The current pension funding discussions focus on how to fill the gap between pension assets and pension liabilities. Our opinion is that the exact formulation of the pension promise, ‘what does the plan participant get?’, needs to be addressed first. We also believe that the pension promise should be based only on what the underlying investments are able to deliver: nothing more or less. However, the pension plan participant should not be bothered with the ‘how’ of investing – a clearly defined pension promise is the only thing that counts.
We therefore make a case for structuring the pension product promise in terms of horizon, risks, contributions and benefits. This approach combines the advantages of defined benefit (DB) and defined contribution (DC) into a ‘defined benefit – contribution risk’ (DBC) pension product. When implemented the discussion about “the pension crisis and surprises” will be history. It will be about how much pension you need and at what cost you can get it.
The ‘how’ of pension investing is a topic widely discussed by regulators, consultants, employees, employers and several other interest groups. The reason for all of this is that there has been a rude wake up call for the stakeholders of DB schemes. The ‘don’t worry everything has been taken care of’ myth has been replaced by arguments over ex post interpretation of benefits. Benefit levels in DB schemes are being reduced, and employers are confronted with the potential severe effect of the volatility of pension asset prices on their company financials. Pension regulators are becoming stricter. Pension funds are required to create reserves at the moment they cannot really afford them and are stimulated to shift assets towards fixed income. As a consequence the chance of generating future reserves through asset returns are being diminished and the widespread belief in benefits of diversification of investments seems to be seriously questioned.
Taking the liability structure into account when formulating the pension asset mix is clearly becoming a theme in the latest discussions. For example, Bernd Scherer of Deutsche Bank Asset Management stated that the core of the pension problem is that “assets often showed little relation to liabilities at all and consultants will need to take more care of establishing the relative risk versus long term liabilities for a given asset mix rather than focusing on asset only solutions”.
Obviously from the asset management point of view there should be a firm link with the pension promise. Linking the investments with the promise is, in our eyes, an iterative process and not a one-way street from promise to investments.
The result of this iterative process is that the pension promise is nothing more or less than what can be delivered by the investments (and the contributions) underlying promise. It also identifies ‘false’ promises that cannot (yet) be locked in by investment structures. Examples include the indexation of the future pension payments to future wage inflation, the guarantee of a pension related to individual final salary, or the future availability of intergenerational solidarity.
Everybody wants this solidarity as long as someone else pays for it. So let’s take generations with different horizons as a starting point for taking a closer look at linking investments with the promise.
This means allocating promises and investments to horizons. There could be for example 35 different horizons (each representing a different generation of future pensioners) with the first being for example 2004 and the last 2039. This does not mean that promises are differently formulated for each target horizon, but that the assets needed at a future date (‘Target Capital of year X’ or ‘TC’) differ in composition. For example assets backing up the promise for the employees retiring in 2009 are very likely to be totally consisting of (inflation linked) bonds. Contributions received by the pension fund for this particular group would be invested in these bonds with a certain yield to maturity at the moment the contributions are invested. As this yield would differ at each of the future dates of contributions, the premiums for this age group could fluctuate within a predefined band. This is what we call the defined risk of contributions.
Arguments can be found to assume that the current investment portfolio for the employees retiring in 2025 might be structured differently as compared to the ‘TC 2009’ portfolio. However a similar methodology of contribution rate calculation would be applied : the premiums are calculated each year on a combination of certain and expected yields on new contributions and premiums would be allowed fluctuate within a pre defined range. So another characteristic of the DBC model is that contribution rates are expected to be different for each horizon.
The defined risk for the benefits is in the above-mentioned example a full guarantee of the Target Capital at a specific future date (Horizon), or zero risk in obtaining the nominal Target Capital. This can be supplemented with a “risky” top-up saving product where both contributions and benefits are defined within a bandwidth.
The Target Capital promise can be translated into a pension income guarantee if the future annuity rate applicable at the date of the Target Capital can be guaranteed as well. If not, an indication with a probability can be given. This is an example where the availability of investment instruments is influencing the pension promise.
Now back to the heat of the current discussions about pension funds struggling with funding levels and corporations being concerned about the impact of new accounting rules. What measures could be taken by the pension stakeholders to migrate from an existing DB into the new DBC system?
Based on what we see in practice, we believe a two-step approach – ‘quick repair’ and ‘full migration’ – would be most common.
The quick repair would be done in order to buy some time and get a release from the current pressure exerted by the sponsor and the regulator, for example to meet 105% funding by the end of 2004.
The quick repair is often implemented through boosting assets and reducing liabilities simultaneously. On the asset side we frequently see a lump sum payment by the sponsor - sometimes through a subordinated loan - in combination with structuring a (conditional) portfolio guarantee. Examples of short term repairing on the liability side we see include the conditioning of indexation, moving to average salary benefits, or increasing the age of retirement.
The “full migration” step is reducing liabilities to what can be backed by asset guarantees as described earlier. When the pension promise has been adjusted this will lead to a new aggregated liability calculation for each horizon year (2004 – 2039), as depicted on the right side of pension funding balance sheet as depicted in the chart.
Matching the total liability levels at different Horizon dates with target assets (Target Capital, see left side of picture) will result in a full balance between the two.
But what to do with existing total former DB plan assets? Part of these could be allocated to each bar, other assets could be restructured to match the asset profile of the particular bar. The allocation could be done taking the remaining life of the particular Target Capital portfolio into account, and could result in a different asset allocation amongst bars.
The balancing process will result in a situation where each age group would have a fully funded (or stocked J) bar. The guarantee provided to the employees retiring at 2027 is then based on the assets of the Target Capital 2027.
This translates into a pension promise for employees retiring at for example 2027: “With an annual contribution of between E14,500 and E15,500 paid until 2027 (jointly by the employer and the employee) an annual income of between E50,000 and E51,000 is guaranteed as from 2027 increasing each following year with 2% annually during your remaining life”. This is how the DBC pension product concept is formulated to the participant.
Once this situation has been reached there is no material difference anymore for the participants between a DB or DC plan. Also there is no difference anymore between individual or collective rights within the target Capital 2027 employee group. For the employer the DC characteristics of this DBC pension plan would give a major relief from the reporting requirements and possible future balance sheet implications. There would be no reserve requirements anymore for a guaranteed DC plan, as there are none for non-guaranteed DC plan. If the pension plan would give individual rights to the assets in each target capital group, the possible available reserves inherited from the previous DB plan could be partly allocated as a bonus for the participants. The pension right administration could be simplified for the pension fund, the regulator could be happy with a safe version of DC, and the employer will be relieved from any possible future under-funding obligations.
In the meantime work remains to be done on developing the DBC product concept further for full implementation. It can be enriched with several features like for example life insurance modules and a financial planning model.
The core remains however a “clear to understand” fully investment guaranteed product, where costs, benefits and risks have been made fully transparent. The DBC concept combines DC and DB characteristics in one. It does deliver on the promise of no more future under-funding.
Jeroen Tielman is CEO at FundPartners,
a pension product engineering firm based in the Netherlands