The reason for the slow progress of DC is not that German employers feel so much more responsible than others for the size of their employees' retirement income. Nor are they reluctant to externalise book reserves because of any dislike of market investments. There are two main reasons for the phenomenon. The first is that the pension, tax and labour laws make it a prerequisite to define and guarantee a benefit, rather than a contribution. Second, tax laws do not allow the separation of pension assets" from an employer's other assets.

To understand the consequences one must understand what a DC plan means for the employer and his assets and liabilities. The main points are:

q The employer's only obligation under a DC plan is to make the agreed contribution. Pension expenses should normally equal contributions.

q The beneficiary takes the investment risk. The employer's expenses and assets remain unaffected by the performance of the pension assets.

q The pension assets are physically and legally separated from the company's assets and are solely allocated to benefit purposes.

q Vesting is restricted to the accrued assets. Any shortfall of assets will not be met by the employer.

q The employer is not responsible for any loss in retirement income from inflation or increasing life expectancy.

None of these major requirements can be met by a book reserve; the last can be overcome only if a lump sum design is chosen. However, book reserve lump sum plans are heavily taxed at retirement.

Why is it not possible for an employer just to agree on a contribution under a book reserve plan?

Suppose an employer says he will contribute 3% of employees' income to a book reserve account. Such a simple statement would not allow the employer to include the resulting expense in a tax-effective P&L statement. To obtain a tax deduction (a prerequisite for funding) he must specify the benefit at retirement. Once a benefit is defined, an expense (book reserve allocation) can be calculated and included in the P&L statement. Thus, the only way to set aside 3% of payroll tax-effectively is to commute it into a retirement benefit. The employer's obligation is not to allocate 3% of payroll but to accrue sufficient assets to guarantee the resulting benefits. Only if he guarantees the benefit will he receive a tax deduction for his commitment.

Needless to say, under such conditions the actual expenses for the guaranteed benefit will at no point match the technical contribution of 3% of payroll; only the total of such contributions, increased by interest, will match the guaranteed benefit.

The second obstacle is the separation of assets. Although the assets the employer builds up over the years may be physically separated from other assets, they cannot be legally separated. The only way of achieving such a separation is by a decision concerning the investment mode. Such investment is then split from the investment of other business assets. However, no matter how and where the book reserve-backed assets are invested they remain company property and share the fortunes or misfortunes of the company. Such a physical separation does not make these assets legally separated pension assets irrevocably allocated to pension purposes.

If the advantages of a DC plan can be achieved neither by market investments nor by plan design one asks whether there is any point in either designing a DC plan or investing "pension assets" in the capital market.

My answer would be "no" if it is done with the hope of achieving the advantages of a DC plan. My answer would be "yes" if the aim is a pension plan that avoids the direct link between income and benefit as under a defined (preferably final pay) benefit plan and simply defines a cost number for benefits. In communications with beneficiaries the expenses of a pension plan will be stressed rather than the pure benefit expectation. This will add a new option to the standard final pay plan, the career average plan or the fixed DM plan, which all focus on benefit rather than cost.

There have never been any restrictions on market investment of book reserve backing assets. The obstacles are different - mainly the tax issues.

A major obstacle is that one would expect assets backing book reserves to be exempt from immediate taxation and that taxation would be deferred until such assets are paid to the beneficiaries. Current legislation does not allow this. The investment is recorded on the company's balance sheet with the lowest value (purchase price or market price), meaning that no taxation arises on the investment's growth as long as it is not changed or traded. Interest or dividends are, however, taxed as soon as they are earned.

A second obstacle to market investment arises from immediate taxation of the growth in value once the investment is traded. This makes companies very volatile and may make market investment less attractive than the traditional use of book reserve assets as long-term loans and business capital.

Despite these disadvantages, the market investment of book reserve assets creates an understanding of the relationship between "pension assets" and "pension liabilities". The prevailing practice of reinvesting the tax savings achieved by the book reserve within the company means it is never clear to what extent pension liabilities are balanced by the assets earmarked for pension purposes.

This is another reason why book reserve backed assets are rarely invested externally. For it to be worthwhile, the yield of a market investment should at least equal the total interest paid on a loan of equivalent size, increased by the internal rate of investment.

The security of the beneficiaries remains unaffected by the kind of investment. As long as the pension assets are not legally separated from other business assets and must be used to back non-pension liabilities, the beneficiaries will have no particular advantage and no increased security.

Thus under existing legislation, market investment of pension backing assets does not yield significant advantages, regardless of whether it is combined with a DC plan or not.

A much better option is a support fund. As allocations to a support fund are based on the liability, the amounts of assets and allocations allowed under a support fund compare reasonably well with those under a book reserve, although they generally perform at a lower level. The investment of support fund assets in the capital market removes two significant disadvantages. Both the interest yield and the growth in value can be cashed tax free by the support fund, as it is generally exempt from tax (provided maximum permitted limits are respected); this improves performance. Second, as the support fund's balance sheet is not included in the employer's corporate balance sheet, the investment performance has little influence on operational results.

In many respects the support fund fares much better with a market investment of its pension assets and can be viewed as a real pension fund with assets taxed only if paid as benefits. Unfortunately the support fund rules also require a retirement benefit to be defined to qualify for a tax deduction on allocations to the fund.

I expect no real change in the immediate future. Government will only reluctantly tackle the complete restructuring of tax and pension law that would be necessary for a significant increase in market investment of pension backing assets and of DC plans funded by book reserves and support funds.

Dieter Kleylein is head of Bureau D Kleylein in Frankfurt"