Why some transfer values may be just TOO HIGH!

over priced


Earlier this year I wrote a blog “why some transfers are ridiculously high” which set out to explain how DB transfer values are calculated and why they are at present at historical highs.

My friend and colleague Alan Smith presented the actuarial version of that blog at the Great Pension Debate. I don’t think we distributed the slides, but if you want to see the technical arguments for high transfers, please read , download and share these slides, they are meant to help – not to sell us as actuaries!

The link to the slides is here

I won’t be delivering these slides to an audience at the Pensions Network tomorrow, but I will be referring to them.

If you are going to the event in Tring, and have questions about this blog- please ask them at the session on Friday Sept 8th (tomorrow as I write)


Transfer values are a regular topic of breakfast discussion in our household.

I live with someone who manages a DB scheme with assets north of £45bn and with outflows from CETVs of some £250m a month. She reports that there are schemes in her peer group with even higher levels of exodus.

Her feeling and that of many senior pension professionals is not just that pension transfers are high – but that they may be “too high”.

The fundamental objection to the current means of calculating CETVs

One argument, an argument that Con Keating articulates, is that there is an unstated but implicit internal rate of return on which pension promises are made at the outset of someone’s joining a DB plan. This is not “time- variant” – it does not vary with time.

If you accept this argument of an internal rate of return, Con calls it the accrual rate, then all transfers would be calculated using this as the discount rate and transfer values would be much lower than they are today, reflecting the longer term assumptions underpinning the scheme, not the short-term best estimates discount rate – based on the scheme’s actual asset allocation a the time of transfer.

The pragmatic arguments against the current means of calculating CETVs

Another argument, based on a less fundamental approach to the problem, is that CETVs are unfairly rewarding those transferring with the prudence with which the trustees are managing the assets. Put in easier words, people are getting fair shares plus.

The highest transfers, those which give cash sums of 40+ times the pension being forsaken, are so high- typically because the remaining assets within the scheme have been selected to maximise the security to existing members – it could be argued that schemes that are invested almost entirely in bonds are “self-sufficient” and don’t need any future contributions from the sponsoring employer. Indeed most of such schemes want to sell themselves to insurance companies who will guarantee to pay the pensions.

Not only do members walk away with the “prudence” in the funding, they also get a CETV based on 100% of the benefits being paid as pension. In practice – most people take around 75% of the benefit as pension and the rest as tax-free-cash. As we all know, tax-free-cash commutation figures are set in favour of the scheme. Every time a CETV is paid out on the basis of 100% pension being paid, the scheme bleeds the cash-commutation windfall.

Transfer values from such schemes are close to the buy-out cost of the scheme. They don’t offer much of a discount to the employer to the book-cost of the liabilities in the company accounts (under IAS 19) and they offer individuals a right to something well in excess of the original promise – the transfer includes the price of the guarantee of the insurer (or a good part of it).

Why does this matter?

It could be argued that these super high CETVs are windfalls for those lucky enough to have them – rather like the huge bungs paid out when mutual demutualised. This argument simply says that some people get lucky.

But when some people get lucky, the premium they receive is paid by others. In the case of defined benefit schemes, the high transfer values are paid for at the expense of other stakeholders of the scheme.

One of those stakeholders may be the sponsor, who is relying on an investment strategy that benefits from a carefully planned cash-flow strategy that may have to be unwound to meet exceptional early payments of CETV. If you are paying out £3bn a year in CETVs against a fund of £45bn , you have to do a bit of rejigging, re-jigging does not come cheap. The cost of changing an investment strategy is born in the future funding rate of the scheme- that cost is either born solely by the employer or shared by those members still accruing. Either way there is a cost.

And if the point of moving the scheme into low-risk assets, was to immunise the sponsor from future cash calls, the impact of CETVs on the sponsor will be most unwelcome. The sponsor can rightly say they have paid once for prudence, why pay again to regain prudence that has been un-necessarily paid out in the CETVs. In extremis, the sponsor may walk away from the problem and that puts members in jeopardy of a scheme going into the PPF.

Is there a solution?

It’s not often that actuaries are challenged, least of all pension actuaries. But I am hearing stories of expert trustees challenging the payment of high transfer values on the grounds laid out above.

Of course, there are all kinds of reasons that employers like people taking transfers, but for trustees, they are not good reasons. Trustees want pensions to be paid and paid in full. They do not run pension schemes as a kind of launch pad for the wealth managers.

Nor do they run their pension schemes for the benefit of finance directors of sponsors who can write large chunks of pension liabilities from the sponsor’s balance sheets on the basis of what has happened in the past.

The Trustee solution may be as simple as challenging the basis of CETV transfer and going back to the first principles  of how the scheme was set up. It is unlikely that they will adopt as radical a discount rate as Con Keating’s IRR or “accrual rate” but it is quite likely that they will argue that CETVs should not include the full prudence within the scheme funding and should reflect the commutation of tax-free cash.

If that challenge happens, and is successful, then the current transfer bonanza may be drawing to a close.

However – I do not suppose that this will happen soon and this is not a call for IFAs to put out the “transfer now while CETVs last” sign.


One final point – and it’s an important point. There is a much simpler way for trustees to reduce CETVs and that is to demand an insufficiency report which tells members that the Transfer Value has been reduced because the scheme is in deficit and the trustees need to protect the scheme as a whole. Those who remember market level adjusters in with-profits will see the analogy.

Trustees are understandably nervous about telling members they are nervous. It tends to cause a run on the scheme and give all the wrong signals to the Pensions Regulator.

Many people have commented that there are too few insufficiency reports out there. There are very few employers who make those arguments, it hardly does your credit rating good – if your trustees are telling all and sundry you may not afford the pension scheme.

It may be time for insufficiency reports to be reviewed by the Regulator, they aren’t doing what they are supposed to be doing and we may need something better.


I find myself once again having to talk about something which is sensitive and I’m quite sure that a lot of actuaries, including some of my colleagues, who will be asking why a non-actuary is asking questions about the way they do things.

Once again I will point such people to the rubric at the top right hand corner of this blog. I am  speaking as a non-expert but as someone who cares about pensions.

About henry tapper

Founder of the Pension PlayPen,, partner of Stella, father of Olly . I am the Pension Plowman
This entry was posted in pensions and tagged , , , , . Bookmark the permalink.

4 Responses to Why some transfer values may be just TOO HIGH!

  1. Con Keating says:

    Some of the really high CETVs are motivated by the regulatory treatment of liabilities – for financial institutions the required regulatory is yet another cost of the scheme.
    The people who really should be worried by the bulk annuity players – if companies can shed liabilities must more cheaply by this route than by buy-out, their business model is clearly challenged.

  2. Bob Compton says:


    Your article is spot on. Trustees should be challenging their Scheme Actuary, if the CETV calculation is anything better than best estimate. It should be an Employer decision to enhance TV’s to higher than Best Estimate, as this may be a way as Con infers of the Employer speeding up the reduction of liability shown in the Employer accounts, by making the transfer option more attractive. However it is rare for Trustees to have access to independent advice as often the Scheme Actuary and Trustee Consultant, not only work for the same organisation, but are the same person! This then means the Employer has to bring in outside expertise, which then tends to raise the hackles of the trustee advisers, and this can end up in becoming an expensive “time cost” bean feast for the advisers at the ultimate expense of the Employer.

    But at the end of the day, it is the Trustees decision, and if they are genuinely fulfilling their function, they should be asking questions of their scheme actuary, and most certainly should not be overpaying TV’s without the explicit support of the Principle Employer. l can foresee members suing trustees in future when schemes fail, where TV’s have been overpaid, when the assets backing the scheme were not sufficient. How long before the Regulator realises this is a real issue, and if they are fulfilling their statutory function of protecting the PPF, they should be taking the issue seriously, otherwise they too will be seen by a future Frank Field inspired enquiry of failing in their duty……..

  3. John Moret says:

    Excellent article Henry on a very topical and challenging subject. Thanks too for the mention of The Pensions Network(TPNW) – I’m looking forward to chairing the transfer panel debate – with yourself, Romi Savova (PensionBee), Michelle Cracknell (TPAS) and Justine Pattullo (Origo) it should be a cracking discussion. We’ve also got John Greenwood, Steve Bee, Susan Martin (Local Pensions Partnership), Dr Iain Clacher (Leeds Uni) and Elisabeth Costa (Behavioural Insights) as speakers so it should be a fantastic meeting. If you’re reading this comment and want to find out more about TPNW go to http://www.the-pensions-net-work.com It’s my privilege to chair this meeting & discussion forum for pensions professionals.

  4. I read this looking for a strong case for a significant issue of inequity between leavers and retainers. I just don’t see it. There are, as you point out, Henry, difficulties identifying the right technical measures to ensure equity but it would be a step to far to ban or impede transfers (as some would like) just because it was impossible to ensure perfect calculations.
    The observation about the significance of the discount rate differences arising from the scheme’s asset mix is important not just to what makes the income multiple a high (not ‘generous’) one. The asset mix is also critical to the calculation of relative utility. The transfer boom is only increasing most leavers’ utility if, given their own tolerance of poor outcomes (in the form of sustainable income) and the value they assign to better outcomes (whether for spending or bequest), they would anyway hold an asset mix different from the scheme asset mix. This cannot be tested just by looking at what assets they hold to complement safeguarded income because those assets will need to change as long as the overall plan is subject to unchanging outcome tolerances.
    Practical evidence from our own experience may be useful to illustrate the point. We run stochastic models for clients’ assets assigned to retirement in which the risk attitude is constant but the risk level varies as a function of the horizon (because of setting horizon-specific tolerance constraints on the probable outcomes). We rarely found in the past we could increase utility by that combination of a CETV and a lower proportion of risk assets, compared with retaining the DB pension and building on its foundation (unless other factors dominated the transfer logic). It was only as a consequence of both further scheme derisking and further falls in ILG yields that we found ourselves, with unchanged models, having to say to the same clients, hang on, we now need to alter our advice.
    It’s possible that our model points to greater utility gains because our replacement strategy is not the typical broadly-static asset mix matched to a constant risk level (equivalent to the old ‘balanced’ model schemes once relied on). It’s hard to justify a transfer when the DB underpin is simply replicated, but poorly, in the replacement strategy by a permanent holding of nominal bonds, possibly of long duration, with lots of inflation risk. It’s easier to justify when the CETV, like the complementary resources, will be assigned to a combination of what is really an uninsured temporary annuity for near-term liabilities, with little or no inflation risk, and risk assets for longer liabilities.
    Even allowing for weaknesses in the commonest approaches to modelling or quantifying projected replacement income assuming drawdown (which CP17/16 does not fully address), I am convinced that the utility gains from transferring out of derisked schemes at typical individual risk tolerances are of a different order of magnitude to the possible costs to pension schemes or to remaining members.
    But it’s probably only a temporary situation anyway. The pitch really is ‘while stocks last’. The transfer window will close when sanity is restored to financial markets and we’ll go back to only doing them when there are other factors that increase utility. There are more such factors with pension freedoms, but not enough to explain a high level of transfers.

Leave a Reply