Under the current actuarial rulings (APP), pension funds are allowed to use interest rates up to a maximum of 4% to discount liability streams. This has to be done in a prudent fashion. Specifically the words ‘maximum’ and ‘prudent’ seem to be ignored quite regularly, as it is almost a given that pension funds automatically use the 4% limit.
Irrespective of using variable, market-related discount rates to be used under the new financial assessment framework (nFTK) , or using a fixed rate, a certifying actuary cannot close their eyes to the duty of being prudent both in the way the technical provision (VPV) and the (cost recovering) contribution levels are being calculated.
To get some sense of reality, effective yields of (government) bonds with a duration of 15-20 years lie in the area of 3.7% for nominal bonds and 1.5% for real bonds. Is it prudent for a certifying actuary to stick to the ‘4%-rule’, if the market tells you differently? Frankly, the answer to that question has nothing to do with the current ruling (APP) nor the new ruling (postponed until 2007), but is one of being impartial and prudent.
Postponement of the nFTK might lead to a situation where pension funds won’t find themselves in a so called ‘under reserve position’. This is a situation where funding rates will fall below pre-described solvency levels, leading to the installation of recovery plans. Recovery plans are to be submitted to the regulator, the Dutch Central Bank, and the plan must show that the pension fund is ‘solvent’ again within a 15 years time frame. Submitting a recovery plan for approval and sticking to it is, among others, quite an administrative burden. Using a discount rate of 4% instead of, say, 3.7% might just keep you out of the recovery docks.
A simple example might show this. Assume there is an average Dutch pension fund (50% equities, 50% five-year duration bonds) with a funding rate of 120% (under the 4% ruling). If this fund was to change to the nFTK, whereby market interest rates are to be used - in the former example 3.7% (ie, a drop of 0.3% ) - then its funding rate would drop to 115.7%. Under market value conditions, and with no interest rate matching policy in place, this pension fund would most likely fall into an ‘under reserve position’, ie, the average Dutch pension fund would.
Reason enough then to postpone the nFTK until January 2007? No, not from an economic view point, because it might prevent you from solving the interest rate mismatch risk for another year. In our view, the interest rate mismatch shown on an average pension fund balance sheet is a so called ‘unrewardable risk’ and should, strategically speaking, be matched. Any freed-up risk budget as a result of matching should be allocated to ‘rewardable’ areas with a positive risk premium.
Anyone with an ultra bright and clear vision that the absolute level of the euro yield curve will rise between now and January 2007 might disregard the foregoing. Anyone expecting interest rates to rise, but can’t afford them to fall, might consider asymmetric hedging (swaption structures). The latter always comes at a price.
Why would postponement be a good thing then? Well, practically speaking, pension funds are currently very busy implementing new arrangements, new indexation policies, installing new administrative systems, complying with new monitoring and reporting requirements (see new consultation document, dated 26 September 2005), dealing with new laws like the VPL and, last but not least, the impact new accounting rules (IFRS, FAS) have on the sponsors’ balance sheet and profit and loss account.
Pension funds are just very busy and forcing them into the nFTK environment too soon, might lead to unacceptable levels of governance risks. Not having enough time, knowledge and resources to cope with the foregoing points is a recipe for disaster for the whole of the Dutch pension system and this should be avoided.
Under the nFTK, longevity risks, the dispersion of the trend in longevity (TSO/NSA) risks, and insurance risks are explicitly taken into account in the calculations of both the liability streams and the required solvency buffers of a pension fund. This is an extra - but prudent, some might say over-prudent - burden, that easily lowers the funding rate of a pension fund by
3-4 percentage points. The general verdict is out on whether this impact shouldn’t be transferred solely to the required solvency buffer. More of this is to be expected.
What about investments and the market? Solving the interest rate mismatch has gathered pace in the Dutch market. Postponement can now lead to odd developments if one considers funding rate developments. Suppose a pension fund has made its balance sheet ‘nFTK-proof’. In other words, the asset portfolio has become a replica (at least for a certain percentage) of the market-related development of its liability streams, so that one has assumed that both sides of the balance sheet more or less move in similar ways, as a result of changes in interest rates.
One does not need to be a genius to see that being matched in a market value environment - ie, all things being equal, the funding rate volatility has been lowered to an acceptable level - means being unmatched in a 4% environment. The latter will actually lead to an increased level of funding rate volatility, which cannot be the goal in the first place. Pension funds facing this situation are wise to report their funding rate development both in the 4% world and in a market value environment, in order to prevent being penalised for the wrong reason.
I suppose a good thing about postponement is that pension funds can buy time to figure out how to fulfill their indexation ambition. This is often called the ‘flip side to matching’. In a matching and return environment, the remainder of non-matched assets face the goal of delivering indexation payments for the future. This is also called the return-seeking portfolio. The way to structure the return-driven components should preferably contain elements of sensitivity to unexpected inflation, absolute return drivers and a high level of diversity. In other words, this is not a portfolio that is being structured in a split second. Therefore, some serious thinking about this part of the portfolio is a necessity.
Final remarks are on the market. Is there an impact on the market to be expected? Many people argue that the ‘pressure has come off’ the long end of the yield curve now. Supposedly, the demand for the 50-year French OAT has decreased, therefore downside pressure on long-term interest rates have decreased. Instead of 100 people rushing at the same time to get into a tube, 100 people are now willing to stand in line for a year and to pick and chose the tube they wish. Frankly, they are still the same 100 people wanting to travel and the driver (the market) knows it. If the forward rate says anything meaningful about the future spot rate, then I doubt if the downward pressure on interest rates have completely levelled off. Then again, I am just a passenger.
Gerard Roelofs is a partner and head of investment consulting at Watson Wyatt in The Netherlands