individual asset classes and thus of any portfolio allocating to asset classes. Pension fund managers often intuitively understand the behaviour of economic risk sources such as inflation better than the behaviour of asset classes which often involve a mix of such risk sources. So a risk system that is based directly on risk sources can give pension fund managers a better grip on the investment policy and puts them more in control.

The exact effects of a risk approach based on economic risk sources depend on two key factors.

A) The selected control variable for the investment policy. In the examples used in the FNV Bondgenoten study, the central objectives are to provide transparency and to control the surplus risk. The focus, therefore, is on the potential change of the surplus. This is the simplest control variable because surplus is defined as a linear relationship – investments less liabilities – and it is generally easier to use linear relationships in statistical calculations rather than non-linear relationships.

In practice, however, the potential change in the funding ratio is much more crucial for pension funds than the potential change in the surplus. Many vital policy decisions (e.g. indexation and risk appetite) are based on the course of the funding ratio. For this reason, we have developed a model that breaks the funding ratio risk down into separate risk sources. This makes our model more complex in terms of statistical rigidity because the funding ratio does not consist of the simple linear difference between investments and liabilities, but rather of their quotient.

B) The choice of risk sources that are deemed to be important for control purposes. The FNV Bondgenoten publication identifies these risks:

1) interest-rate risk

2) inflation risk

3) country risk

4) credit risk

5) business risk

6) currency risk

Notably, the sources are organised along a stepped scale. Only the extra risk that a higher risk source (e.g. business risk) adds relative to its immediate predecessor (credit risk) is taken into account.

In our approach, we too have in principle applied such a stepped approach.

Exhibit 1: Selected risk sources

tableSource: BNP Paribas Investment Partners

We start with the real interest rate (risk source 3) and move up via the break-even inflation (these two sources in combination form the nominal interest rate), the swap spread and the credit spread, to step 7: the business credit curve.

With risk source 8, we add the extra risk dimension of high-yield bonds, followed by the remaining additional risk of equities with risk source 9. In line with the FNV Bondgenoten study, we also defined the currency risk (risk source 2).

In addition, we further expanded the inflation dimension by including, alongside forward-looking break-even inflation, risk source 5, which looks back at the risk impact of actual inflation over the past 12 months.

Finally, we incorporated a risk source that explicitly measures the scale of the funding ratio risk resulting from the fact that the liabilities-related interest-rate risk is, as a rule, not fully hedged via the allocation to the matching portfolio. This is risk source 1 (‘interest-rate hedge mismatch risk’).

Allocation starts with hedging decisions

The resulting insight into the impact of the scale of the interest rate hedge is a fundamental difference between our approach and that of FNV Bondgenoten. As interest-rate and currency hedges are generally among the first decisions made in the strategic asset allocation approach, we have put these risks at the top of the list. We have thus defined a set of risk sources that provides an alternative way of understanding total risk.

X-raying a funding ratio by using economic risk sources

A breakdown of the risks around the expected funding ratio according to risk sources can provide insights which can be the basis for portfolio adjustments. Let’s suppose that the expected funding ratio in one year’s time carries a risk of 8% in terms of its standard deviation. An example of the breakdown of this 8% can be found in exhibit 2. It turns out that, given the pension fund’s current asset class allocation and given its current interest-rate hedge of the liabilities, 94% of the standard deviation can be explained with the aid of the defined risk sources. The other 6% is specific risk.

Exhibit 2: Risk-allocation funding ratio

graph

Source: BNP Paribas Investment Partners

What is striking in this example is that the risk is tilted heavily towards high business risk. This tilt can be intended, but if it is unintended, pension fund management will need to adjust the strategy. That can be done in a smart way. Given that we calculate an allocation to risk sources for each individual asset class and the liabilities as well, our method also makes it possible to manage the portfolio on the basis of specific economic risks.

Exposure to a risk source can be adjusted in several ways, depending on the specific source selected. In other words, the method supports the asset allocation process as well as the management of the interest-rate hedge of the liabilities by providing a more detailed understanding of the risks of the policy pursued or proposed. Thus the funding ratio’s exposure to economic risks can be adjusted as and when required.

Click here to read ‘Pension funds: understanding funding ratio better (part 1)’

For more on our tailored solutions that meet the needs of pension schemes, insurance companies, corporates and other institutional investors, go to http://institutional.bnpparibas-ip.com > our-capabilities > multi-asset-solutions > customised-solutions

[1] For a description, see the paper; ‘Decomposing Funding Ratio Risk: Providing Pension Funds with Key Insights into Their Liabilities Hedge Mismatch and Other Factor Exposures’ by Erik Kroon, Anton Wouters and Raul Leote de Carvalho (http://ssrn.com/abstract=2596543 ).
[2] Beleggingsrisico’s: Aanpak voor het beheersen van de risico’s in het beleggingsbeleid van pensioenfondsen”, by FNV Bondgenoten in co-production with APG, Kasbank and Ortec-Finance. Available at http://www.fnv.nl/site/brochures_en_folders/nieuwe_methode_risicomanagement230512 (text in Dutch)

[starbox id=erik.kroon][starbox id=anton.wouters]

Erik Kroon

Research Manager Financial Engineering

Leave a reply

Your email adress will not be published. Required fields are marked*