This blog first appeared in Professional Pensions, if you’ve read it before, read it again.
“What we observe is not nature itself, but nature exposed to our method of questioning.” W. Heisenberg, Physics and Philosophy: The Revolution in Modern Science
Prudence demands that we avoid monocultures, an absence of diversity, because of their risk of mono-causal catastrophic failure. With trustee boards, we seek diversity of experience, gender, ethnicity and qualification in order to avoid ‘groupthink’ and foster a multiplicity of views. However, we have only one highly prescriptive regulatory framework, and a requirement to comply.
Closely examining an elephant will result in disparate views, which may be difficult or impossible to reconcile without the benefit of some wider perspective. As David Bohm noted:
“One may indeed compare a theory to a particular point of view of some object. Each view gives an appearance of the object in some aspect. The whole object is not perceived in any one view but, rather it is grasped only implicitly as that single reality which is shown in all these views.”
Regulation imposes its own particular worldview, and does not cope well with indeterminacies, promoting a closed narrative. The fundamental theorem of asset pricing, the heart of the market-consistent paradigm, is a prime example of a closed narrative; it was conceived to link and unify many existing but unconnected theories. In doing so, it closed them. Theories deployed hegemonically, as when embedded in regulation, block or discourage other approaches and thereby the scrutiny of other aspects of reality. For pensions, alternate approaches to valuation, such as cash flow projection, are excluded.
To quote Edward Fullbrook**:
“Closed narrative communities typically live in open hostility toward ‘alien’ narratives. … Advocates of closed knowledge narratives often publicly embrace an extreme and primitive form of philosophical idealism, whereby they declare that their conceptual framework rather than offering a point of view on a domain, determines the extent of that domain.”
The zealotry exhibited by market-consistency advocates demonstrates exactly this.
“A knowledge narrative may become invert, meaning that instead of being used mainly as an instrument for explaining reality, its focus becomes itself. Turning away from the empirical phenomena that inspired it, it becomes transfixed with its own existence.”
The Regulator’s view of risk now dominates; it is far from natural. The collective fixation on scheme deficits, liability-driven investment and de-risking is one example; calls to scrap dividends and fund deficits are another.
A further caution is necessary:
“a social-science conceptual system can alter the objects of its enquiry by becoming part of the conceptual and belief apparatus through which humans define themselves, perceive others and make choices.”
In the sphere of pensions, closure to new members and future accrual lead directly to higher pension costs and ultimately cash flow deficiencies, reinforcing the view that DB pensions are unaffordable, as does the pursuit of self-sufficiency or buy-out. The ultimate example must be the over-priced cash equivalent transfers of recent times.
The case for a plurality of approaches is overwhelming. It is the presence of both competing and complementary narratives that fosters understanding, innovation and development. It is a necessary condition for the ongoing provision of occupational DB pensions, which regulation, with all of its prescriptions and ‘guidance’, has suffocated.
All of this said, the prospect for imminent change is limited. As Max Planck noted:
“A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it”.
Whether there will be any occupational DB pension schemes still open at that time is an open question.
-
* A narrative is a spoken or written account of connected events. This includes the theories, paradigms and concepts of financial analysis. Knowledge narratives are those commonly taught in universities.
- **Edward Fullbrook, Narrative Fixation in Economics, World Economic Association 2016. In this article, I have borrowed extensively and unashamedly from this work. I have read and marked my original copy so extensively that I bought a second copy.
I like Con’s analogy of ‘closed knowledge narratives’. But I wonder whether this really fits how a new theory of occupational scheme funding came to be established – against open hostility for quite a long time, before becoming enshrined in international accounting standards and then in the institutions that surround pensions, notably the Pensions Regulator and the PPF. Not all of the hostility was against the principles of mark-to-market measurement of the funding status. Many who accepted the accounting principles focused instead on the application of the principles, such as what trustees and regulators should do about the volatility in the resulting funding position. I remember the proponents of using FI discount rates arguing, when the implications were pointed out, that those were not their concern: it was someone else’s job to work out how or when volatile funding positions needed corrective action. It sounded disingenuous at the time but there is truth in what they argued. It was the Government and the Pensions Regulator who had to take responsibility for what the responses should be to interim accounting shortfalls, just as it had been previously with decisions about the consequences of a volatile MFR. It was their decisions about the practical consequences that created such a strong motive for limiting or even avoiding altogether volatility in the accounting funding position.
For the same reason, I think is is unfair and unhelpful to blame LDI as a concept. It was never realistic, pre-LDI, that scheme actuaries used to carry out a detailed analysis of the nature and duration of the liabilities (for a fee) and then adopted some version of conventional ‘balanced management’. It was too inconvenient to join up the A and the L in ALM. LDI challenged that laziness. LDI has very general merit (and not just in pension funds) by applying the principles of ‘portfolio separation’, as between hedges and bets, to ensure a much more robust risk-control framework than balanced management ever provided. It forces you the think about the nature of the liabilities in order to define the hedges. It also encourages investors to think about the aspects of the risky portfolio that best match their utility, which might be minimising variance in the funding status because of those consequences for contributions but could also be minimising the long-term cost of funding. LDI does not tell you to how much risk to bear but helps you get to the answer.
In my area, wealth management, applying LDI has been transformative in helping clients identify their ‘true’ utility, make fully-costed choices between hedges and bets (so risk tolerance is no longer an abstract), identify the right hedges (rarely, btw, nominal bonds) and make all of these choices dependent on the time horizon (or horizons, as the objective is often a stream of cash flows). Yet the scope for LDI to transform investment planning and management in private wealth has barely been tapped so far. I would hate to see it being sullied by association with some bad choices about how to deal with its consequences in occupational pension schemes, when the blame for those choices lies elsewhere.
These are minor pleas. In general I agree with Con and admire his ability not to be fooled by the ’emperor’s new clothes’.
Stuart
There was indeed considerable resistance to the change of accounting paradigm – that is to be expected with any change like that. I opposed the shift to mark to market accounting in the late 1990s -my objections though were far more general than just DB pensions. As I see it, when a company issues a fixed term bond at a fixed rate, that is what it should account for, not some “value” to some undefined other which varies with the level of dopamine. It turns out that pensions are not that much different, as I have shown with my notes on liability valuation.
The shift in the actuarial profession actually took place after the accountants. I remember speaking against the method at an NAPF conference in 2004. Interestingly, and tellingly, the accountants and actuaries present did not respond to my arguments.
In most fields of scientific endeavour, a mountainous and mounting body of evidence falsifying the hypothesis would lead to it being discarded or at least modified in ways that reflect those paradoxes and ‘puzzles’. In another article, I noted: Elsewhere than economics, such a profusion would usually be taken as strong evidence of a wrong model.
Don’t misunderstand me, I am not against hedging per se; there are places where it is appropriate. The family office I advise has just borrowed sufficient to fully offset the value of its US assets (from a US banking syndicate) – their US operations now have the characteristic of an equity, a call option on the upside. LDI though is a different kettle of fish.
Interest rates are not a determinant of pension benefits – they do not appear in the rules which create them. What rate that there is, is contractual and deterministic – see my article on liability valuation. The valuation standard is mixed attribute – discounted present values on one side and market values of the other – and that will inevitably lead to bias and volatility. It is interesting that these pro-market people accept, indeed insist, that we cannot derive a discount rate from equity prices. In this they are correct; the problem is underdetermined. But then, they use market interest rates which is just another presentation of a price. The problem is still underdetermined. We do not know how much of this rate arises from fundamentals associated with the future utility of the cash flows and how much to the degree of risk aversion of the market participants.
I have not thought in general about LDI in wealth management – my experience is limited to advising just one family office. It is an accepted method in insurance company management that predates my involvement in that industry. However, I will say that to manage on that the basis that A=L is extremely restrictive; it entirely removes any prospect of profit. That of course may be immaterial if the objective is wealth preservation, though I doubt that wealth may be preserved sustainably without the pursuit of profit. When time permits, I will write a more extensive critique of LDI and de-risking. Though the idea of separate sub-portfolios, such as liability-matching and growth assets, is widespread in practice, I am far from convinced that such separation is systematically feasible. Obviously, any arbitrary split is possible.
There is a longer version of the Professional Pensions article which covers some more of the problems (it will appear in the CFA(UK) journal), but even it does not cover the role of the pension regulator in recruiting a congregation for their view. It has though been substantial. There is a meme circulating, largely unchallenged, that pension funds should be self-sufficient. The most recent ILC report expresses it in several ways: “The current TPR remit for pension scheme trustees is to ensure members receive full benefits.” And “to secure full benefits for members.”, and “members …have their entitled benefits protected.” In fact the Regulator’s objective is: “to protect the benefits under occupational pension schemes of, or in respect of, members of such schemes”. That is a very different thing indeed.
This has been promoted by the Pension Regulator. It follows from their secondary objective: “to reduce the risk of situations arising which may lead to compensation being payable from the Pension Protection Fund”. It is not in direct conflict with the primary objective. It is also correct inasmuch as the pricing of new business by an insurer (or mutual compensation fund) is correctly based upon what they can expect at that time to earn on the assets available for purchase. However, when we see extremely low expected returns this is expensive – to which I will return later. It has led directly to scheme closure, but perversely encouragement of provision is outside of the remit of the Regulator.
Let me clarify the position with an illustration. If I buy a fixed rate, fixed term corporate bond, then in a liquidation (not market sale) prior to its maturity I expect to receive the sum originally advanced plus any unpaid coupon; in other words, performance of the original terms of the contract. This is sufficient to secure the future benefits under the original terms. It may or may not be sufficient to secure them under the prevailing market conditions. This is the position which would be enjoyed by a DC member.
There is no reason for DB contracts to be different. Indeed, there are a number of references within the legislation which support the idea that parity with other creditors, albeit secured by the assets of the fund, were an objective. The contractual rate that I show in my earlier articles is the correct discount rate. The liability valuation would in liquidation be sufficient to secure the benefits promised under the terms of that contract. This may or may not be sufficient to purchase assets of sufficient expected return in the market. However, this is largely moot as there is the Pension Protection Fund.
Rather than setting levies as they currently do, the PPF should be looking at the implied terms of the scheme and setting levies based upon this implicit contractual cost. A scheme with a 4% contractual rate would face a much lower levy than one with, say, 8%.
DB schemes with their risk-sharing and risk-pooling are vastly more efficient than individual DC and even the purchase of pooled investment funds does little to alter this. It costs about twice as much to deliver a pound of pension under DC as under DB. The amazing thing about the current vogue for claiming CETs is that in the judgement and opinion of rather a lot of people, some very smart indeed, the terms of DC are now more attractive than sticking with DB.
This reply is now about twice as long as the original article, so I will end it here. Incidentally, with Iain Clacher from Leeds, I will be pulling all these articles and some other work together into a single document. Finally let me thank you for the feedback – it is much appreciated.
you’re right Con, the reply is longer than the blog- but none the worse for that, thanks to both or you for improving my understanding – i really don’t think I’ll bother becoming an actuary after all!