News that transaction charges take up 37% of all investment management costs will come as a big surprise to those who study the disclosures made by IGCs and trustees to savers in workplace pensions. While there are outliers most IGCs are reporting members paying no more than 0.03% pa in transaction costs.
The savers in a workplace pension is taught they’ll not be paying more than 0.75% pa for their investments – so their ought to be some pushback on CACEIS’ number, 37% of 0.75% is over 12 times what the IGCs are reporting. Either the IGCs are under – reporting or CACEIS is over- reporting or the article is just not doing its job.
I will try to make sense of this of all this by looking at each of these three matters separately and I’ll then try to synthesise things into a conclusion,
In thinking this through I find the article very complicated and most definitely not as transparent as we should expect. For CACEIS’ assertion to hold good , it needs to work for all pension investments. That means DB funds and DC funds. Let’s start with the workplace DC funds under the auspices of IGCs and Trustees
Is DC under-reporting transaction costs?
I start with an assumption that CACEIS are right with their 37% figure, no matter how outrageous it may seem to DC savers. Here are the three reasons that CACEIS “37%” number might be right for workplace pension defaults.
IGCs take their reporting not from the platform but from the fund manager. What this means is that what the fund managers are telling you not what you are actually getting. This is a bit like those old 56 mpg claims for the reasonably priced car. We all know we only got half that but in the lab , the engine manufacturers were able to tease their results to whatever showed best.
What we need to know is the engine’s fuel economy on the open road, and it’s the same for pensions, the costs we get from fund managers can be a lot higher when money is coming in and out. IGCs may be being fed transaction costs without the knock-on impact those costs have on performance. DB is transaction heavy and IGCs and Trustees may not be picking up the spreads incurred in the investment of regular contributions and the disinvestment to pay the money back.
Secondly,(and to continue the analogy), even if the MPG testing was based on an achievable performance in the car, you never saw the impact of going up and down hills, idling in traffic jams or spurting to pass other cars. To extend the analogy, IGCs and Trustees are only telling us the impact of the fund charges, not the cost of changing funds when life-styling. IGCs may not be fully reporting the cost of transitioning within their costs and charging
Thirdly, the amount that those who manage the fund platforms for savers in workplace pensions don’t pay what we pay as investment charges. We might pay 0.5 to 0.75% as an annual management charge but only a tiny part of that is spent on fund management. That’s because platform managers drive a hard bargain, only use passive funds and can reduce their fees still lower by ruses such as stock-lending. So 37% of what they pay for fund management may be in line with the amounts the IGCs and Trustees declare as the transition costs on their funds. Because the AMC we pay on our workplace pensions doesn’t tell us what the fund managers are being paid, the transition costs may be a much bigger deal than we might have thought.
I think it very likely that the seemingly tiny transaction costs may well be close to 37% of the total cost, partly because IGCs are under-reporting the impact of transaction costs and partly because the investment element of the AMC is overstated.
In short, IGCs and Trustees may be fulfilling their compliance duties but they are not explaining costs and charges in an intelligible way to DC savers. If CACEIS make a claim that cost and charges are 37% for pensions, they are going to have to show how this works for DC. I will restrict this argument for now to workplace pensions but it is also pertinent to non workplace pensions (legacy personal pensions and SIPPs).
CACEIS research points towards valuable insights for DC schemes but other than comparing DB total cost of ownership to the DC charge cap (see below) , it makes no mention of DC. DC savers stay confused.
Or is CACEIS’ statement misleading?
Having found it possible that CACEIS are right on DC, I then returned to the article to see if I could understand what CACEIS bold assertions meant
Firstly, Pension Age appear to be reporting on a press release not a report. There is no click through to the data and nothing on the CACEIS website to support the bold assertions.
I had to admit defeat (I did put out a message to CACEIS and will include any further news from them as and when I get it.
Having failed to get information from CACEIS , I turned to an independent source – Clearglass and its Chair Chris Sier.
Chris explained to me that the total cost of owning (TCO) a pension fund for a typical Defined Benefit pension scheme in the UK is indeed considerably higher than what DC platforms are paying for the funds they offer us.
Chris commented on the statement in the Pension Age report – presumably by CACEIS data that
UK total cost of ownership stands at 70bps on average – under the defined contribution charge cap of 75 bps.
Chris accepts that their estimate for average TCO is close to a part of his data, but only a part – specifically it is what Clearglass recognise as the cost of ownership for DB schemes with assets between £1bn and £5bn pounds.
He points out that larger schemes (£5bn+) tend to get better deals because of their size and smaller schemes (including most DC schemes) use passive funds (behaving more like DC platform managers).
Chris is worried that many trustees will see the average TCO of 0.7% pa as a benchmark, allowing trustees to think they’re doing alright when in fact they could be doing a whole lot better.
I suspect that there is also a danger of focussing purely on the money part of the value for money equation. If trustees are paying for and getting a targeted absolute return, they could be getting VFM with a high AMC. Workplace pension plans may provide poor VFM despite having a tiny TCO.
I am worried for the small DB schemes and almost all the DC schemes who will see 0.7% as an investment TCO as prodigal. But should they start thinking they ought to be spending more on owning assets to keep up with the billion pound plus schemes. The fund management houses continue to ask for DC funds to have the capacity to invest like DB but from these figures, this would mean a massive increase in the cost of owning a workplace pension – a cost that would be borne entirely by the member.
All this still doesn’t answer my more fundamental question. Why are DC platforms getting funds with TCOs at a fraction of the 0.7% paid by DB schemes. Are they simply getting better deals, are they cutting corners or are they under-reporting their TCOs?
I have no reason to disbelieve that the TCO of CACEIS client is 0.7% or to doubt that more than a third (37%) is going in transaction costs . But as I have no evidence of the scope of CACEIS research , I can’t work out if it covers all pension funds or just a small sample of DB funds.
So I do find the 37% figure a lot less authoritative than this “exclusive” would have us believe . Chris Sier has good reason to consider the 0.7% total cost of ownership figure , a misleading average.
Or is it just the reporting?
The third area I wanted to test was whether the article reporting the CACEIS research is in itself transparant. I find it fails on three counts
- It confuses DC and DB costs
- It confuses actual transaction costs with the reporting of these costs
- It gives a bogus view of exclusivity – it is giving a restricted view of the market
Confusing us on DC costs
Perhaps the most confusing aspect of this article is that the DC charge cap isn’t an investment charge cap at all. It is not a comparator to the TCO of a DB fund because what a saver is paying in that 0.75% is the cost of being supported in every aspect of the savings process. Typically , the investment charge being less than 0.1% the remaining amount is paying for things that DB funds pay for separately.
To make matters even more complicated, the transaction costs (the 37%) aren’t included in the DC charge cap.
So what the article has done is conflate DC and DB cost and charges giving DB trustees a false sense of confidence they are properly controlling their costs. The article is, according to Chris Sier overstating the impact of fund management on the total cost of ownership by providing analysis on too small a data set. And I very much doubt that any of the data that CACEIS has analysed has been for DC workplace pensions.
Confusing transaction costs with transaction cost reporting
Another difficult part of the article is that there is simply no sense in some of the statements it makes. For instance
The proportion of investment expenses made up by transaction costs has increased from CACEIS’ previously reported figure of 25 per cent.
Transaction costs have not increased from 25% to 37% since the last CACEIS report, CACEIS is now capturing not just their costs but their impact. The costs are the same but the attribution of impact now shows a truer burden on sponsors from the trustee’s ownership of assets.
Confusing us on “exclusivity”
This “exclusive” is excluding the analysis of rival firms looking into costs and charges and is infact little more than a press release based on information that is not in the public domain.
And above all , the article’s exclusivity is bogus. It has been bought at a high cost to the integrity of the transparency agenda.
This is very far from the transparency that the costs and charges agenda is supposed to be promoting. Instead it suggests that “cost and charges” analysis has become part of the problem it was supposed to solve.
I do think that CACEIS’ assertion that more than a third of pension investment costs arise from transactions is interesting and deserves to be tested over the whole market.
I find the article guilty of confusing CACEIS’ market with pensions as a whole and confuses us with DC comparators.
I find CACEIS’ failure to provide us with the information on which we can make judgements , wilfully obscure.
If firms that report on transparency, don’t show us their working , we are back where we started.