This expression has been widely used in recent pension reports and consultations. It has wide currency in the world of insurance, and academic finance, where it is a term of art. Many of its recent uses are deeply problematic and unhelpful.

Moral hazard refers to the situation where, once insured, we may cease to show proper care in avoiding the incident covered by the insurance. For example, we may fail to check, as diligently as when uninsured, that all windows are closed and locked after we have taken out household contents cover. Policies contain a number of mechanisms to limit this increased risk exposure; such as a deductible or first loss which the insured bears, or a requirement that there be signs of forcible entry.

One important element is that the beneficiary has control or influence over the situation. Although the reductions in pensioner benefits imposed by the Pension Protection Fund have often been presented as standard moral hazard mitigants, they are not warranted as such. The member does not have control over the behaviour of the scheme, fund or sponsor employer.

The scale of those member deductions is also troubling; they are very large. They may be estimated by considering the valuation of the PPF benefits (s179 value) and the valuation of full all risks buy-out of those same benefits. For the most recent (Tranche 9) triennial valuations, the average PPF benefits lie between 65% and 70% of full benefits. The PLSA consolidation paper is right to be concerned by the extent of these deductions.

It also raises the important question of just how much risk is being covered by the PPF. With average scheme funding at 63.7% of full buy-out, and 95.1% of PPF benefits, the answer (on average) is precious little.

A risk process in which such high deductibles were rationally justified, would be one which was extensively under the control and influence of the insured. For this reason, it would be expected to have few underwriters. However, the risk process here is insolvency, and that is the most widely underwritten class of risk in all finance; it features centrally in banking, insurance and investment management.

The prudent valuation, “technical provisions” standards, are indeed prudent, probably recklessly so, at 71.2% of the buy-out price on average. This is 142.4% of the best estimate of liabilities. In the post-crisis period, banks have objected strenuously to regulation requiring coverage of less than 120%.

The fact that we have bought “all-risks” household insurance does not confer on us the right to burn the house down and claim under the policy, though of course burning the house to the ground is within our rights, provided this action is harmless to others. Indeed, such a claim, arising from our actions and intent, would be fraudulent. The householder may be behaving in an amoral or immoral manner, but the circumstance is certainly not one of moral hazard. It is a clear breach of good faith, and as such voids the policy.

With that in mind, it is surprising to see references to “unscrupulous” employers deliberately abandoning schemes. A very strange employer this: prepared to abandon the scheme, but only when members, of which few are likely to be current employees, have some partial degree of benefit protection.

Properly that abandonment behaviour, and its myriad intermediate shades, should be analysed in the context of the employers’ incentives, and there are many such incentives arising from the capricious and arbitrary nature of much of current accounting and valuation practice, and the associated costs.