As the discussion about the Dutch pensions system moves on to fundamental issues, a realisation is taking hold that a sustainable system must find middle ground between the traditional model of an unwieldy collective with opaque, mandatory solidarity and the ‘modern’ need for freedom of choice, tailored solutions and transparent risk-sharing. The best way to do so is by integrating elements of collective DB and individual DC, according to Theo Kocken, CEO of Cardano and a prominent Dutch pension thinker.

In the current second-pillar pension system, participants’ pension savings are pooled and individual plan members have no way of knowing at any given moment what share of the pot is theirs. That wasn’t considered a problem as long as plan participants could be sure of exactly what their payout would be, as in ‘defined benefits’. Besides, pooling all assets does have its advantages, including risk-sharing across the generations.

“But, lately it is becoming obvious that the benefits of such risk-sharing in our current system are limited to just a few percentage points. And since the financial crisis, it has become abundantly clear that one cannot be sure of the payout,” says Kocken. “When people have no clear view of what is going on inside the pot of pooled assets and neither be sure of the outcome the pot produces, they are bound to lose faith in the system.”

Proposals to resolve the problem have  focused on two extremes – a system of conditional rights, ensuring the possibility of a higher pension outcome but offering no security at all, or a system featuring high capital buffers to bolster security, but ensuring a lower outcome. Neither of these extremes really cuts the mustard.

The industry and the Department of Social Affairs and Labour are currently hammering out a compromise – a financial framework that features both higher capital buffer requirements and the provision that financial shocks must be accounted for without delay, while allowing such shocks to be absorbed over a 10-year period. Although most agree that such an interim financial framework is needed, this compromise doesn’t offer a lasting solution.  

“Such a solution must be found in a dual system comprising two components,” says Kocken. “Presently two very different functions – the accumulation of assets and the disbursement of pension benefits – are combined within a single fund. A simple separation of the two functions would go a long way.”

In this dual system, contribution payments would be deposited in a ‘pension accumulation fund’, where assets are pooled and managed collectively, allowing for the most cost-efficient approach. Plan members would participate with individual accounts, so each knows exactly what share of the fund they own at any given moment. Benefit payments would be managed by a second component – an annuity fund – which does not invest in risk-bearing assets, but instead focuses on cash flow matching of benefit payouts, so participants can be completely sure of the benefits they will receive. “Transparency during the accumulation phase and security with regards to the payout: these are considerable advantages as compared to the current system,” says Kocken.

As they age, plan members would gradually decrease their exposure to risk-bearing assets by exchanging their units in the asset pool for risk-free annuities at market rate. “That way one can gradually buy more security, while retaining the option to keep some exposure to the risky asset pool even after retirement. That is a big advantage compared with the current ‘PPI’ DC schemes, which are legally required to keep their plan members fully invested until retirement, at which point they must transfer the lot to an insurance company.”

Kocken notes that the timing and rate at which risky assets should be exchanged for risk-free annuities needs to be regulated to prevent people from hanging on to risky assets during bull markets, increasing the risk that they will be left empty-handed if markets crash as they are about to retire – a known problem in similar systems. “Our economic knowledge is limited, but as humans we tend to believe we know more than we actually do, so we need robust solutions to protect us from our own hubris.”

The dual system that Kocken proposes would also solve the problem of interest rate sensitivity that plagues the Dutch system. Under the present system, young participants are allotted annuities that are sensitive to interest rate changes and fluctuate over a lifetime. Even when they are older, plan members remain vulnerable to this problem, as older members are not exempt from interest rate sensitivity in the current set-up, he explains.

But as the proposed new system will only begin allotting annuities at a later age, the interest rate sensitivity of the system will be greatly reduced. In addition, the component that does remain sensitive to interest rate changes will be fully guaranteed and protected. “This will largely do away with the uncomfortable split between real and nominal policy objectives,” he adds.

The asset pool component of the dual system allows for various forms of risk-sharing including mutual insurance of longevity risks in terms of the risk that one should outlive one’s age cohort. The more tricky variety of longevity risk – the risk that entire age cohorts live longer than expected – cannot be dispatched quite so easily.

“Buying insurance from a commercial life insurer is not a realistic option as the insurance market simply isn’t big enough to absorb this risk; besides, the new Solvency II rules would make this a very expensive solution indeed,” says Kocken. “It would be best to leave it up to plan members themselves to absorb this risk, with the possible exception of the elderly. In advanced old age, part of the risk may need to be insured at market rate.”

One option would be to design ‘investible’ insurance policies that may be bought by ‘pension accumulation funds’: the risky asset pools that form the first component of the dual system.   

Although the concept of a dual system isn’t new, there has been discussion about the practical implications, says Kocken: “We are now beginning to get a clear view of how such a system might be implemented.  

Making the transition from an existing pension system to a new one is never going to be easy, Kocken realises. But the system he proposes has one huge advantage – there will be none of the trouble associated with ‘importing’ past accrued rights into a new regime.

“As the current system already serves as an annuity fund, it would suffice to launch an accrual fund to be run alongside it for new accrual. When the switch is flipped, participants in the ‘old’ scheme may be given the option to exchange part of their past accrued annuities to the new accrual fund.” Such an exchange is feasible because accrued rights in the old system can be priced unambiguously at a risk-free interest rate. “Joining the new system should preferably be presented as the default option, so that people would be automatically enrolled unless they explicitly choose to opt out,” he says, adding that a middle-of-the-road option might be added.

“As a matter of fact, the changes aren’t really all that extensive. Instead of a single fund in which all manner of opaque things take place, we switch to two funds offering absolute clarity about what everybody will get for their money. That’s all it is.”

The role of the Dutch pension funds, will hardly change. “Annuities basically grant the right to a defined pension benefit,” Kocken says. “In that regard, they are essentially insurance contracts that people buy with the money they have made by investing their pension contribution in the pension asset pool.

“Considering the size of the system, there is no way to buy all the necessary contracts from the insurance industry. As a nation, we have accumulated €1trn in pension savings. That is simply too much – the insurance industry can barely manage to absorb €10bn. Only pension funds have the wherewithal to meet this kind of demand. Pension funds today fill the role of mutual insurers, and they will continue to.”