© 2016-2019 Fowler Drew Limited

  • Black Facebook Icon
  • Black Twitter Icon
  • Black LinkedIn Icon

Authorised and regulated by the

Financial Conduct Authority 

  • Stuart Fowler

FCA decisions on DB transfer advice

Today's FCA policy statement following its consultation on advice on pension transfers shows it has listened to some of the main objections to its proposals for strengthening a clearly fragile advice process but it will leave unsatisfied objections from both opponents and supporters of DB transfers. There are aspects of the paper that are profoundly worrying, notably the implied divergence between Parliaments' intentions and regulatory attitudes, as between freedom and paternalism, which the paper does not satisfactorily resolve, including where the resolution could have been detailed evidence of why it believes up to a third of transfers have been unsuitable.

The headline aspect is that the FCA has withdrawn its stated intention to drop its prior assumption that transferring from a DB scheme was unlikely to be in a member's best interests. We don't attach much significance to this 'starting point', as it was anyway contradicted in the consultation paper and besides it is what the Financial Ombudsman (FOS) adopts as a starting point that really counts. This change in FCA approach it really only tokenism as advice should always stand or fall on the conclusion and the logic leading to it.

The FCA has not budged on the advice industry's protests that the new Transfer Value Comparator, replacing the current Transfer Value Analysis, still assumes that an annuity will be purchased at retirement. This assumption is in stark contrast to the Government's own observation, confirmed at the recent Select Committee hearings on Collective Defined Contribution pensions, that annuities were suboptimal - that being the prompting for the Pension Freedoms and the subsequent shrinkage of the annuity market. But this is perhaps to miss the point of the FCA's stated purpose of the TVC, which was to force advisers to show the (very high) monetary value of the benefits being given up. The TVC is a prescribed method for calculating a market-derived present value, as a sum of capital, required to replicate the DB benefits. But the FCA has at least acknowledged that its original proposal did not do this: it assumed a risky approach up to retirement and then a risk free approach at retirement. But the point about the benefits being given up is that they are risk free at both stages: before and after retirement. We were probably not alone in pointing this out in our consultation response. The FCA has now conceded that the return assumption for the period up to annuitisation at retirement should be based on a notionally risk free gilt yield. There are some issues around the prescribed yield basis but the principle is at least now consistent with the stated intention.

Though the FCA has stuck to its guns on the annuity assumption, it has at least stated clearly that the relationship between the TVC capital sum and the CETV capital sum should not become the basis of a simple test, or crude focus, of suitability in the way that, for instance, an arbitrary implied 'implausible' critical yield under the old TVA was adopted by some PI insurers as a proxy for the risk of bad advice. That's good to hear but it's actually the Financial Ombudsman Service that needs to take note. It too has fallen into the trap of placing too much emphasis on the required yield and not enough on the term over which the differential yield will compound. That's why it's critical that suitability tests reference the intention to draw down, not buy an annuity. To avoid circular reasoning, it will be important to demonstrate that the importance of the time exposed to risk, which in most cases requires extension into retirement, was demonstrated clearly, so that the rewards of not 'retiring the capital' when the member retires were part of the motive for the member stating an intention not to annuitise.

A corollary of the FCA's decision about the TVC calculation is that it should now be harder to justify transfers in general except when risk free rates are unnaturally low. This mathematical truism has never actually been used by the FCA (or FOS) to justify a bias against transfers and the new policy statement is no different. Though it has withdrawn its proposed removal of the bias - perhaps goaded by the Treasury Committee hearings into the situation that arose at the British Steel Pension Scheme - it has not chosen to relate that decision to the mathematics of the implied suitability reasoning when two present values, one using only risk free rates and one using partly risky rates, are compared. This is bizarre because it is clearly the anomaly created by Quantitative Easing that best explains, at a time when many pension schemes hold mainly, or only, risk free assets, the boom in transfers. It is not due to agency biases or to fundamental shortcomings in the old transfer advice regime, even if these have played some part. We think it is a great shame that this reasoning was never incorporated into the FCA's policy positions, either at the recommendations or decision stage. It loses an opportunity to treat the underlying maths as educational but it also leaves a bit of a question over whether the FCA even understands it itself.

If you think of the merit of transfers being a function of two essential conditions, firstly that the risk tolerance of the scheme and member are different and, secondly, that risky assets offer a generous risk premium over risk free rates, it is then obvious that the use of fairly static mean growth rates for risky assets, barely responsive to changing market conditions, as are the FCA's prescribed rates, is likely to obscure one of the two conditions. It has not mattered much recently because the risk premium is not high because equity prices are very low but rather because risk free yields are very low. But in any future scenario in which equities are depressed relative to risk free rates the FCA's rules will probably prevent the improvement in member welfare from transferring being exploited.

One way to avoid relying on normalised assumptions, instead of actual market conditions as interpreted by an adviser, was to use stochastic modelling, as this is what is more likely to introduce (inter alia) actual market conditions. The FCA probably fears this will cut across the standardisation of calculations that it sees as preventing 'gaming' to justify a recommendation. So it has reproduced an absurd position on the use of stochastic projections, that they are fine as long as the mean growth assumption in a stochastic distribution is not above the FCA-prescribed 'intermediate' projection. This was roundly drummed out by actuaries and insurance companies when it was first floated nearly 20 years ago as defeating the very object of stochastic modelling. Now the FCA has chosen to ignore these objections, the difference is that stochastic models are now more widely used in private wealth, such as by Fowler Drew, so this is a much more worrying decision. We will be taking this up with the FCA (having previously been asked to wait till the policy statement to make any representation beyond our consultation response).

There has always been a glaring hole in the both the FCA's original rules and the proposed changes. The success or failure of drawdown, of which a DB transfer is merely a special case, is largely a function of the rate of draw and this is arguably the most important information clients need from an adviser. Of course the investment policy and the rate of draw are intimately connected and must be internally consistent but the FCA's approach to assessing the investment policy ignores the client's intentions in respect of the level or time profile of the draw or the idiosyncratic constraints a client places on it, such as flexibility to adapt spending. The FCA might argue that it is avoiding being overly prescriptive but it is still curious gap when it is attempting to address suitability of the overall advice process.

The policy statement has addressed concern, such as from the Work and Pensions Select Committee, about contingent charges. This is being pushed into a new consultation, cp 18/07. Our views are clear and we will respond to the consultation. But an important connection here is with two other aspects of the policy statement. One is that advice not to transfer is just as much within the scope of regulated transfer advice as advice to transfer. This logic will raise eyebrows as it is radically novel in UK regulation, undoing a tradition of testing advice only in relation to the 'sale' of a product. Yet it is a contradiction that needed to be addressed. The other is that the practice (which we also adopt) of applying a 'triage' process is likely, in the FCA's view, to be regulated advice. Triage is usually a quick calculation of relative advantage based on a cash transfer quote and is intended to prevent a member paying the full cost of a transfer advice process when the chance of it being suitable is perhaps speculative. The FCA needs to understand the implications of effectively preventing such a phased process. We shall be asking the FCA whether our online calculator, which allows an anonymous and free test of relative advantage powered by a stylised version of our drawdown model, counts as non-regulated triage.

There are a number of smaller technical issues raised by the policy statement, such as in the area of changes in assumption rates that will mainly affect software providers, which we will return to, but I have highlighted here what I believe are the most important issues after a single reading (while conveniently trapped on a train) of both today's publications.

Taken as a whole, the impression given is that the FCA will not care too much what happens as long as it is a lower volume of transfers. It has probably taken fright at the recent goings on in Port Talbot and the dressing down it received from Frank Field's poorly-informed but very impassioned Select Committee. The FCA may not care about the perfect being the enemy of the good because it appears not to appreciate the good that can come from transfers as a direct and exceptional result of government's intervention in markets. This is a pity as we are sure that this is a process that can be appropriately regulated with the right technical procedures rooted in sound theory. Without clarification as to what the FCA believes the failures to be, we are all left wondering whether its findings from file checks about the suitability of DB transfer advice are technically well-founded or a reflection of a paternalist bias at odds with the intentions of Parliament.

#definedbenefitpension #Pensiontransfers #Regulation #CP1716 #PS1806 #CP1807