Is it possible to compare different employers’ pension plans?

Buck Consultants in Geneva certainly thinks so and has devised for the Swiss market a method of calculating a ‘technical value’ for each employer’s plan. The exercise was based on a database consisting mainly of internation-al companies operating in Switzerland.

We have established a systematic, objective technical valuation method, which has been used to assign a value to each benefit of each plan and for the aggregate value of each employer’s plan,” the consultant claims.

The aim is to set up a procedure to assign value. For example, if two plans are identical except that one’s benefits are exactly double the other’s, its technical value will be double. But the consultant recognises that this will not occur in practice, so the methodology assigns values separately to each benefit, but in a way that correctly reflects the difference in value between them.

Values are calculated and expressed as a constant proportion of payroll payable throughout the career of the member, for both defined contribution (DC) and de-fined benefit (DB) schemes, as well as for the risk benefits.

To do this, the consultant says the “individual entry normal constant percent” actuarial cost method has been adopted. “The so-called ‘normal cost’ of a plan under this method is the constant percentage payable from entry to exit for each individual participant.”

In the case of a DC plan with a constant contribution rate, the normal cost is in fact equal to that contribution rate. But for a DB plan, the future pension benefit is projected, then converted into the capital equivalent and the constant payroll percentage required to obtain that projected benefit is calculated.

In its initial study, Buck says it has evaluated all plans as-suming an identical employee profile. The profile is a 45-year-old man on a current salary of Sfr100,000 ($68,000), who entered the plan at age 35, married with two children.

As the technical valuation process consists of projecting the amount of future benefits paid out by the plans and converting these future amounts in the form of constant percent annual contributions, a number of assumptions need to be made. These include: a 4% interest rate, which is used to convert future amounts into present values and an-nual contributions; for DC plans the rate of return credited to accounts is also put at 4%; salary growth rates are put at 3.5% pa. Other as-sumptions are basing mortality and disability on specified Swiss tables, using set leaving service probability figures. The actual figures used for leaving probabilities at age 45 are 1.2%, at age 50, 0.4%, and at age 55, 0%. These are low probabilities and result in the assignment of “relatively low value to the leaving benefits of the plan”.

Plan participants are assum-ed to retire at age 61, given the pattern of earlier retirement in Switzerland. But the comparative calculations for age 55 and 65 retirrals should also be looked at, Buck says.

To arrive at the aggregate value, the retirement, death, disability and termination benefits of each plan are calculated and then totalled up for each plan. Where a plan is contributory, the actual proportion contributed by the employee is deducted, to ar-rive at an aggregate net value. Should the contribution not be a constant percent of total salary, an “equivalent constant percentage is calculated” and deducted.

A problem that can arise when arriving at the aggregate technical value is what the consultants refer to as a kind of ‘double counting’ of values. Taking the example of a DC plan with a 10% contribution rate of 10%, where on death, retirement, disability the benefit is always equal to the sum of the accumulated contributions plus interest, so the plan should have a value of 10%.

But if 80% of people retire, 4% will leave, 8% die and 8% become disabled, then “in a sense one may say that the retirement value is 8% (80% of the 10% total value), 0.4% is the termination benefit value and 0.8% each is the value of the death and disability benefits”. Under the conventions adopted, 10% is the value atributed to retirement benefits, being equal to the contribution rate, but to this would be added the 0.4% and the 0.8% twice for risk benefits, giving a plan value of 12%.

The median aggregate technical value of participating schemes was 20.83%, with the best case having a value of 32.64% and the worst 11.85% (see chart). Deducting employee contributions produced a median net aggregate figure of 15.49%, with a low of 5.38% and a high of 24.51%.

The consultant says that having set up the database, it is now possible to generate additional analyses, including changing the assumptions or having alternative employee profiles. Buck says that a systematic procedure that provides an index of value can be used to compare plans. But it adds: “We believe it to be dangerous and foolish to suppose that all the information that is needed about a pension plan can be summarised in a single figure.”

The consultant, in setting up the Swiss Pension Forum, which is open to companies that are not its clients, has created the database with details of pension and employee benefit plans. Altogether 37 companies provided sufficent information.

About two thirds of those taking part were private international companies, a quarter were Swiss private companies, with a number of Swiss public companies and international organisations. Among these were Amoco Chemicals (Eur-ope), Battelle, DHL, Digital Equipment, The Red Cross, Novartis, Hewlett Packard, Elma Electronic, Honeywell Lucifer, Merck, Motorola (Suisse),Otis, Union Carbide Europe.

A fifth were in engineering technology,with financial services and banking ac-counting for around 18%, chemicals and pharmaceuticals 13%, and computer and high tech 11%. “

Topics