In my recent article on the ‘false logic’ in the FCA’s guidance on transfers from DB pension schemes to personal pensions, I alluded to the risk of awards by the Financial Ombudsman Service (FOS) being biased by that logic. Searching FOS published case histories involving transfers from the British Steel Pension Scheme, to see if they included the details I wanted of the terms available at different points for different ages, I found a clear example of bias, made worse by a fundamental error of reasoning. The case is here:

In setting out the reasons why the ombudsman decided a transfer was not in the claimant’s best interests, the first and longest section of his argument is about the ‘critical yield’. This is the investment return that must be achieved in order to provide the capital value at retirement which, when applied then to the purchase of a pension annuity (assuming current annuity rates), would generate a guaranteed real income equal to the defined benefit pension. If you could not achieve that investment return, you would be worse off than keeping the pension. If you could achieve a higher return, you would be better off. But only if you planned to buy an annuity. It does not measure whether you could be better off with neither the defined benefit pension or an annuity. Such as with drawdown.

Post pension freedoms, reference to the critical yield is logically only relevant if the case for drawdown has already, for other reasons, been found to have failed. The improbability of reaching the critical yield by retirement age, by whatever return assumptions are used as evidence, cannot be a reason for finding drawdown unsuitable. That is circular reasoning because the superiority of the guaranteed income is presumed, not tested for. What the ombudsman needed to consider was whether the option of drawdown chosen by the claimant was likely to increase personal welfare relative to the defined benefit pension or its annuity equivalent. We don’t know the answer, even though the adviser must have included an outcome comparison assuming drawdown, because it simply isn’t addressed by the ombudsman. I have no idea how prevalent this reasoning is but it’s worrying even to see a single instance.

There are many aspects to the person-specific utility provided by flexible benefits over fixed benefits, other than comparative income, and they were significantly more complicated in the particular circumstances of British Steel. The ombudsman goes on to address these. It sounds like the adviser did themselves no favours by not anticipating and addressing possible reasoning against their own – a common failing in financial advice generally. But what is clear is that the ombudsman’s determination of other aspects of utility was prejudiced by the false logic applied to the income comparison. Whether you are likely to enjoy better real retirement income, and with what (if any) risk of shortfall, has to be the primary test of utility, even if it is, in almost all cases, a necessary but not sufficient condition. Fail the income comparison and the other factors will not normally compensate. This was not a case where they obviously would.

The ombudsman having found in favour of the claimant, the problem of gaming the redress approach (a key point in my article) applies here. The redress will be paid as a top-up to the claimant’s existing personal pension. There is no requirement to ‘derisk’ the claimant’s existing investment strategy (which in this case included self-investment of tax free cash in property, with a stated risk-diversification motive) in line with the asserted ‘true’ attitude to risk. No consideration is given to the possibility that the choices made about the replacement investment strategy might themselves have some bearing on the true attitude to risk. This is the kind of internal contradiction that a legal process might be expected to expose and prevent.

It is always invidious to take a single example and rely on a summary of an ombudsman’s findings, rather than a complete assessment of the case. Most of us in the industry, and not just smaller firms like ours, accept that we benefit from the halo effect of consumer protection schemes that are more easily accessible than a full legal process, even if they are not perfect. But it is not acceptable that the alternative is prejudiced by mistakes made by the FCA that then taint FOS’s procedures. Or that it is easily open to gaming. But it is also unacceptable to see individual errors of reasoning, such as in the case I have quoted, go uncorrected.