Test 1 is a self-defeating Catch-22

In this post, I argued that the following principle underlying Test 1 is already satisfied by USS’s current equity-weighted growth portfolio: Employers should have a greater than 95% certainty of never having to raise their contributions by more than 7% in order to provide promised pensions.

One might respond that, even if an equity portfolio carries such greater than 95% certainty, the less than 5% of scenarios in which it fails to deliver promised pensions might involve downside risks of greater failure than a self-sufficiency portfolio which is heavily weighted towards bonds. That would provide a reason to avoid equity even if it satisfies the above principle in bold.

But if First Actuarial’s cash flow analysis is broadly accurate, then, even if there’s a catastrophic crash in the stock market, the combination of regular employer DB contributions, the extra 7% in extremis contingent contributions, and dividend income will be sufficient to keep the scheme cash flow positive. Therefore, the scheme won’t be exposed to the disinvestment risk of having to sell equities when their price is depressed. As this graph by Robert Shiller indicates, dividend income remains on a steady, upward trend, even through such catastrophic drops in the price of equities as happened during the Great Depression:

Source: Robert Shiller, Irrational Exuberance, 3rd ed.

I am assuming that the scheme remains open to future DB accruals on salaries up to £55,550. If, however, an in extremis scenario involves a closure of DB (or if contributions into DB become insignificant because the DB/DC threshold has been moved way down), then the scheme would become cash flow negative. In this case, USS would need to start selling its assets to make pensions payments, which would expose it to far more disinvestment risk than at present. (See this post for further details on USS’s cash flow position and its significance.)

But why close (or greatly reduce) the DB section? Only because it has become unaffordable on account of the lower investment returns that arise from Test 1's requirement to shift the DB portfolio towards bonds. If, by contrast, USS is allowed to remain open, ongoing, and invested in its current growth portfolio, it would be able to continue to fund the DB section at the current level of employer and employee contributions.

So we need to ensure that a self-sufficiency portfolio is within reach of a 7% increase in contributions — i.e., we need to meet Test 1 — only if closure to future accrual is a likely scenario. But closure to future accrual is a likely scenario only if Test 1 de-risking is imposed. Test 1 is a self-defeating Catch-22.

Test 1 is intended to protect against “the need to sell or mortgage assets to fund the scheme” (see p. 10). But the imposition of Test 1 actually generates the very need to sell assets that it is supposed to protect against. This is because the test’s requirement to shift the portfolio towards bonds makes DB unaffordable, which forces its closure, which pushes the scheme’s cash flow into the red, which forces it to sell assets to make pension payments. Get rid of the Test 1 requirement, and the DB section can afford to remain open at its current level, which will keep the scheme cash flow positive in a manner that protects against the risk of having to sell equity.

(Please read the quotations from Derek Benstead of First Actuarial in the PS to this post for more on the manner in which Test 1 leads to the replacement of DB with demonstrably inferior DC pensions.)