There is a lot of hand-wringing going on in Canada about the generally poor financial condition of corporate, public sector, and industry-wide defined benefit (DB)pension plans. Plan members, employers and regulators are worried too.

Interestingly, despite all this concern, there is no consensus or how to fix the problem. Some say tighter funding rules are the answer. Others want looser funding rules. Some say pensions should vary with ability to pay. You get the idea: little consensus or anything.

Why is it impossible to reach agreement on how to fix DB plans? I fear the answer is because they simply can’t be fixed, at least in their current form. Where did this heretical idea come from? It came from my experimental attempt to build the ‘ideal’ pension plan from the ground up. I started with the premise that post-retirement income streams(i.e., pensions)are intrinsically good things. So why don’t people organize to create a stream of future pension payments themselves?

There are two problems: behavioral and institutional. The behavioral problem is that most people have neither the knowledge nor discipline to engineer the construction of such future payment streams themselves. The institutional problem is informational asymmetry between individual pension savers and the for-profit financial services industry. The predictable result is that the pension savers end up paying too much for too little.

Fortunately, both problems have a single solution: an auto-pilot savings-investment-annuitization protocol managed by an expert co-op organization of sufficient scale to operate at low unit costs. Does this formula work? Absolutely. New York-based TIAACREF, one of the world’s largest retirement systems, has used this formula to successfully provide adequate post-retirement income streams to millions of U.S. university/college retirees since 1917.

In what sense is a DB pension formula superior to the TIAA-CREF formula? In theory, because it provides intergenerational risk insurance. In other words, the DB formula offers successive generations of plan members the possibility of equal pensions despite experiencing differing investment regimes(e.g., an extended high return period, followed by an extended low return period, etc.). That is the theory.

Can it be made to it work in practice? Unfortunately, not very well with the underfunded DB plans undertaking significant amounts of asset-liability mismatch risk. Such structures require fair-value costing as well as aggressive ‘steering’ (i.e., combinations of dynamic contribution, benefit, investment policy adjustments). Otherwise, they will eventually degenerate into unfunded pay-go plans costing 30% of pay or more. Such aggressive ‘steering’ in turn cannot be accomplished without the legal requirement to apply modern finance principles (e.g., contingent claims analysis, risk-based capital reserve requirements, etc.)to measuring, managing, and regularly disclosing, DB plan balance sheets.

The Dutch DB pension system is actually moving to this modern finance-based management regime. Funding and costing is moving to a transparent, ‘fair value’ basis. Risk-based capital reserve requirements for DB balance sheets are being imposed. Flexible benefit levels are being tied to balance sheet funded ratios. Contribution rate adjustments are being tied to explicit formulas, rather then opaque actuarial rules of thumb.

So why is Canada mired in the dysfunctional DB model while the Dutch have been busy re-inventing theirs based on modern finance principles?

First, the Dutch have carried on an active, informed pension debate for years now. Second, when a better way was discovered, the pension regulator was prepared to take flak from the Dutch pension industry, and force it to adopt the modern pension finance framework. Is it too late for Canada to take these progressive steps? Time will tell.

Keith Ambachtsheer is president of KPA Advisory Services Ltd., a strategic advisor to major pension plans around the world, based in Toronto.

For a PDF version of this article, click here.