Pensions Council rejects Colm Fagan's proposal despite its independent consultant endorsing it

[jasdpace@gmail.]

Is your last question not the same as the first?

Well, when you put it like that, I suppose it is!

[On a serious note, it was simply intended to reinforce the point that (if the identified risk of negligible ERP materialised) the chosen system doesn't protect the AE contributor from this risk. In a world where the ERP vanishes or is minimal, we can accept that your proposals will be less beneficial to contributors than the central modelling anticipates. However, the PC's decision means that, in this Non-ERP world, the design that they have chosen will, in all probability, lead to even poorer outcomes. Otherwise put, the chosen design protects contributors less well from the identified risk.]

 

[Colm Fagan]

So basically the state steps in to shore up the fund in periods of poor performance. Completely in opposition to one of the key objectives of AE which is to reduce future reliance on the Exchequer!

The state will not have to step in (other than to pay excess costs over management charges in the early years. This applies irrespective of whether the scheme is structured as smoothed equity or on the lines proposed by DSP.)
When net scheme cash flows are positive (expected for the first 30 years or more), exits (at smoothed value) are balanced by joiners (also at smoothed value). Any excess of joiners over leavers means new money for the scheme. If SV>MV at date of joining, then new joiners suffer an immediate loss equal to the difference between the two for the net new cash flow in that month. If SV<MV, then the opposite is true: new joiners make an immediate profit. Joiner profits and losses should balance out in the long-term. The state has no involvement.
Cash flows must turn negative at some stage (in what I think actuaries call a "stationary population"). In that case, net exits receive the smoothed value on exit, but market value is withdrawn from the fund. Any difference between smoothed value and market value for net exits is dealt with through the buffer account, which will start to be funded from around year 20 (by which time I expect management charges to exceed expenses). Once again, the state will not be involved.
If things go badly wrong, if for example cash flows turn negative before the buffer account has been established, then the state may be asked to help out, to prevent the scheme being wound up, in which case members would still get the market value of their account, which, I accept, might be less than smoothed value in those circumstances. As an additional belt and braces, to prevent wind-up, the government could start funding the buffer account from the start (or close to the start) by contributing a portion of what it saves relative to what it would have to pay under the DSP scheme (1% of earnings under the smoothed scheme v 2% of earnings under DSP scheme). I'm confident that the funds contributed by the government in this manner will never be required and will eventually be returned to it (with interest!).

 

[Dr Strangelove]

Colm first you say:

The state will not have to step in (other than to pay excess costs over management charges in the early years.

Then you say:

If things go badly wrong, if for example cash flows turn negative before the buffer account has been established, then the state may be asked to help out, to prevent the scheme being wound up, in which case members would still get the market value of their account, which, I accept, might be less than smoothed value in those circumstances. As an additional belt and braces, to prevent wind-up, the government could start funding the buffer account from the start (or close to the start) by contributing a portion of what it saves relative to what it would have to pay under the DSP scheme (1% of earnings under the smoothed scheme v 2% of earnings under DSP scheme).

I’m not trying to score debating points here, seriously, but you can’t have it both ways!

Don’t get me wrong I like your idea in principle. I think “lifestyling” is poor advice for almost everyone and people should be heavy in equities for as long as they can.

But you are proposing heavy state involvement in something that is supposed to de-risk the state. And a large mutualisation of risk (and reward) in what is designed to be an individual savings vehicle.

I wish there was a way to preserve more of your ideas without dreaming up a very different model. Could you at least argue in favour of an equity-heavy portfolio mix as the default?

 

[Brendan Burgess]

Could you at least argue in favour of an equity-heavy portfolio mix as the default?

I think that this really is the point.

Put simply, Colm is suggesting that everyone should have their pension in equities during accumulation and drawdown. That will give a much better return than investing in cash and bonds.

Colm's scheme does not increase the return for someone who is invested in equities anyway. He just makes sure that they are not subject to the wild fluctuations of the stock market.

Brendan

 

[jasdpace@gmail.]

Hi Brendan,

I think we need to be a little careful in what we say.

Put simply, Colm is suggesting that everyone should have their pension in equities during accumulation and drawdown. That will give a much better return than investing in cash and bonds.

I believe that Colm is not suggesting that 100% equities is appropriate for the standard AE contributor in retirement (during the drawdown phase) if such a person is investing on an individual basis. He was advocating 100% equity investment on a pooled basis where the inherent volatility of equities was managed by the smoothing mechanisms of his proposals. In other words, the all equity approach was argued to work in a pooling environment where specific control mechanisms are in place and because of the somewhat unusual nature of the AE scheme itself (like guaranteed on-going cash-flow, strict anti-selection protocols, etc.).

Colm's scheme does not increase the return for someone who is invested in equities anyway. He just makes sure that they are not subject to the wild fluctuations of the stock market.

I'm going from memory but the, broadly, doubling of anticipated income in retirement under Colm's proposals was made up from 3 main sources.
1. Impact of reduced charges pre & post retirement
2. Impact of staying in equities longer in the lead-up to retirement
3. Impact of staying in equities during the draw-down phase.

I hope Colm will confirm and comment, as appropriate. Personally, I'd be really interested in understanding how much is wasted as a result of point 1 above?

 

[Gordon Gekko]

Why doesn’t someone (i.e. a life company) just take this concept and commercialise it?

(appreciating the history of ‘with profits’ etc)

 

[Dr Strangelove]

Why doesn’t someone (i.e. a life company) just take this concept and commercialise it?

I’d imagine as it has some characteristics of a Ponzi scheme.

Namely dipping into other clients’ funds for your drawdown during periods of poor market performance.

 

[Itchy]

I'm going from memory but the, broadly, doubling of anticipated income in retirement under Colm's proposals was made up from 3 main sources.
1. Impact of reduced charges pre & post retirement
2. Impact of staying in equities longer in the lead-up to retirement
3. Impact of staying in equities during the draw-down phase.

I hope Colm will confirm and comment, as appropriate. Personally, I'd be really interested in understanding how much is wasted as a result of point 1 above?

In addition, I would have thought any risk (burden) caused by the potential state funding of the buffer account would be minimal compared to:

1) the vastly inferior returns that will accrue under the proposed system - whereby Colm's proposal proposes to massively reduce the relative burden on the state as a result of larger pots accruing to people who would otherwise be subject to investment risk e.g. suffer the burden of higher fees, lifestyling, poor asset allocation etc. hence greater (and ongoing dependence on the state pension/health resources etc.) and

2) the effect of the smoothing formula means that if the State had to meet the gap in some way that it would be time limited, reduced by the effect of new contributions, lower fees and the integrity of the fund would be retained providing the opportunity for recovery. In addition, I think Colm addressed the negative cashflow scenario by indicating that the management fees could be increased (thus the additional spread could be captured) and the state could temporarily incentivise greater contributions to the fund and thus have an asymmetric effect from increased employer and employee contributions 6:1 (thats my perception anyway).

 

[nest egg]

In addition, I would have thought any risk (burden) caused by the potential state funding of the buffer account would be minimal compared to:

1) the vastly inferior returns that will accrue under the proposed system - whereby Colm's proposal proposes to massively reduce the relative burden on the state as a result of larger pots accruing to people who would otherwise be subject to investment risk e.g. suffer the burden of higher fees, lifestyling, poor asset allocation etc. hence greater (and ongoing dependence on the state pension/health resources etc.) and

2) the effect of the smoothing formula means that if the State had to meet the gap in some way that it would be time limited, reduced by the effect of new contributions, lower fees and the integrity of the fund would be retained providing the opportunity for recovery. In addition, I think Colm addressed the negative cashflow scenario by indicating that the management fees could be increased (thus the additional spread could be captured) and the state could temporarily incentivise greater contributions to the fund and thus have an asymmetric effect from increased employer and employee contributions 6:1 (thats my perception anyway).

I think you're hitting on two interesting points.

1) Comparison: As far as I'm aware, only Colm's AE approach has been independently evaluated. Given the impact AE will have on the population for years to come, if any of us had this decision to make, wouldn't we want a similar evaluation of the proposed AE approach? There will be no perfect solution after all.

2) Risk: Today risk has shifted entirely from the employer to the employee. Gone are the days of a guaranteed / DB income (unless you work for the civil / public service). I'd agree the risk is low for Colm's scheme but even in saying that, it's entirely reasonable in this context to ask why the State shouldn't take some of the risk burden, especially considering the reward on offer, and that AE is targeting people who have no pension provision today, those who can least afford poor returns, higher charges, and the complexity & additional costs arising when setting up an ARF/annuity on retirement.

 

[Colm Fagan]

But you are proposing heavy state involvement in something that is supposed to de-risk the state. And a large mutualisation of risk (and reward) in what is designed to be an individual savings vehicle.

The state is on the hook big-time with AE, no matter how the scheme is structured. The NEST scheme in the UK is "pure" DC - no guarantees, no smoothing, no nothing - yet the UK government was in the hole for a cumulative £918 million at March 2023 (including a further £110 million loss in 2023, when the scheme was already over a decade in existence). If the NEST scheme can't wash its face after more than a decade in existence, with charges that equate to more than the 0.5% of assets under management proposed for the Irish scheme (note: is that still the case??), what hope does the Irish scheme have of ever breaking even if it's structured on similar lines, given that the Irish scheme can only hope to get about 10% of the membership of the NEST scheme, but its costs will be far more than 10% of the UK scheme's costs.
The financial position of the smoothed scheme from the government's perspective will be vastly better than that of a scheme run on NEST lines. There will be just one fund, which will only need to be valued once a month. Everyone will get exactly the same 'interest rate' every month. Furthermore, members will stay in the scheme post-retirement, generating high margins which will more than cover costs, enabling the charge to be reduced further. Workers leave the NEST scheme at retirement.
The state is also on the hook for its contribution to the scheme which, according to DSP's plan, is 2% of earnings. Under my proposal, government would only be required to contribute 1%. If you had read on from "the state may be asked to help out", you would have seen that I said that, if the state contributed just half the 1% saving relative to the DSP's proposed scheme to the buffer account from the start, it would completely eliminate the risk of it ever being "asked to help out".

 

[Colm Fagan]

I hope Colm will confirm and comment, as appropriate. Personally, I'd be really interested in understanding how much is wasted as a result of point 1 above?

You're right that total costs should be massively lower under the smoothed scheme I'm proposing. Members will get the benefit of those lower costs.
There will be just one fund, which only needs to be valued once a month in order to calculate the smoothed return for everyone in the scheme (active and retired). In contrast, a scheme structured on the lines proposed by the DSP will need something like 50 default funds (pre-retirement), one for each projected maturity year, in addition to the low-risk, medium-risk and high-risk funds. Each of those funds will need to be valued daily. Investment strategies will need to be devised and then implemented for each of the funds - the mix between equities, bonds, cash, alternatives, etc. For the smoothed fund, everything will be invested in a cheap-as-chips passive world equity fund (at least in the early years). No messing around with asset allocation strategies, etc.
Workers won't have to choose, either at the start nor at any later stage, what fund they want to invest in, with all the potential for mis-coding and resultant hassle in order to correct the errors. There will be no choice - everyone will be in the smoothed fund.
There will also be far fewer queries to the administrator. Can you imagine a workplace (say the local tyre repair shop), where workers discuss at coffee break how their fund is doing. Under the DSP's scheme, two or three people working side by side will have different returns on their funds, depending on their fund choice. Even two workers, both of whom have chosen the default option, may experience different returns just because of a difference in ages. In the smoothed scheme, everyone in the country (active and retired) will get exactly the same "interest rate" every month. It will be a live topic of conversation in pubs, hairdressers, etc. There will be no confusion.
As anyone involved in the business knows, reductions in unit price generate queries. Under the scheme proposed by the DSP, workers will see the value of their fund fall once every three months or so (particularly if they've chosen the high risk option). Current market value gets only a 1% weighting under the smoothed scheme. There will be a positive "interest rate" almost every month, resulting in fewer queries.
Volatility of returns also causes people to drop out. Brian Woods and I did a few back-of-a-fag-packet calculations, which indicated that over half of the 10 million or so claimed members of the NEST scheme have stopped contributing to the scheme. Drop-outs adds significantly to costs. There will be far fewer drop-outs under the smoothed scheme.
I haven't yet got near what happens at retirement. Speaking from experience, there are massive frictional costs at retirement under any sort of conventional DC scheme. Those almost disappear completely under the smoothed scheme. Everyone keeps exactly the same account. They just start drawing from it rather than contributing to it. Much the same happens throughout retirement.
It's getting late - it's already tomorrow! - so I must stop, but you get the message: costs are massively less under the AE scheme. Those lower costs must go to the benefit of the membership.

 

[Colm Fagan]

I’d imagine as it has some characteristics of a Ponzi scheme.

Namely dipping into other clients’ funds for your drawdown during periods of poor market performance.

Unfortunately, this misunderstanding is widespread when people haven't studied my proposal or haven't understood it. It is totally wrong. There are no characteristics of a Ponzi scheme; there is no dipping into other clients' funds.
Probably the easiest way to explain why neither is true is to consider that the scheme is always valued at market value (MV), not smoothed value (SV). The SV concept is only relevant for people (contributions actually) joining or leaving at a particular valuation date.
If SV>MV at the valuation date (this is the situation most people worry about), someone joining is paying over the odds to get in. If they're being balanced by someone leaving, the two trade at SV - the person coming in loses, the person going out gains. My argument is that people will (normally) be happy with that agreement. People see that it balances out in the long-term. In the process, everyone gets a smoother ride while they're in the scheme. (We can discuss in a separate post what happens if they're not happy with that agreement).
Therefore, at any time, the MV of someone's interest in the scheme equals their smoothed value (which they know precisely - the balance as per last statement plus contributions and 'interest' since), multiplied by the ratio of MV to SV for the scheme as a whole.
To illustrate, suppose I'm told that my (smoothed) balance in the scheme is 100. I also know that the SV of the entire scheme is 1,000 and the MV of the entire scheme is 900 (BTW, I'm proposing that this information is always in the public domain), then the MV of my interest is 90 (=100*900/1000). Every member of the scheme can do the same calculation, and the total for all of them adds up to the total MV of the entire scheme. There is nothing spare for future members of the scheme. Neither does the scheme owe anything to people who've already gone or due to join in future. There is no dipping into other clients' funds at any time.
I hope I've explained it adequately.

 

[ronaldo]

I'm sure the Pensions Council had their minds up long before an independent report was commissioned. Why go to the expense of commissioning a report if you're only going to reject it? On average, members likely consider themselves infallible and never envisaged the possibility that an independent consultant hired by them would give their stamp of approval to a concept not thought up by themselves.

I'm not sure why they argue that they should avoid the approach because they "did not find any precedents for an investment approach akin to the proposal in global, national, or provincial pension provisions". I imagine they concentrated their efforts on studying state pension provisions across a multitude of international countries. I'd love to hear their response to the question "what percentage of the global, national, or provincial pension provision schemes that they looked at for guidance are underwater versus what percentage are considered sustainable into the future?".

I wonder will their next argument be that the proposed solution isn't ideal because European markets have dropped ~1.5% today.

If designing a new scheme, why look at a multitude of broken schemes to come up with your own? Surely, if they had an answer to the following three questions in relation to the schemes they looked to for precedent, they'd realize that a scheme with no precedent is what's needed:

• Is it sustainable (I'd imagine there are only a handful, if any, state pensions throughout the world that meet this);
• Is it costly (fees have a massive impact on performance over the timescales involved, as we all know);
• How much are people likely to get out the other end

Regarding point three, some people are arguing that the government shouldn't need to subsidize the scheme at all and I get the argument. However, according to the table in post 5, someone joining at age 25 is likely to end up with 53.5% of their salary in pension provision under the proposed system versus 22.3% under the lifestyling approach.

Imagine the scope the government would have in decades to come to freeze state pension rates, or target annual increases at sub-inflation levels, and let the states pension expense dwindle with inflation if the 25 year-olds of today were set to have private provision to the tune of 53.5% of their salary.

 

[jasdpace@gmail.]

Edit: My post has been moved from the thread you started today - no idea why?

Hi Colm,

I find myself getting very interested again in all this stuff. In fairness, it's a remarkable insight into how decisions are made. [I've just seen Ronaldo's post and apologies for a wee bit of overlap.]

I can't say for certain that I wouldn't have come to the same conclusion as the PC but I'd hope that I would have treated you much better during the deliberation process (principally by engaging meaningfully with you) and given you credible reasons for ultimate decision. FWIW, my sense is that if there had been robust discussions with yourself and Brian and honest communication by the PC in which the PC explained clearly the areas causing them real concern, I think it may very well have been possible to address these concerns in a way that still manages to capture, at least the majority of, the gains identified in your proposals (i.e. broadly dialing down the bounty slightly for greater security). Frankly, it's pretty ridiculous that it's got to the stage where you feel compelled to be having debates on social media.

Just three points for now (.....else I'll be at risk of navigating my own retirement sooner than planned!)

1. What was the purpose of the independent valuation - as in, should the C&AG or whoever has oversight here be asked to comment? It feels to me like a complete waste of taxpayers money!!!! [Not perhaps Uni of Limerick levels of waste, but still....]

2. Are we absolutely certain there's no comparator anywhere to your proposals? [The likes of Mercer compile details in summary form of the pension structures (in existence and in the pipeline) in every country in which they operate (so pretty much everywhere).]

3. The Authorities have been designing this since 2006 but we're not really sure if, what is quoted below, is what will happen in retirement. [The Duke explained the situation in this regard yesterday].

Under its proposals, workers will be turfed out at retirement and forced to hawk their savings pots around the market.

 

[Duke of Marmalade]

Brian Woods and I did a few back-of-a-fag-packet calculations, which indicated that over half of the 10 million or so claimed members of the NEST scheme have stopped contributing to the scheme.

Possibly a fag-packet was a tad too small. Using the back of an envelope I get the following. NEST in its accounts boasts of 11 million members. It further boasts that these pay £400m per month. These are impressive numbers except when you divide £400 by 11 million you get an average of about £35 per month. That would seem to me to indicate that whilst 11 million have enrolled and presumably did not take the first opportunity to opt out more than 80% of them are no longer active contributors.

 

[Duke of Marmalade]

More on NEST.
Earlier back of envelope calcs were from 2021 accounts. From the latest accounts published in October 2023:

It boasts 12m members but only 4.8m are active. The scheme is running 11 years so we can broadly estimate the half-life of active members as about 5 years. If that were to continue for say 20 years we would expect only 5% of current members to be still active. Okay, one expects persistency to improve with duration but nonetheless these are persistency rates well below industry norms. Average funds per active member are roughly £5k after an average of 5.5 years contributing. Is this really the model to defuse the pension time bomb and for Ireland to follow?

 

[Dr Strangelove]

The scheme is running 11 years so we can broadly estimate the half-life of active members as about 5 years. If that were to continue for say 20 years we would expect only 5% of current members to be still active.

Not really. Many inactive members will opt in again. Think of a 35 year old taking time off to raise kids. It’s very likely they’ll be in paid employment again.

Many people move between inactivity, employment, and self-employment through an adult lifetime.

 

[Dr Strangelove]

Average funds per active member are roughly £5k after an average of 5.5 years contributing.

How do you get this? Total assets divided by scheme members more like £2.5k.

 

[Colm Fagan]

I can't say for certain that I wouldn't have come to the same conclusion as the PC but I'd hope that I would have treated you much better during the deliberation process (principally by engaging meaningfully with you) and given you credible reasons for ultimate decision. FWIW, my sense is that if there had been robust discussions with yourself and Brian and honest communication by the PC in which the PC explained clearly the areas causing them real concern, I think it may very well have been possible to address these concerns in a way that still manages to capture, at least the majority of, the gains identified in your proposals

Thanks for that - much appreciated. As you say, the sensible approach was to engage with Brian and me throughout the evaluation. As professionals, we want to get the right outcome, not to push our point of view. I don't know why they decided against that.
One of the aspects that's in danger of being forgotten (partly because it was strangely missing from the Minister's April 2023 letter to the Pensions Council) is the massive sums involved. My claim (supported by the expert appointed by the Pensions Council) is that the potential saving to the nation (employers, employees and the state combined) is of the order of €1.5 billion a year when the scheme is mature. That merited a far more detailed evaluation than could be done with a €50,000 budget and a short timescale.
Given the sums involved, it was natural to ask the ESRI to look at it, and to encourage them to speak with Brian and myself, as well as others interested in the proposal, from either a negative or positive perspective. I agree that they might conclude that my claim is over-optimistic (I don't think so!!) but even getting half-way there would be brilliant.

 

[Duke of Marmalade]

How do you get this? Total assets divided by scheme members more like £25k

29.6bn/4.8m = approx. 6k, knock off something for inactive members to get 5k, say

Not really. Many inactive members will opt in again. Think of a 35 year old taking time off to raise kids. It’s very likely they’ll be in paid employment again.

Many people move between inactivity, employment, and self-employment through an adult lifetime.

My experience was more with insurance policies and the incidence of what we called "paid-ups" coming back to the faith was negligible. I am surprised that you find the NEST figures reasonable. If this is an indicator of how the Irish DSP scheme will unfold there isn't a snowball's chance that it will be viable at 0.5% AMC which is actually less (for a considerable initial period) than NEST's 0.3% AMC + 1.8% of contributions.