[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

Pensions: Value at risk

In an average month in an average year, you will find at least half a dozen pension-related consultations published by a number of regulatory or governing bodies. Of those that were published in 2010, three in particular caught my eye as having the potential to herald changes to the defined benefit (DB) pension landscape:

1) Published in July 2010, the EU Green Paper on the future of pensions focuses on adequate, sustainable and safe European pension systems, suggesting that a Solvency II regime could be applied to pensions.
2) The changes outlined in the Exposure Draft, issued in April 2010 by the International Accounting Standards Board (IASB), might mean that many schemes will experience a significantly negative impact on their company’s profit and loss (P&L) due to increased pensions expense — especially for schemes currently with a high proportion of risky assets.

The Exposure Draft also suggests the inclusion of additional disclosures, which would provide users of financial statements with further information about the pension risks faced by sponsors. Such information would include sensitivity analysis, details of asset- and liability-matching strategies and factors that will cause the pension cost to increase significantly.

3) The Pension Protection Fund (PPF) levy consultation published in October 2010 sets out proposals for a new framework to include an investment risk element in future calculations of the PPF levy, including the requirement for pension schemes with assets above £1.5bn to carry out asset and liability stresses.

Taken as a whole, the impact of these consultations, if finalised, will mean that an understanding of pension risks becomes a necessity.

Poker chips

The reactions to date
The reactions to these consultations have been somewhat mixed. The adoption of a regime similar to Solvency II by the DB pensions sector is seen by some as a means of strengthening security, while others have expressed concerns that such a move will undermine the level of benefit provisions.

Other industry commentators believe that the PPF levy framework will better reflect investment risks but will penalise schemes investing in risky assets and encourage investment in matching or lower risk assets.

The dilemma of trying to decide on the fine balance between security versus adequacy has always existed. Likewise, the judgment between the amount of risky assets (and potential returns) versus less risky assets (and lower returns) has always generated a great deal of debate. And if Solvency II were to apply to pension schemes, even in a modified form, it is likely to lead to:

>> Higher funding requirements, or at least the pressure of higher funding
>> A focus on asset liability risks — potentially resulting in a shift from holding risky assets to liability-matching assets
>> Additional cost of compliance.

These issues are being debated, in part, because higher expected returns are allowed for in schemes investing in risky assets. This has an impact on both pension accounting and ongoing funding. In an accounting context, holding risky assets means a lower P&L for sponsors. For schemes that are large relative to the size of the sponsor, allowing for advanced credit on risky assets could lead to an accounting ‘profit’ that is unreal.

On funding, holding risky assets could reduce the required cash contributions, as advanced credit could be taken in constructing the recovery plan.

While it is important to understand how policy can influence behaviour, risk management has sometimes been seen by organisations as either increasing actual funding cost, or as a box-ticking exercise to ensure regulatory compliance.

So is there a positive way to look at this?
A simple example, such as that of a house purchase, suggests an upside to some of the changes proposed. Three years ago, I was looking for a somewhere to live in London and found two perfect properties. Both were situated in the same location, were of a similar size and each was on the market for £500,000. The only difference was that one was a period property with elegant period features and the other was a new-build apartment. For what seemed like an identical cost to me, I settled for the beautiful period property while my friend bought the flat.

Over the next three years, I incurred £50,000 worth of maintenance and repair bills relating to damp, water pipes leakages – all the usual problems one would expect from an old property. My friend, on the other hand, spent £10,000 over the same period. Although at the outset I was aware that the maintenance cost for a period property might be higher, I had no idea about the range or possible costs.

If, in the Home Information Pack, I could have found information such as ‘the Value-at-Risk (VaR)’ — at 95th percentile, say — ‘for maintenance cost is £50,000 for the period property and £10,000 for the flat’, I may well have chosen to buy a different property. However, as my focus three years ago was on the headline cost of the property, I made the decision based only on information I had at the time. Without a quantitative way of assessing the risks, my qualitative information on its own let me down and led me to a decision that I came to regret.

If I had understood the risks more clearly, the decision may well have been the same but the decision-making process would have been much more controlled.

What are the hidden costs of DB schemes?
While some pension schemes carry out a degree of stress testing or setting a risk budget using measures such as VaR on the assets and liabilities, pension costs continue to be communicated using a single figure on a present value method.

The results are reliant on which discount rate has been chosen and give the decision-maker no information around real risks, or of the likelihood that the pension cost quoted might turn out to be different. Over the years, it has given trustees and sponsors the impression that scheme actuaries ‘magically’ change the liabilities by setting subjective discount rates.

While we cannot settle on which discount rate is the right one, perhaps communicating liabilities in a different manner might help the debate. So how about quoting a range of pensions costs? Or a confidence level? So for a scheme with £100m liabilities on an economic basis and £40m liabilities on a more optimistic basis, the liability could be quoted as £60m at the 70% confidence level. That naturally leads the trustees and sponsors to ask: What is the probability that we might have to put more money in and how much more?

We can argue that a confidence level is only theoretical and probabilities may not be accurate but, from a communications point of view, the sponsor can clearly understand that there is no such thing as a ‘single’ cost and actuaries could avoid being accused of ‘playing God’ by setting subjective discount rates which change the value of liabilities.

How will the pension landscape change under the new proposals?
With the pressure to improve risk management for all organisations, simply going through a box-ticking exercise to satisfy compliance will not be what decision-makers want to do, especially in a world where we all have too much to do with too little time. High-quality condensed risk management information will enhance the qualitative information, enabling decision-makers to select the appropriate reactions.

Risk management is not about funding rules and regulations; it is about better decision-making. I therefore hope that the current raft of consultations is widely viewed as an opportunity for companies to take stock and to give due prominence to the need to have good risk management information as an integral part of their decision-making processes.


Celene Lee is head of pensions at Barrie & Hibbert