Peer review is the evaluation of work by one or more people of similar competence to the producers of the work. It constitutes a form of self-regulation by qualified members of a profession within the relevant field. Peer review methods are employed to maintain standards of quality, improve performance, and provide credibility (wiki).
In Britain, as in the US, peer review also happens in politics. We have a house of Lords and a house of Commons who review each others work. It is the same in the States with the House of Representatives and the Senate.
I am not going to get dragged into an argument about USS, but I am prepared to stand up for politics. Our political system cannot be dismissed so easily. Academic peer review is not superior to political review because academics have carried it out.
There is no divine right of academics.
The political system in this country is designed to apply common sense to debate. Consequently Government, whether in Westminster, Brighton (tPR), Stratford (FCA) or in any of the local authorities, right down to the parishes, is the application of common sense to debate.
Common sense cannot be dismissed by academics.
Academic peer review is important, but it can be very introspective. If you note the people liking John Kiff’s tweet, they are like-minded scholars.
It is very like academics to use their networks to sign joint letters and to extrapolate this inter-connectivity to suggest that they represent global authority. So the 39 academics who are co-signatories of “our letter” are supposed to represent worldwide support for a particular idea. But we know that the internet, and especially the social end of the internet, can bring people together in small clubs very easily. These clubs – groups – pages – can easily become their own echo chamber.
What appears to have happened with the club of 39 is that they have bestowed on themselves false authority, through the mechanism of a joint letter.
I too have signed a joint letter recently. I signed a letter calling for a resolution of the “net pay anomaly”. I did not sign that letter as someone authoritative, I don’t want to promote myself as expert in an academic or any other way. I simply wanted to make my and my company’s position clear on a matter of gross inequality.
We submitted the letter , not to a pension scheme, but to a Government Office – the Treasury. It was a petition on behalf of a certain group of people whose interests we think are not being properly promoted.
The distinction is quite clear. Instead of by-passing and then dismissing politics, we appealed directly to Government. I hope that our appeal is successful but I am prepared to accept the ruling of Government.
Those academics who dismiss the law, are promoting their ideas as above the law. That is always a dangerous thing to do. The law says that schemes can adopt an “expected rate of return” funding basis and some choose to do so. That law comes from parliament and was subject to peer review – by a House of Peers.
Academics (and financial economists do) criticise schemes for taking unnecessary risk, but they have no special privileges to be right. There are other academics that support a best estimate funding approach and they too have no right to be right.
The rule of law is everything, without it – we would descend into anarchy. From time to time over Britain’s political history, certain groups have claimed special authority. Financial Economists seem to be claiming special authority right now.
But they have no special authority, they are petitioners – like everyone else – in our common weal. The rights of the common man are upheld and enforced by our political system and they cannot be undermined by any special interest group.
The expected return on assets approach was in fact recommended to the Goode committee by the Institute of Actuaries and then appeared in the 1995 Pensions Act – which should put an end to this peer review nonsense.
Interesting. Could you please send a link to the work that underlied that Institute of Actuaries recommendation?
I’m hoping that today is the day the Chancellor uses his Budget to address the net pay anomaly which was the subject of the joint letter you recently signed
Adrian
I never said that “peers” had to be academics. In fact for the application of an actuarial concept like the liability discount rate, I would hope to find analysis done by professional actuaries and reviewed by other actuaries. My repeated requests for such actuarial analysis that supports the use of expected rates of return in that context have never been answered.
Regarding the actuarial analysis you request, John, see ‘Determining Discount Rates Required to Fund Defined Benefit Plans’ (Society of Actuaries’ Pension Section Research Committee). Link: https://t.co/cOmdZwtq3B
In this tweet and the two below, I discuss some of the passages in this paper: https://twitter.com/MikeOtsuka/status/1028251207566016512
Thanks Mike. I don’t know what happened with the notification on this blog, but I just got the notification today! Anyways, I’ll add that to my weekend reading material. I had hoped that “weekend” would have been weeks ago, but my day job keeps swallowing up my weekends. Tomorrow starts a long one here (Veteran’s Day) but crypto-assets and fintech have killed this one too!
One should not push the idea of peer review too far. It is just a check that the work is a serious piece of research that uses rigorous methods. It does not follow that because it has been peer reviewed it is unassailable truth. Academia is full of schools of thought that publish different journals with their own norms of what they will accept and not accept. There are journals that take it as axiomatic that assets prices follow a lognormal brownian motion stochastic process and will not publish work that questions that, however sound the research. There are other journals that take a contrary position and are happy to publish work that critiques the tenets of FE. Another factor that is quite important is that academic journals have a bias in favour of positive contributions and against those that are critical. Thus it is easier to get work published if it is an extension of an existing model than if it is attacking the foundations of the model.
But as regards the actuaries, who are a profession as well as an academic specialism, the former have had to endure a lot of criticism. They were told (eg by Sir Derek Morris) they had to embrace the exciting new techniques from financial economics that made it possible to understand and manage risk better than what they were doing. The fact that, for all their peer review, “Nobel” prizes and all, these techniques were not soundly based got lost.
I agree, one shouldn’t hang too much on peer review. However, it does serve as a sort of quality check, and, from my own experience, generally improves the work. Maybe i should rephrase my request to ask for a paper published in a recognized actuarial journal that builds the case for linking liability discount rates to assumed rates of return. If you find the concept of peer review objectionable, I’d settle for a paper that has been informally published that builds that case.
BTW I have no problem with politicians overriding expert analysis. Perhaps they have overarching policy goals such as encouraging pension schemes to provide long-term risk capital to the economy? Although Dennis Leech thought he was being very “smart” by dissing my employer’s policy positions, in fact, for all I know, the IMF might support such an overarching goal. (As was clearly stated in the “letter” my views don’t reflect on my employer’s.) But at least that overriding of the analysis that should have been done by actuaries and other learned professionals should be clearly acknowledged.
There is plenty of peer-reviewed published academic research showing the damage done to peoples lives by the neoliberal political economy that is behind the Washington Consensus of the IMF as much as the extreme market fundamentalism behind financial economics.
You’re off topic again. This has nothing to do with the IMF and I might even agree with your comments in that regard. All I’m asking for is a reasoned argument from the actuarial community for the use of expected rates of return to discount DB pension scheme liabilities. Is that too much to ask?
It’s kinda “funny” how, whenever Inask for any substantiation, everyone goes quiet… I understand that political considerations sometimes trump logic, but the substantive arguments should still be put forward. And there have been no substantive rebuttals of the logic expressed in the infamous “letter” – just ad hominem attacks on the signatories and demands that we butt out. Very sad…
There really is a lot of (peer reviewed) published economic thought that rebuts the logic behind the letter. For example both theoretical and empirical research published over thirty years ago in the American Economic Review (the most widely read and respected academic journal in the subject) has shown the weakness of the Efficient Markets Hypothesis, which is the foundation stone of the approach.
I don’t think you have to rely on the EMH to show how absurd it is that increasing the assumed rate of return on pension scheme assets will decrease the discounted present value of scheme liabilities. It’s not financial economics, it’s common sense. If you want to go to extremes, you can reduce the present value of liabilities to near zero by increasing the assumed asset return to infinity in your framework! That’s not financial economics, just basic math!
Indeed what you have just written is commons sense. But you have not said how you choose your discount rate.
Where EMH comes in is in the use of mark-to-market valuation of assets. That in fact introduces risk.
I’ll see what Derek’s “truth” paper has to say about using assumed rates of return. But market values may not be perfect ex-ante reflections of ultimate value, but I’ve been led to believe (or maybe indoctrinated!?) that they’re the best we have. Although many flaws in EMH have been exposed (in the peer-reviewed FE literature I might add) where’s the hard evidence that sponsors and their actuaries can consistently beat the market? That to me would be a pre-condition for allowing scheme sponsors and their trustees to effectively substitute their own asset valuations for market ones.
Reading for you, John:
http://www.simoncarne.com/truth.pdf
Hi Derek
Happened upon this post by accident but this paper is very insightful.
It’s somewhat bizarre reading it post-FRS 17 in a world where discounting seems to only be based on atrocious returns but there you go. I guess no-one foresaw the credit crunch at that time.
DC
“how absurd it is that increasing the assumed rate of return on pension scheme assets will decrease the discounted present value of scheme liabilities.”
Why is it absurd? It’s just arithmetic.
The liabilities of a pension scheme are the benefit promises.
The present value of the liabilities of a pension scheme is a number written down on a piece of paper having decided what discount rate to use to discount the expected benefit payments. A smaller discount rate gives a bigger number to write down on the piece of paper and vice versa.
What discount rate to choose depends on lots of things: your objectives for the pension scheme, the objectives while the employer is solvent and the objectives in the event of employer insolvency, what you intend to use the calculation for.
Meanwhile, it is possible in principle to manage a pension scheme by studying its expected cash flows in and out without doing any discounting.
Credit where its due, the author of the paper I gave the link to is Simon Carne.
John, is it fair to clarify your comments as follows:
But BOND market values may not be perfect ex-ante reflections of ultimate value, but I’ve been led to believe (or maybe indoctrinated!?) that they’re the best we have. Although many flaws in EMH have been exposed (in the peer-reviewed FE literature I might add) where’s the hard evidence that sponsors and their actuaries can consistently beat the BOND market? That to me would be a pre-condition for allowing scheme sponsors and their trustees to effectively substitute their own asset valuations for BOND market ones.
No I meant markets in general – the ones in which typical pension scheme assets trade. So that includes equities and bonds. I believe that was what Dennis was referring to in his EMH comment. Also, maybe I might learn more from Simon’s paper, but I’ve assumed that the idea behind using expected rates of return (rather than risk free rates) to discount liabilities is to “correct” for perceived inaccuracies in market prices of asset holdings. That’s the only rational rationale for using expected rates of return that comes to mind right now.
To derive an expected rate of return on an asset, I would project the income I expect from the asset and find the rate of return which valued the income at the asset’s market value. This method uses market data to derive the expected return. It is applicable to any market. Indeed, a gilt yield is exactly this.
We’re not in any disagreement, John?
No disagreement with the math. What I need to think further on is how this rate of return, when applied as a discount rate on liabilities, weaves into the overall economics. Given the pressures of my “day job” that may have to wait until the weekend.
I look forward to your thoughts after you’ve had a chance to read Simon Carne’s paper. Meanwhile, I too should earn my living.
So not this weekend as my day job spilled over and into it. If you glance at my other Twitter activity you’ll see that it’s very much focused on crypto-assets and more broadly fintech, which is a big obsession here at the IMF these days. Hopefully next weekend.
But just one last thing on liability discount rates. You may have seen me occasionally referencing a paper I co-wrote on U.S. DB pension plan sponsor actuarial assumptions (see link below). We find that when you give sponsors wiggle room to set their own discount rates, they tend to take it and run with it, and particularly cash-strapped sponsors.
So besides any “financial economics” rationales for my “absurd” comment above, I have some incentive issues with giving sponsors that wiggle room.
I would prefer to move away from packing so much into the liability discount rate (sponsor creditworthiness, risk of asset portfolio, degree of de-risking, pension scheme objectives, the objectives while the employer is solvent and the objectives in the event of employer insolvency) and move to the model I described to Sam Marsh in a recent Twitter thread. (Basically discount at the risk-free rate and then apply adjustments to the resulting liability target to account for all of those factors packed into the currently used discount rates.)
So I’m heartened by your comments above (“it is possible in principle to manage a pension scheme by studying its expected cash flows in and out without doing any discounting”)!
Here’s the paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2506628
Your language is pejorative: “give sponsors wiggle room to set their own discount rates”. If a pension scheme has a portfolio whose expected rate of return actuaries can project that is a relevant parameter. It is very insulting for you to describe that, prudently arrived at estimate following professionally laid down rules, as “setting their own discount rate”. If they use what you describe as the “risk free rate” – by which you mean bonds – instead then they are making the valuation arbitrary, because the scheme is not invested in bonds.
Two points: 1. there is no such thing as a risk free rate; bond rates are volatile. 2. the method is destabilising because the only way it makes sense if for the pension scheme to immediately (if not before) to shift its investments from return seeking assets into bonds. There is a big question mark over whether it is possible for all the pension schemes in the world can switch into bonds without substantial and serious consequences for the world economy.