What (little) pension trustees should know about investment
In the continuing debate about how to strengthen the effectiveness of trustees of salary-related ‘defined benefit’ pension funds, and particularly their investment understanding, the essential point seems to have been overlooked. Logically, trustees in this case should have nothing to do with investment decisions, only investment arrangements. Otherwise they are acting as if the consequences of investment decisions only impacted the scheme members when in reality, as intended by a series of legislative and regulatory changes, they impact the fund’s sponsor and, when a compensation pool exists, the sponsors of all funds collectively. Following these changes, the conventional hope that sponsors and members share common risk utility functions looks wildly unrealistic. If trustees turn up to courses to learn about investment principles, this is something they could usefully take away with them and use to redefine the sponsor’s expectations of their role.
Before I set up a ‘family office’ for individual investors I briefly flirted with the idea of adding the role of occupational pension fund trustee to my other activities. I was responding to a call put out by Christine Farnish, head of the National Association of Pension Funds, that schemes needed experienced investment people like me. I backed off when I realised that with new accounting rules I would be in an impossible situation along with the entire body of trustees of whatever fund cared to hire me.
I arrived at this conclusion by thinking about how I would take, or help my colleagues take, high level investment decisions (those explaining virtually all of a fund’s return path) in the specific context of a salary-related (defined benefit) scheme where deferred promises are earned as worked and reserved, but not paid, as earned. Though the idea of a salary-related pension is increasingly talked about as if deferred pay, it is the promise (once only thought of as an aspiration, economics permitting) which is owned by the member, not the money reserved on an accrual basis to meet it.
As a financial adviser, rather than a trustee, I would expect to be assisting the sponsor to set and manage the reserves. In the absence of any regulatory requirements dictating this function, I would advise on the basis of a risk utility function shaped by costs and consequences in much the same way as an individual would think about meeting a future liability such as the repayment of an interest-only loan. As an adviser to an individual, I would likewise seek to inform personal choices based on costs (resources required) and consequences (which I could imagine if I were in their shoes but could never think about exactly as they do). The effect of new accounting rules, as explained in an two earlier items on equity risk, is to make the costs and consequences highly specific to the rules, changing radically the rational choices players make. Though it was always simplistic and optimistic to assume sponsors and members could share a common utility, it becomes foolhardy now.
The main effect is that the traditional scope to trade off certainty of outcome against the contribution rate has been closed off to sponsors. You can still invest in risky assets but effectively you have to fund at the same higher level as would be required if investing in riskless assets. It’s like saying to an potential mortgage endowment purchaser ‘you can invest in this balanced fund to meet your liability in 25 years but the premium or contributions will be based on an assumed growth on line with a gilt’. That is in fact how endowment premiums were set until somebody decided that you didn’t need to be 99% certain of a surplus to allow some of the expected risk premium over 25 years of investing in riskier assets to be applied to reducing the premium.
What did for endowments was what also did for salary-related schemes: setting the premium at a level equivalent to about a 50% certainty of hitting the target instead of somewhere nearer but still short of 99%. The sellers of pension promises and endowments probably exhibited a very similar utility function, as the decision makers stood to benefit immediately from the savings and a different lot would pick up the bill if they got it wrong. Rather than speculate about the integrity failings here, I simply take these as object lessons in ‘streetwise’ finance: treating all agency relationships as contests and trying to fathom the rules of the contest before playing.
Salary-related pension promises now, by act of parliament, must be treated as guarantees, not a promise with no better than a 50% chance of being kept. In funding terms, this means they now require require certainty as near 100% as it is practical to be. For various reasons to do with inflation risk, I suspect it is actually no higher than 95%.
International accounting standards have the effect of forcing any underfunding of the promise along the way into the open. Prescribed rules for calculating the fair value of a transfer out, similarly based on market-related discount rates, will have the same impact. The effect is to close off the trade-off funds have traditionally made between outcome certainty and contributions. The issue of investment risk is thereby transformed.
First, investment risk becomes not only separate from but also subordinate to contribution adequacy. Since in a salary-related scheme the latter is now an actuarial function (based on identifying cash flows out and matching bond-based cash flows in) and an accounting function (recording any underfunding or overfunding in a prescribed fashion) there is little scope for either discretion or control by boards of trustees.
Second, if a fund chooses to take investment risks that call forth a higher level of contributions (and higher premium payments to the compensation fund) than is otherwise required, it is the sponsor’s direct stakeholders that have to consider the utility of that, not trustees. The members clearly do not have the same utility. They bear little of the cost of a riskier policy when it is funded accordingly but might hope to benefit from the surpluses in higher benefits, as it is easier to put money in than take it out. It is typically therefore shareholders who should ask why trustees are being trained to take these decisions better when they should not be taking them at all.
When you take away the trustees’ investment function, individual boards of trustees look like an expensive and inefficient way to secure the stewardship, custodial and oversight functions of these schemes. As pension plans themselves move to matched and fully funded reserving, then, just like closed with-profits funds, specialised collective solutions to manage the cash flows in and out without investment risk look more efficient than individually managed structures. Even if the sponsor’s fiduciary function cannot also be outsourced (though I cannot immediately see why), it is not what it once was or what most of the people involved assume it still is.