In a dignified article in the FT, Steve Webb  warns that the Financial Reporting Council’s plans to project pension returns based on how investments have done in the past comes with the real risk that ” the new statement figures will make little more sense than the old ones“.

This is the house line at LCP, where Steve is a partner and it’s a line that has been pension orthodoxy for a few years now

Standardisation will provide consistency – but needs to be simple. A cross industry working group in 2019 concluded that the use of future expected returns for each of 7 key asset classes was appropriate for setting growth rate assumptions. @stevewebb1 https://t.co/xJCCWvzbm1

— Ruston Smith (@RustonSmith1) June 4, 2022

I am out of step with Ruston and Steve in supporting the FRC’s approach and you don’t disagree with such people, without having a good reason. My “good reason” is that the FRC are taking us down a path that they hope will allow people to compare their pension pots and “make sense of the numbers”.  Members – not actuaries or employers – are the people who take the risks in pensions not meeting expectations. With a pension dashboard looming, where people will see all their pots projected on a standardised basis, we need to get that standard right. In this blog, I argue that rather that it is member experience not actuarial speculation, that should drive projections.

I will put aside the question of whether annuity rates can be used as a means of converting pots to pensions. The pension industry don’t seem too keen to move on from 2014 on annuities, but more worryingly they don’t seem too keen to move beyond “DB thinking” on projected growth rates of our DC funds.

Making sense of the return you have had , are getting or will get on your money is almost impossible if you don’t know what has happened so far (your experience). What LCP and others want us to do is to look at the future with no regards to the past but with regards to assumptions agreed by a concensus of industry experts (see Ruston’s tweet).

Following this logic, we might ask why these expert predictions haven’t applied in the past.  Part of the answer is of course that the returns we actually get on our money aren’t just decided by how our investments have done, but how much middle-men have taken out of our pots and how lucky we’ve been with the timing of contributions. I mention this last point as we see considerable variations in the returns people get for no obvious reason than investment administration.

Generally speaking, the more meddling that is done with people’s money, the more likely the costs of meddling will outrun the value , meaning that actively managed funds will struggle to provide better returns than those with low maintenance set and go approaches.

This is particularly the case when pension funds try to protect people against nasty surprises ( what is called de-risking). This usually involves a smoothed transition from what are seen as “volatile” investments – such as equities, into what are seen as “stable” investments – such as bonds and cash. These “glidepaths” have been constructed and administered at considerable expense to member’s funds because actuaries believe that certain asset classes have inherent characteristics that should deliver predictable returns.

This kind of abstract thinking is based on the management of DB schemes where – should the theory go wrong, there is somebody else to pick up the cost (usually the sponsor). Unfortunately, the only person picking up a shortfall in our DC saving is the saver.

Here is Steve’s and LCP’s grouse with the FRC’s approach to projecting forward using the evidence of the past.

Had it been in force five years ago, for instance, a fair number of equity fund values would have been found to be relatively stable while some government bond returns would have been relatively volatile. Applying the FRC’s logic, this could mean that where a saver was invested heavily in equities the provider may have to project a low rate of growth; where the saver was invested heavily in bonds, the provider may have to assume a high rate of growth. This is the exact opposite of the approach currently taken ….. and also the opposite of what we would expect to see in practice.

There are two problems here

1. The statement implies that people who were pitched out of calm equities into troubled bonds – to stabilise their fund – experienced precisely the opposite of what was intended (by those designing the lifestyling)
2. The example given is based on a selected period of time which is likely to be at odds with a longer-term trend – on which the experts are basing their assumptions

My challenge to the experts on

(1) is “has anyone tried to explain to members how the experience of lifestyle compares in retrospect with the theory of how it was supposed to work?” I am pretty sure that there is no explanation given to members as to why things didn’t go as expected.

On (2), my challenge is to LCP and others – “why  don’t they look at actual member returns over longer periods to provide themselves with evidence of what is working and what isn’t?”

What the FRC are attempting to do is to make sense of the past to explain to savers what is likely to happen in the future. If there approach has weaknesses it is that their view of the past is too short (like Steve’s example it risks being base on historical anomalies – such as the impact of QE). The second weakness of the FRC approach is that it assumes that whatever is likely to happen to someone’s money in the future is driven purely by the return on the assets. It ignores the impact of charges, or transition costs and of good and bad  administration. In short, projections assume a standard outcome based on everything but the assets being the same. The FRC are going in the right direction but , to me, they could go a lot further.

This doesn’t mean I don’t support the FRC – I do – but I see their approach as the start of a process which will lead eventually to projections being based on experience and not speculation.

The saver, comparing various pots has no obvious means to make a choice but on the basis of which pot is likely to grow fastest (based on the FRC’s projection numbers). Which suggests that the obvious direction of travel for projections is not back into the conventional wisdom that future returns can be modelled by economic models (the actuarial approach) but to a more radical view that future outcomes are determined by the long term capacity of a pension fund to deliver good returns.

If this approach was followed to its conclusion , we would be allowing pension funds to project forward based on the experience of its members (and should there be insuffecient evidence – this should mean no projection is given).

This approach has two advantages. Firstly it makes sense to savers who have no other anchor than what has happened to their money so far, and secondly it makes for a degree of accountability to pension scheme providers, for their delivery of good or bad outcomes.

Data exists within all schemes for this approach to work. All that is needed is that the experience of each cohort of savers is measured and averaged. The consolidation and maturing of DC schemes should mean that the vast majority of people’s pots could be projected on an experienced basis. Is it too big an ask of pension funds to provide member specific illustrations? I don’t think it is. They need only concentrate on the experience of savers in their default fund.

The outliers – those pots that fall outside default pathways, should be considered as “outliers” and not get a projection – this would exclude the vast majority of self-select schemes and policies (SIPPs) and those in workplace pensions that have opted out of the default strategy.  For such people, any projection based on the experience of others , should be distrusted. The people who self-select must be assumed to have self-confidence or advice.