Michel Iglesias del Sol and Gerard Roelofs assess the wider implications of FTK2
On 12 July 2013, the Dutch ministry of social affairs and employment published the long-awaited consultation document on the new Financial Assessment Framework (FTK) for pension funds. On the one hand, the document does not contain much new information and could be considered an extension of the so-called headline paper from May 2012. On the other hand, there are some noteworthy changes, including a further move away from market valuation.
It was already well-known that the new FTK offers a choice between a nominal and a real (inflation adjusted) contract. The nominal contract can be considered as an adjusted version of the current regulatory framework. It does, however, aim for a higher level of security of nominal rights. This means that the current confidence level of 97.5% remains, but that capital requirements are updated, leading to an expected increase of required capital of around 5 percentage points on average. It also contains rules that limit the amount of cost-of-living adjustments that can be granted, depending on the level of solvency available. These rules emphasise the difference between the two contracts.
The real contract contains a number of new elements when compared to the current situation, many of which were already listed in previous preliminary regulatory documents, such as automatic indexation.
A striking feature of this contract is the proposed way that the yield curve for discounting of liabilities is constructed. It is based on the euro swap curve, with ultimate forward rate (UFR), like the nominal contract. Then a risk premium is applied on this curve, which is maturity dependent and converges to 1.5%. Finally, the expected inflation is subtracted, where the Dutch Central Planning Bureau (CPB) forecasts are used for the first two years and the 2% ECB target is used for 10 years and longer (with linear interpolation used for the period in between). The result is a lower discount curve than for the nominal contract.
This valuation methodology has some interesting consequences. First, there is hardly any balance-sheet sensitivity for expected (or break-even) inflation levels. There is some sensitivity to CPB forecasts but the present value of short-term liabilities is usually very limited compared to overall liabilities. The limited supply of traded inflation-linked assets is mentioned as one of the reasons for this valuation method.
Undeniably this is the case for Dutch inflation and even for euro-zone inflation and we have seen some unusually high (ex-post) inflation risk premia in the past in some markets. Dutch pension funds will therefore not be forced to pay this premium. It also means, however, that pension funds are not ‘rewarded’ in their balance sheet management for inflation hedging, even though inflation adjustments are the main purpose of this contract type. Implicitly, it also assumes that the ECB will be successful in maintaining an inflation rate near its long-term target of 2%. Clearly this still remains to be seen.
Second, the value of liabilities in the real contract will have similar sensitivity to interest rates as if they were in a nominal contract. Actually, interest rate sensitivity overall will be higher because of the (fixed) inflation component in the liabilities, leading to higher duration of overall liabilities. This leads us to expect that most pension funds will not lower their level of interest rate hedging (at least not directly), no matter what type of contract is chosen.
Impact on investment policy
Although we do not expect a significant reduction in the level of interest-rate hedging for either contract, there could be an impact on asset allocation. As mentioned above, the nominal contract implies a greater focus on security. So if capital requirements are higher, this could lead to a more defensive investment policy for funds that run a nominal contract. Consequently, some funds might choose to ‘build in’ a floor, to continuously and dynamically protect funding rates from falling further. This strategy could become part of a ‘journey planning’ process.
The real contract, on the other hand, is designed to facilitate a more flexible investment policy focused on long-term inflation-adjusted pensions. This could lead to an increased allocation to equities and other return-seeking assets and, where possible, with a direct or indirect link to inflation adjustment. An important element of the real contract is its ability to accommodate financial shocks and handle volatility in financial markets, with the minimum smoothing period being three years and the maximum 10 years. It is also possible to create a voluntary buffer. This methodology makes absolute return-type investments more attractive.
In the last few decades we have witnessed a clear trend towards market valuation and transparency, making the new FTK stand out somewhat as it clearly moves in the opposite direction. Examples are the introduction of the UFR, a fixed-risk premium, an almost fixed-inflation adjustment and the use of a 12- month average funding level. These mitigate or delay the impact of changes in market prices and we believe there are a number of drawbacks to using this approach.
We expect the financial position of pension funds would become very difficult to compare and on an absolute level to understand. When, for example, an employee moves from one pension fund to another, it would become a complicated issue to determine if accrued rights should be transferred or not. Pension funds would also run the risk that a change of economic regime would be recognised too late because of all the delaying factors mentioned above. In addition, risk management becomes a challenge because all hedging instruments are naturally based on market valuations. Furthermore, best-practice risk management is further compromised because of the divergence between the policy-based balance sheet and the market-driven balance sheet.
We believe that all these factors would complicate communication; especially if managed under the real contract, which seems to add more complexity. In this contract, for example, indexation is granted each year and – in the case of a deficit – a (smoothed) cut of pension rights is applied as well, while (smoothed) extra indexation is granted in case of a surplus. This could prove very difficult to communicate.
The new FTK is currently in the consultation phase and further changes are expected, while the exact setup of the UFR is still being researched by a committee. The ministry has also announced that it is looking into the possibility of a combination contract, which would combine elements of both contract types. Whether this resolves the unintended communication challenges and risk-management difficulties remains to be seen, and would largely depend on picking the best elements from each.
Michel Iglesias del Sol is head of investment strategy at Towers Watson in the Netherlands and Gerard Roelofs is head of investment for continental Europe at Towers Watson