Doesn’t pension freedom extend to Defined Benefits?

The debate about the assumptions we use to decide the state of our defined benefit schemes is in full swing.

On the one hand you have those, like John Ralfe who use an approach which ensures that there’s always enough money in the pot to pay future pensions from an insurance company (a mega annuity policy covering everyone).

On the other hand you have those, like my colleagues at First Actuarial, who favour an approach which in the DC world we’d call  drawdown.

Of course it’s not quite as simple as that- but you get the picture.

Now when you’ve made a promise to people that the “annuity” will be at a defined level, somebody has to be on the hook for any shortfall between the price of the annuity and the amount in the pot to pay the pension. In practice defined benefit schemes don’t earmark parts of the great big pot to pay individual annuities. Actuaries value everyone’s benefit with one number and everyone’s pot as one pot. This is called pooling and is the basis of collective pensions.

Extremists dislike collectivism on principal. Collective pensions are crypto-Marxist wealth re- distributors. We’d all be better off making our way home. The stragglers at the back will be picked off by the wolves.

Extreme collectivists would have no competition but simply state benefits. There were attempts in the 70s to remove the private sector from providing pensions and to run one big national pot from the national insurance fund. Vestiges of this philosophy are still evident in the DWP and tPR’s championing of NEST.

As every, there is a happy medium that neither denies the merits of collectivism, nor disempowers people from making the best of it for themselves. I suspect that that’s what the pensions debate has been about since the War.

But I am , as my friend Vince puts it – talking from the top of the ladder of abstraction.

As regards the nitty-gritty, there have been calls for First Actuarial to provide the detail of how it has produced its FAB Index – how it has magically made DB pensions so much better off.


I am not a party to the detailed calculations. There are requests both on and off this blog for us to show our workings and I will supply the nitty-gritty to those who want it – on request. We are nothing if not transparent. But I only want to climb down that ladder one rung at a time.

Let me play you instead a debate being played out inside the mallowstreet wall-garden. I have kept the debate anonymous as it’s very Chatham House in there (you can report on what’s said but not saying it).

If you don’t want the technical argument, go to my summation below the quotes

Ralfean chappie

As you say, these regulatory numbers don’t drive Trustee action. On a TP (technical provisions-ed)  basis the cashflows are discounted by gilt (swap) yields a spread reflecting a conservative estimate of the assets’ expected return. A scheme only moves to gilts flat if they’re targeting buy-out, in which case you have to play by the insurer’s rules: if they discount at gilts flat, you should too.

As I said, I’m not sure Hilton understands risk in an asset-liability context.

Your points on sponsor strength and drawdown risk are well made. My colleague wrote a piece on that. To summarise: “Expected returns don’t pay benefits!”

An investment chappie

Hilton is quite correct, we don’t have to do it this way. We can build a growth investment portfolio which we expect will have a much higher return and we can discount on this higher basis – resulting is less money needed today. But there is no risk transfer involved. If something goes wrong with this investment strategy, the trustees have to look at going back to the employer covenant to make good. The employer also has to worry about how much damage the scheme could do to the company in the process.


First Actuarial chappie

I agree that TPR does not call for bond based discounting, and the funding regulations do not either. Nevertheless, a look at TPR’s Scheme Funding Statistics Appendix of June 2016, tables 4.1 and 4.2, show that average single effective discount rates for technical provisions have stayed remarkably close to gilt yield 1 % throughout. Although there is no requirement to follow gilt yields, it is clearly what most actuaries and trustees are doing.

If gilt based funding plans cause a problem, there’s an easy answer: stop preparing gilt based funding plans, it’s not required.

There are three arguments here

1. Investment returns don’t matter (the annuity approach)

2. We should move to a drawdown approach – if the employer can stand the risk of it going wrong. (drawdown without conviction)

3. We should trust the investment markets and have confidence (drawdown with conviction).

The Pension Plowman’s view.

I am in camp 3. I am about to enter drawdown with my own money – with conviction. I believe that with good husbandry , my drawdown plan can last me a lifetime and meet my financial needs.

The arguments put forward by FA chappie are no different. He is simply talking collectively and about other people’s money.

Here is FA chappie, a little later in the argument.

The funding regulations, as Simon and I have been pointing out, allow for the discount rate to be based on the expected return on the assets. The expected return on the assets is, by another name, the internal rate of return, which is the rate of return which values the expected income on the asset at the asset’s market value. The internal rate of return is a term which is accepted universally in all financially related professions: actuaries, accountants, surveyors, business managers, economists are all taught what an internal rate of return is.

Having worked with “gilts plus” and “internal rate of return” discount rates from before the SFO began, I’m confident that “gilts plus” is a rubbish model which is unsuitable for both long run planning of the cash flows of a pension scheme and for short term solvency valuations. And that an internal rate of return based discount rate provides a better model for long run planning.

We now have to deal with Einstein’s fourth dimension -time. As FA Chappie later accepts, you cant put a + to a gilts place funding model or use an internal rate of return if you are hoping to buy out tomorrow.

This is really the nub of it and I wish (as one Linked in poster put it) to “nip this argument in the butt!”. If you are pursuing a short-term strategy over pensions, then pensions become very expensive indeed as you have to value everything at annuity rates. Hence the deficits in PPF7800 (they’re even worse in the Hymans Robertson index).

If you are accepting that your pension scheme is going to be around for a decent length of time, then you can adopt an internal rate of return approach. .

FABI is adopting the internal rate of return approach because it is assuming that schemes are not looking to buy-out tomorrow but will continue till the last payment needs to be paid.

My view is that the problem is being caused by a rush to buy-out which is quite without economic justification. It is precisely the opposite view of John Ralfe’s.


Making sense of it all

If you’ve read this far, then you are probably an expert, but I hope there will have been a few amateurs (like me) who have followed me down this little road.

We have two views of the pensionworld – Ralfe’s and First Actuarial’s; they are very different views of the world.

John Ralfe wants to annuitize. We have conviction in drawdown.

John’s view tends to gilts- ours tends to equities.

It really isn’t any different to the conversations we are having as part of our personal pension freedoms.


Do you want to be free?



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.

2 Responses to Doesn’t pension freedom extend to Defined Benefits?

  1. John Mather says:

    Henry isn’t that a very long winded way of saying IRR of 8% which is my back of envelope way of calculating the results of your graph. Its a similar argument to the one put forward with Prufund the difference seems to me is the blue line is drawn by an actuary rather than an insurance salesman

  2. henry tapper says:

    see the assumptions blog John- not quite that high – more like gilts + 3.5% at the moment – though that is not the way IRR is calculated

Leave a Reply