The DB Pensions Crisis

The Irish Times carries an important op-ed by Michael Walsh of Mercer on the crisis in the defined-benefit pension system.  Compared to the fiscal, banking and employment crises, the DB pensions crisis is largely below the radar — but still hugely important.  

The focus of the article is on the need for policies to ease pressures on sponsoring businesses, and to prevent the more radical response of winding up existing schemes.  Of course, the proposals have potentially huge implications for a key component of the wealth of many Irish workers.  Unfortunately – but perhaps inevitably in an 850 word article – the actual reforms being proposed are less than clear.  I hope the diverse readership of this blog can provide some clarifications and perspective. 

There are two proposals:


The Society of Actuaries and the Irish Association of Pension Funds have put a proposal to government to address this issue. It would involve insurers being allowed to sell, and pension schemes being allowed to buy, a new kind of annuity. These so-called sovereign annuities would be directly linked to Irish government bonds. They would therefore be much cheaper than conventional annuities. This would increase the chances that pension schemes can continue to operate and make good current funding deficits over time.

How exactly would these sovereign annuities work?  That they lead to cheaper annuities immediately suggests a reduction in the present discounted value of the expected benefit stream – that is, a loss in wealth. 

And second:

Our proposal at Mercer is that pension schemes that wind up be permitted to pay lump sums to pensioners instead of buying annuities. The lump sum would be the capital value of the person’s pension calculated on a prescribed basis. The calculation would allow for current life expectancy and expected future mortality improvements together with a specified long-term rate of interest or a rate linked to average euro-zone bond yields. Pensioners could then put the money in an Approved Retirement Fund from which income could be drawn down. Alternatively, they could use the money to buy an annuity, although if this is done at the current time it would likely be for a lower amount than their previous income from the pension scheme.

The proposal as written comes across as relatively painless.  But as Michael says, the devil is in the details.   What exactly is the “prescribed basis”?   Again, there is a presumption that any reduction is the contingent liability of the sponsor is also a reduction in the contingent asset of the member.

Michael Walsh may well be right that, all things considered, such reforms are warranted.  But the stakes for many individuals are such that a public debate is crucial, notwithstanding the complexity of the issues involved. 

21 replies on “The DB Pensions Crisis”

What are we thinking?

We are living longer! Celebrate! The fact that our “economic model” is no longer as relevant, is the point. We need a new one. Half pensions? Allow workers to work, in a smart economy, this is possible? Of course it is.

The challenge is to create consensus on diversity by consent. No forced retirements. Reduced benefits that may allow for private schemes, invested in property or wherever. The fear of poverty drives the need for pensions. A political guarantee to ensure certain minima may remove fear. But perhaps we want the fear? It shortens lives, as stress!

I think one of the major omissions from this article is the poor performance of Irish money managers in the last decade – primarily due to over-investing in equities and primarily Irish equities. Mr Walsh makes the point that average ROI has been nil over the last decade but that does not excuse the poor performance of money managers – Mercer included.
The assumptions of rate or return were wildly unrealistic and remain so even today. Thus the scale of the problem is perhaps even bigger than has been acknowledged. We are certain to remain in a near-ZIRP environment for a substantial part of the next decade so returns will remain poor and the deficits will get even bigger. Add to that the prioritisation of those already retired, the outlook for those in their 30s and 40s is quite bad – not quite as bad as those without pension provision but pretty bad. IMHO they are unlikely to get back what they have contributed in real terms.

His plan on not tying people to annuities is a good one. As it is money managers thus take a cut on both ends. It is a great business. heads i win, tails you lose. I also do not understand his complaint over German bund yields. If yields are low, it means the bunds pension managers carry have risen in value. So what’s to complain? And given the high yields on Irish government bonds, this is a clear case for reallocating capital from German to Irish equities. Unless of course, most pension funds already have too many Irish bonds!

unintended consequences: how do you make retirement optional without moving towards a situation where it becomes unviable for the majority to retire?


annuities are currently being priced off bunds, hence PV of pension fund liabilities has risen. Proposal is to price off Irish long bonds, so liability falls. If pensioners were to receive cash, would they buy Irish or German?

Changes in funding rules do not create or destroy ‘wealth’, they just re-assign etitlements/obligations. Some reassignments (eg against bust or weak companies) may not be feasible. Even where feasible, the community of [pensioners + shareholders] is playing a zero-sum game.

@ John

currently annuities are based off German Bunds – so therefore (a) very expensive and (b) pension funds have to put up a lot of capital to buy these expensive annuities in the event of wind-up.

By re-basing to Irish govvies (a) they are much high yielding so annuity is cheaper and (b) improves solvency situation of pension funds.

This would also encourage/allow Irish life insurers to buy more Irish govvies to form the asset base for the annuities.

I can fill in some of the background on the first paragraph. There is a real and a technical/regulatory dimension to this.

What is being proposed is allowing pensioners (buying annuities) to have them backed by Irish government debt, rather than Euro sovereign debt. Pensioners will obtain access a higher yield, so each, say, €100 in annual pension will cost less in discounted capital.

But, and it is a pretty big but, they are accepting the higher implied (actual?) risk of default of the Irish Government. In effect, the pensioner isn’t really any better off in terms of expected wealth. There is an apparent monetary benefit, but as I mention, expected wealth is indeed no different unless you have some way of knowing that the market is definately pricing too much risk of default into Irish government bonds.

The technical points comes from the fact that the regulatory funding requirement for DB pension schemes in Ireland is driven by annuity prices. Each current or future pension needs to be valued at mark to market annuity prices and discounted at higher risk-based rate for those pensions not yet in payment. Since these are based on risk free long term rates (as they should be) the cost at the moment (2.3% for Gemeran bonds) is prohibitive. So this regluatory valuation of liabilities, as required under the Funding Standard, is currently very high and asset values have fallen. Companies are required to inject (what are some large amount of) cash into these schemes over a maximum of 10 years, but in some permitted cases longer, in order to restore funding to the “Minimum Funding Standard”.

If the annuity rates used in this Funding Standard calculation could be calculated at 6% rather than 3% (which could happen if the governemnt allowed such an Irish Sovereign Annuity” program), then the regulatory liabilities would be lower, the funding gap would be lower and the cash demand on companies would be lower.

What would be higher? The chance that this blows up after pension scheme base their balance sheets on and scale up their investment in Irish Government bonds and default does indeed occur.

Just something that needs to be kept in mind. The EC, as always has something to say on these matters:


“the minimum amount of the technical provisions shall be calculated by a sufficiently prudent actuarial valuation, taking account of all commitments for benefits and for contributions in accordance with the pension arrangements of the institution. It must be sufficient both for pensions and benefits already in payment to beneficiaries to continue to be paid, and to reflect the commitments which arise out of members’ accrued pension rights. The economic and actuarial assumptions chosen for the valuation of the liabilities shall also be chosen prudently taking account, if applicable, of an appropriate margin for adverse deviation;”


“The Commission shall propose any necessary measures to prevent possible distortions caused by different levels of interest rates and to protect the interest of beneficiaries and members of any scheme.”

I am drawn to phrases such as “sufficient for payments to continue to be made”, “an appropriate measure for adverse deviation” and “prevent possible distortions caused by different interest rates”.

The Commission could jump in pretty quick smart if they didn;t like the idea of Irish pensioners being bundled completely into Irish government bonds withoutt heir knowledge or consent.

Does anyone feel that sometimes Irish pension funds under management are treated as institutional war-chests?

Exhibit 1: From the Sunday Tribune – “Bank of Ireland Asset Management (BIAM) and Irish Life Investment Managers (ILIM) are expected to be the anchor investors in a €400m share placing which will form part of the €3.3bn capital raising plan Bank of Ireland is announcing this week, industry sources said. The two domestic institutions, which currently hold 2.53% of the bank’s shares each, are among a handful of big names prepared to divide the placing “into substantial chunks” ahead of a €1bn rights issue, according to one institutional source.” (I don’t know if they actually went through with this).

Exhibit 2: From the Sunday Business Post – “AIB has used money from its staff pension fund to buy a stake in the controversial special purpose vehicle (SPV) set up to own the toxic loans acquired by the National Asset Management Agency (Nama).
In a regulatory filing, the bank revealed it had raided the staff pension fund for €12 million of the €17 million it was required to contribute in return for a shareholding in the SPV.”

@ Eoin,

“This would also encourage/allow Irish life insurers to buy more Irish govvies to form the asset base for the annuities.”

You risk becoming predictable 😉

This links to Brian Woods II’s observations about what level of default risk is acceptable to an annuity holder. Shifting to Irish government bonds may be good for pension fund solvency but would currently expose the annuity writer and annuity holder to enhanced default risk . Will Irish insurance companies base their annuity pricing on Irish bonds if the difficult to accurately estimate default risk is judged significantly higher than the likelihood of Bund default ? Can Irish bonds be said to be risk free ? What happens to annuity holders in the event of a default ? Will pension funds compensate them?


Many thanks for that.   On your first point, you make a valid observation about the risk premium.  But isn’t it the provider of the annuity that is affected?  The pensioner’s benefit would be based on the rules DB formula.   Allowing investment in higher yielding bonds makes it “cheaper” to back the annuity.    As you say, the higher return is just offsetting the higher risk borne by the annuity provider.   

The point about the increased risk to the pensioner of an underfunded scheme blowing up is well made.  Shane Whelan has written insightfully on this.   I have written to him asking him to weigh in. 

It would be good to hear your reaction to the risk premium point.   It reminds me a bit of the US debate about the investment the social security trust fund in equities.   It was made out to be an easy way to boost returns.   But those returns only came because of the higher risk. 

On the zero sum game, I don’t think you can completely dismiss the distributional implications of changes to the rules.   Someone depending on their DB pension to support them in retirement would not be so disinterested.    I think it is important to put a spotlight on the implications of any changes.


I am not ‘dismissing’ distributional changes, just making the point that aggregate net wealth cannot be created by actuaries, more’s the pity.


the Commission could say ‘you must price annuities off AAA sovvies’, ie bunds? How many solvent DB schemes would be left then?


As far as I know, the provider of the annuity (a lifeco.) will simply be required to meet the regulatory capital requirements as direct by the Financial Regulator. Should this shift to “Irish Sovereign Annuity Fund” and should the debtor underlying the “Irish Sovereign Annuity Fund” default (i.e. the Irish Government) I am assuming that it will be the pensioner that loses (along with the lifeco. shareholders I presume). All manner of scenarios can be spun from there though.

Anglo, AIB and BoI guarranteed and recapped today, Irish Life, Bank of Ireland Life etc. guaranteed and recapped tomorrow in orer to save in this case pensioners rathher than savers/bondhoders.

@ Colm

Currently it is marketted annuities, which I think pretty much means German issued sovereigns in practice.

My reading of the relevant Commission Directive does indicate to me that the Commission could “in the interests of protecting beneficiaries” tell the Irish government that they can not play such a game that shifts the interest rate applied to calculate the technical reserves (Funding Standard) to a more favourable higher and risk laden rate.

The relvant part of the directive seems to allow the Commission to argue that such use of Irish Sovereigns represents a “distortion” caused by different levels of interest rates.

Roughly 80% of DB schemes in Ireland currently fail the MFS test

Just a few random comments

1. DB benefits are not guaranteed and solvent sponsors can wind up a scheme in deficit at any time (witness SR Technics). Most sponsors however are under a moral obligation to deliver the benefits and do their level best to meet these promises. Trustees of DB schemes can only demand so much of their sponsors who in the main are supportive of euity investment as they get a nice P&L benefit from it per current accounting rules. Financial economics hasn’t infiltrated DB pensions in any great shape or form as yet.
2. The MFS test is a weak hurdle unless the scheme is very mature (ie predominantly pensioners). It is a non-market related test and is therefore impossible to hedge. Therefore sensible “economic” decisions are often not taken due to the implications of Pensions regulation.
3. The sovereign annuity idea effectively results in a transfer of risk from non pensioners to pensioners but potentially results in a more equitable share of the pot. In theory pensioners backed by German bunds have a greater degree of security than the semi state PAYG system. It is intended that a “default” clause is included to allow the Insurer reduce pension payments in the event of default
4. Under the current system a retired former CEO aged say 55 could walk away with most of the pot of assets in a wind up scenario and the active employees (who could be close to age 65 in a lot of cases) may get nothing.
5. There has been a long standing DB industry practice (not just in Ireland – same in UK, US etc) to invest heavily in risky assets as a result of Actuaries taking advance credit for risky asset growth in irregular (triennial) funding valuations. Assuming you’ll get 7%pa on equities (say) leads to lower immediate cash funding requirements than if the assets were all invested in safer assets.
6. The traditional belief was that pension investment was a long term game. The advent of including pensions in company accounts and increased regulation roughly a decade ago has shown up the folly of this and the volatility of a mismatched investment strategy is only too apparent
7. The regulator in Ireland has little power in enforcing change and instead has to walk a tightrope between trying to enforce better funding and investment strategies without causing a raft of wind ups in deficit. Not an easy balancing act in the current environment.
8. The UK has many more layers of protection – debt on the employer, Pension Protection Fund, a Regulator with significant powers. Ireland is a long way short of this.

The key point is the current system is under severe strain at the minute due to the over reliance on equities to deliver generous benefits. Overpromising and under delivering – sound familiar?! Contribution holidays and benefit improvements (to probably unsustainable levels) was the norm in the 90s and with the severe equity crashes in the 2000s most sponsors now find themselves with significant legacy issues.

The ideal solution would be a move to an insurance based arrangement where capital reserves are required for risky investment – sounds reasonably sensible doesn’t it? However that would require significant legislative change and the appetite is not there for that at the minute. However imminent changes to things like the accounting rules may mean we get to this point eventually – if all the schemes aren’t wound up in the meantime.

Those of you with private DB pension provision may want to take a closer look at your benefits and how secure they really are…

Those with Undefined Benefit schemes, or DC as they’re called (or with AVCs), may want to be offered the choice between an annuity based on German 15 year yields of 2.51% or on Irish government bond yields of 7% in the same maturity. Annuitants might welcome being treated as mature and sentient adults. God knows, they’ve taken enough (involuntary) manager risk over the last few decades with the corpus of their savings – it might be an idea to let the owner of the funds roll the dice, for once.

The company writing your annuity reserves the right to invest in the higher returning bond, while offering you the lower return under the guise of statutory obligation. Investment in the lower yielding bond merely ensures their certainty of profit margin over the annuity’s life, to the benefit of two parties – the life company, and the German government which has access to distortedly cheap funding. (Is “distortedly” a word?)

First thinking through pension proposals from an investment perspective can be fruitful. For the first proposal- The Irish Sovereign Annuity- this means that someone will purchase Irish bonds to back the annuity. Whether it is the pensioner or a life company that takes the investment risk is a moot point. The key is that pensions in payment will now be dependent on the flow of income generated by an Irish bond rather than a German Bund. When clients ask me about the enticing attractions of the higher yield on Irish bonds, I usually ask them why they don’t invest in Greek bonds – the yield is even higher again. Of course the answer is that Greece is a basket case, the risk is too high …. !
Surely, if there is one big lesson to learn from the crash is the importance of diversification. Therefore, this proposal should be dismissed.
The second proposal makes a lot of sense:
–linking actuarial calculations to a weighted average Eurozone bond yield is better than just the German Bund(and French OAT).
–the current system of forcing pension members to effectively purchase the German Bund at very low yields is daft, and members should have the option of a lump sum invested in an appropriately diversified portfolio.
A final point – much of the confusion surrounding the pensions debate is caused by the peculiar nature of the defined benefit pension scheme. Pension funding is simply about building up a pool of assets that can ultimately generate an income. The DB structure turns things around by first promising what that the future income will be, and then trying to work out a contribution and investment plan. The implosion of DB schemes is due to the fact that these promises were simply too generous, and consequently the investment strategies adopted were far too risky.


DC pensions are unlikely to be widely successful as a national strategy because (1) people don’t like to invest their life savings without knowing what they will get when they retire and because (2) the pensions industry performance has been abysmal and lacks credibility.

That is partly due to the absence of suitable investments, such as inflation-linked long-term government bonds which give relatively modest but guaranteed results (~ spread over several states if you wish to diversify the risk). My understanding is that such bonds are also a relatively cheap way for governments to raise long-term debt, much of it domestically.

Inflation-linked bonds do not require a re-invention of the wheel. There is experience in France, Canada, US (and on a small scale) UK.

@Brian – “Whether it is the pensioner or a life company that takes the investment risk is a moot point.”

No it’s not. The pensioner should not be infantilised, and should be offered the choice.

Further, the pensioner through their working life has effectively, especially in DC-land, already taken a huge risk by appointing an overcharging, underperforming fund manager (if you can show me an exception I’ll be pleasantly surprised). The range of underperformance by the worst performing fund managers in Ireland over the last 5, 10, 20 or 30 years, versus even the average, is frightening. The average versus market performances, particularly by those on active fee levels, has been little short of appalling.

If your client had taken out an annuity base on Greek yields they would, to date, have been far better off than taking one based on bund yields. The life company wouldn’t, for silly accounting reasons.

I’m a bit late replying to these comments but here goes:
Inflation linked bonds would be great if you could get them at a sensible yield. The current real yield on index linkers in US,UK etc is less than 1% and some of them are priced to give a negative real yield(UK 2.5% 2016!). I know we live in strange times but generally index linkers pay very low real yields.

My point on whether its the life company or the pensioner that takes the risk is that if the underlying asset on which the payments are made defaults, then there’s simply no money in the pot. Let’s say Irish life companies sell annuitiies backed by Greek bonds. Rates are great but Greece defaults; the life company can’t meet its obligations and goes bust leaving the pensioners out of pocket, exactly the same result as if pensioners held the bonds directly themselves.

With many company DB schemes underfunded and in desperate trouble has anyone investigated the possibility of the DB pension scheme being initially mis-sold ? I draw a parallel here to the mis-selling of endownment type mortgages in the UK which was realised in the last 10-20 yrs. I beleive many insurance companies were proven to not adequately to clients that there were risks involved with endownment mortgages and also not sufficiently checking out the buyers views and attitude to risk. Could it not be possible that DB pension schemes in Ireland were equally sold without sufficient explanations concerning the risks and lack of guarantees to buyers benefits ?

Comments are closed.