Imagine buying a bond at par, say a ten-year, ten-percent coupon issue; then let five years pass, when market interest rates decline to one percent. At this time, the bond is trading in the market at £143.68 percent. Now, rather than selling in the market, you instead present it to the issuing company for redemption at this value. This is the fundamental position when taking a Cash Equivalent Transfer Value (CETV) from a DB pension scheme.
This possibility of such redemptions presents several problems. For example, management of the cash flow under exercise of the option, as this is larger and occurs sooner than was initially contracted. It is an intriguing variant on the bank-run. The cost of this funding, which was contracted as 10% p.a. for ten years, has risen to 16.31% and reduced to just five years. The option has proved extremely expensive to the corporate issuer or equivalently the scheme.
The exercise of this option will lower the credit standing of the issuer. If that is difficult to envisage, just think of retiring another liability, equity, where it is more directly evident. The extent of the credit impairment is due the excess cost paid over book value, which in this case would be £100. This is also the value of a creditor’s claim in insolvency and solvent liquidation. Moreover, the range of assets which may, prudently, be purchased by the corporate treasurer is more restricted than might otherwise be the case without this option.
Defined benefit pensions contain an implicit investment return, defined by the initial contribution and the projected benefits payable, the contractual terms of the pension. For most DB schemes, by virtue of their historic stock of awards, this inherent investment return lies within the range 6% – 8%. Valuing these liability cash-flows using lower discount rates will throw up excess costs in the same manner as the previous bond illustration.
The guidance on cash equivalent transfers makes it clear that the basis for an initial cash equivalent (ICE) appraisal is a best estimate valuation, not the prudent assumption basis used in the estimation of technical provisions requirements. This removes the prudent conservatism which overstates liabilities in the assumptions about longevity, inflation and other risk factors. The discount rate rises, and the present value of liabilities falls.
If the level of scheme funding is below this best estimate, then, after an actuarial appraisal, the CETV may be lowered further to reflect the member’s proportionate share of those assets.
At first sight then, this arrangement appears to be fair between the departing member(s) and those remaining, but that would only be the case if the discount rate used in the best estimate valuation was equal to the implicit contractual investment return. When this discount rate is lower, we have the equivalent situation to the earlier bond illustration, and the departing member’s withdrawal has a direct cost to the remaining members, which include the sponsor.
The effect of a departure also has a deleterious effect on the risk sharing and pooling properties of the fund, though this is usually small for any single member. However, it is possible for higher paid members to game the presence of the Pension Protection Fund. With a company in distress, these members may take CETVs even though asset values are below the ICE value, because their pensions would be capped under the PPF. Even if the trustee had the power to refuse CETVs, it would not be possible for them to do so on the grounds of harm to remaining members, since by this “best estimate” methodology there is none evident.
Notwithstanding the guidance on valuation, there appear to be many schemes which are offering transfer values based on gilt yields and extremely conservative inflation and longevity assumptions, making for very generous transfer values. This raises the much more interesting questions: how and why did we get here? But that will have to wait for another blog.
That makes perfect sense and poses an imponderable question. Thanks for the clarity Con.
Just how did we get into this position?
Con, as I pointed out in a comment questioning Henry’s mooted moral dimension of transfers, there must be many cases in which the deferred pension liability is hedged and the discount rate to apply is clear and consistent with the hedging asset. If hedged, we would expect a scheme to be able to identify very easily the price at which it can be indifferent between retaining the liability and removing it from their books altogether. The difference in prudence of assumptions you refer to would not apply and neither would the option you described (the transfer liquidates the bond at market value).
By contrast, where there is (as far as we can tell) less hedging, when we run the numbers through our modelling we note the CETV terms are not always creating economic value for clients with typical risk tolerance (allowing for the downward adjustment to existing tolerance implicit in the pre-transfer plan/portfolio asset allocation when replacing a risk free asset effectively already held, assuming an unchanged constraint in respect of minimum outcomes).
We therefore deduced from the partial sample we handle that the fundamental source of economic value is not gaming but a difference in utility between the scheme and the member, which includes not being motivated to hedge all liabilities, or (if you like) treating retirement income as aspirational rather than guaranteed, much as DB schemes once did.
My question to Henry was how inconsistent with this theoretical explanation is actual current practice? I can’t tell because my firm only works on the other side, as the transferring member’s agent. Is yours another theoretical observation, like mine, not rooted in coal-face experience? If so, perhaps Henry’s readers include actuaries who are doing the pricing. Anyone care to share?
Stuart
The property rights of the member are defined at the point of award and payment of the contribution. This means that the contribution and the projection of pension benefit fully define the investment return promised to the member. How these obligations are funded is absolutely immaterial. Interest rates do not expressly figure in the terms of a pension. Matching assets to liabilities and using a risk measure defined on sensitivities to interest rates is nonsensical in this context. The only valid discount rate is that investment rate. This means that the transfer value would be determined by that rate. In the case of the bond, it would be £100, not the £143.68 which is based upon future performance. At this £100 price the member is getting exactly what was contracted for, up to that point in time. If the scheme pays any value in excess of this, it is overpaying to the detriment of either the sponsor or other pension beneficiaries.
Hedging the liabilities with assets does not introduce any justification for paying more than the original best estimate pension promise. It seems to me that the only reason for a scheme and sponsor to offer the high CETV s that we see is that they want shot of the scheme and anything which is cheaper than the full buy-out cost is acceptable to the sponsor. It is debatable whether it is to the remaining pension beneficiaries.
I am reliably told that very few insufficiency appraisals take place, but that is not surprising given that the Regulator’s position: If you believe the employer covenant is strong, then you think stayers will have all benefits paid in full so you shouldn’t be reducing the benefits of leavers. That brings to mind Alice, or rather Twumbledee – Contrariwise, if it was so, it might be; and if it were so, it would be; but as it isn’t, it ain’t. That’s logic.
I will reiterate the position. The contracted internal rate of return of the scheme delivers the correct value for liabilities. The way in which these are funded is at the discretion of the scheme trustee and sponsor. The level of funding needed is then determined by the expected returns on the asset mix chosen – for example FABI or gilts.
Make no mistake though, paying the higher transfer value, the £143.68 crystalizes the £43.68 as an additional cost over and above the 10% promised.
You will tend to see higher CETVs in the cases where schemes have bought gilts rather than equities, as there are more of them, for the same funding ratio. If the gilts yield 2% and the equities 4%, twice as many in a perpetually open scheme.
Given the risk pooling and risk transfer properties of the DB structure, which save on the cost of production of the pension benefits, the idea that a transfer value can buy equivalent benefits in the market elsewhere is more than a bit challenging.
The basic idea of hedging everything in some matching manner is misconceived. LDI and de-risking can only be justified by the desire to avoid the consequences of the misconceived accounting and regulatory regimes. It is encouraged by the Pensions Regulator and PPF, not because it is appropriate or correct in any financial sense, but because it reduces their risk exposure. The perfect and limit hedge would be to hold cash equivalent to the total benefits projected, and then the transfer value would be simple, full and immediate payment of all the pensions, but where does that leave them as deferred pay? The cost to the sponsor would be rather high, and it would not even create any investment from which the economy might benefit.
I only work with one scheme here and I have checked but that has only ever had ill health transfers out, and we treat those very generously indeed, so I can’t answer the coalface question. Maybe some of the others can help on that.
I am not an actuary but work with actuaries. The view among actuaries I talk with is that the perfect hedging of liabilities is simply the cash needed to purchase the annuity. As the CETV is going to be less than this, you could say the person transferring is doing the scheme a favour even when taking a CETV at a ridiculously high multiple (40+).
But this is to make a nonsense of investment as your clients know very well. The purpose of a collective investment scheme is to achieve defined outcomes more efficiently than an individual could do. So the purpose of the transfer must be to reshape the benefit in some way or other (unless the heroic individual thinks he can beat all the odds.
That we are now arguing that it’s a win win for scheme and member that members go it alone, suggests the CETC is inappropriately high (I accept I am paraphrasing Con),
I don’t see why people shouldn’t pick up £50 notes lying on the floor but I would rather that trustees did not empty their wallets in the first place. Bank notes make expensive litter and this is money that employers have paid to plug deficits – they might have used the money more productively than to provide golden goodbyes to their pension scheme members,
Stuart
It appears that I was not as clear in my two pieces as I had hoped to be. So:
The discount rate used in the best estimate valuation for transfers (ICE) is the expected return on the asset portfolio held. So, if you hold gilts, you will have gilt yield as the discount rate, and if you hold equities, you have equities. This explains what you observe in the pricing of quotations. This complies with the Regulator’s position.
This really is Alice in Wonderland stuff. Think of my limit example of the scheme which just holds cash.
My position is that both of these are in fact wrong and that it is the contractual investment/discount rate which is correct, though currently not used by anyone to my knowledge.
Henry’s point that actuaries think of buy-out as perfect hedging – they may and do indeed do that. But if the insurer goes bust before the scheme has wound up, then the trustees can have members back knocking on their door if the FSCS does not fully meet their claims.
Con