As Japan gets closer to making some sort of tax concession in favor of defined contribution (DC) plans, we thought we would, once again, bring up some of the issues that should be considered when considering replacing an existing defined benefit (DB) arrangement with a defined contribution one.

Like most insurance-related illustrations, DC cash-flows are very misleading, even when they are designed with the client's interests at heart.

One of the most problematical areas is the selection of appropriate actuarial assumptions for the comparisons. Some actuaries and everyone who makes his money investing assets would say that, for an "apples-to-apples" comparison one should use the same discount rate. But is that right in all, or even most, cases?

When doing "real" manager selection, you will find your asset consultant talking about a concept known as the "efficient frontier." The basic concept is that for a given level of expected return for a portfolio, there is one diversified mix of assets that is lower risk than all of the others.

As one graphs these points for a range of expected returns, one will see a curve develop. This curve shows the increasing relationship between investment risk and investment return. It also shows that the relationship is not linear and that for relatively small increases in risk, fairly large increases in return can be expected.

A client can then select a level of risk and see what long-term expected return results. The trade-off between risk and return can then be examined and the client can choose the diversification that is at a level of risk he can live with.

Point is however, except for a small set of singularities related to lower returns, higher average risk translates to higher average return. Of course, the problem with averages is that they always have tails and if your investments end up in the "bust" tail, no amount of efficient frontier theory will make you feel you made the right decision!

So, the greater the asset risk, the higher the expected portfolio investment return.

Now, what are the characteristics of players who can bear higher risks? Without going into details, we will claim that employers can bear greater risks than employees.

An employer can accept a higher level of risk in his investment portfolio than an employee can.

This translates to higher expected return. The efficient frontier shows that even small increases in risk are worthwhile choices, if you can stand the down-side. It is entirely reasonable to expect an annual difference in expected return of full points (hundreds of basis points) based on the difference in risk that an employer can bear versus what an employee can bear.

Those extra investment points belong to the employer and translate into enormously lower retirement plan costs over a long period when the plan is a DB plan. A DC plan costs exactly what the employer agrees to contribute to it. To the extent that employees are not paid less by a company providing a DC plan than one providing a DB plan, the DC plan will cost more when designed to provide exactly the same benefits.

Another cause of higher costs in a DC plan that occurs in the U.S. but not so much in Japan is the fact that, in the U.S., DC plans provide much greater benefits to departing employees than typical [1] DB plans. This reflects important differences in basic design between the two countries. Given equal investment returns, the retirement plan providing greater benefits will cost more.

A concern that is not a cost issue is the variation in individual results. Most retirement plans, even DC, are designed with a certain amount of benefit at specific ages in mind. DB plans, by their very nature, provide those exact amounts without variation.

DC plans never produce the designed results. Where an employee has investment discretion and he has a bit of luck, he gets a lot more than designed. Bad luck produces the other result.

If the employee does not have investment discretion, required conservatism (such as the need to protect principal) produces a lower - substantially lower - long-term return. Of course, that return is not flat and luck still plays an important role.

These situations also help explain why money managers and brokers are so keen in getting DC plans approved in Japan; more money will be needed to be invested in order to get the same results that are currently achieved with DB plans. Look at what has happened in the U.S. as regulation has discouraged the DB approach and companies have adopted DC plans. More money invested translates to higher incomes for those in that field, regardless of how employees fare in retirement!

So the bottom-line remains the same as it always is: the money you spend in the design stage of your retirement plan will more than pay itself back as you avoid foreseeable problems ab initio. Your retirement plan is a major use of company funds, make it do something useful!