PensionReforms' summary and comments
There are three different groups of people who need to know whether a Defined Benefit (DB) scheme is 'properly' funded: first, there are the members and pensioners.  Will the scheme deliver on its promises?  Is the combination of current assets, future contributions and investment income likely to provide enough to meet the benefit outgo until the last present beneficiary retires or dies?

Next there is the sponsoring employer - or, more particularly, the employer's owner(s) and their creditors.  The employer underwrites the DB scheme's current and expected obligations.  Are they a burden too far?  Is the amount paid this year a fair reflection of the obligation that accrued this year so that this year's accounts are a 'true and fair' view?  What surprises might the scheme deliver in future years?

Finally, there are the tax and regulatory authorities.  DB schemes usually attract significant tax concessions on contributions so the revenue authorities need to ensure they are not abused.  Securities regulators also need to supervise the DB scheme's operations from a disclosure perspective.  Are members being told what they need to know?  Where there is some form of insurance scheme (such as the US's Pension Benefit Guarantee Corporation), taxpayers should also be very interested in their financial health.

Investment markets have been quite unkind to both DC and DB schemes, even to the time the report was published (2007).  In a DC scheme, members shoulder that economic damage but sponsoring employers bear the investment risk in a DB scheme.  Markets have certainly worsened since 2007.  Not only that but mortality has been improving particularly in recent years.  As a society, we should celebrate that but it makes DB pensions more expensive.

In the meantime, accountants have been generally 'tightening' reporting standards:
"The trend towards market-based valuation of pension plan liabilities is in general a welcome development as it may offer a more realistic picture of the solvency position of DB pension funds and improve the transparency and international comparability of company accounts."

The DB schemes' actuaries also have different ways of measuring the way the benefit liabilities are calculated and the discount rates applied to future obligations:
"Differences in actuarial cost methods are likely to remain, however, as pension regulators and accountants often take different perspectives when valuing pension liabilities.  In particular, pension regulators are increasingly concerned about the termination value of benefits (ABO), tend to make more conservative assumptions about vesting and withdrawal rates and use lower discount rates."

Not only that but there can be differences in economic and demographic assumptions; even differences in the way the ".accounting standard for pension liabilities (IAS19) . has been implemented in most OECD countries."  

In the past, 'smoothing' techniques allowed actuaries to even out what they thought were temporary fluctuations.  That has now changed, not necessarily for the better.

"A move to fair value accounting, where smoothing or amortisation of actuarial liabilities over time is no longer possible, therefore risks misrepresenting the long-term economic cost of pensions as an ongoing concern.  It can also lead to sub-optimal asset allocation decisions and have negative implications for financial stability as employers and pension funds overreact to short term changes in asset values and engage in procyclical investment behaviour."

While a 'mark to market' approach seems attractive, what really is the 'market value' of assets that aren't saleable or where there is no intention to sell, now or even ever?  It all seems rather difficult and the report suggests that the sponsors of DB schemes may prefer to walk away from them:
"Ultimately, however, many sponsors may prefer to move away from defined benefit plans that expose them to significant risks without any upside potential, as surpluses cannot normally be recovered by the sponsoring employer, and when they can they are often subject to heavy taxes."

We have already seen a shift to Defined Contribution schemes in many of the countries covered by the report's survey.  The members bear the investment risk in those cases.  However, "[n]ew pension arrangements, like cash balance and other hybrid pension plans, can encourage more meaningful risk sharing, where, in particular, the cost of anticipated increases in life expectancy are borne by each generation of workers."

PensionReforms notes that the issue of liability for the investment risk is more nuanced in the case of Tier 2 schemes where the state bears an underpinning promise of a minimum income at Tier 1.  To the extent that the Tier 1 pension is higher because of losses suffered at Tier 2, taxpayers carry that part of the investment risk.  The same applies where there is a full-blown income/asset test on Tier 1, such as in Australia.

PensionReforms answers the question posed in the report's title, by suggesting that DB schemes are probably not safer especially as a result of what has happened in the last 2-3 years but we probably know a lot more about what's really happening.  However, PensionReforms thinks there is a bigger issue here - whether employers should have DB schemes, regardless of the financing issues.  It really matters that employers know how much they are paying their employees in total this year and next.  Where there is a DB occupational scheme, employers should run actuarial numbers for each current employee in that employee's own circumstances.  They will probably be surprised at the total and may wonder whether they really need to pay as much to get that job done.  What they will also discover is that DB promises distort total remuneration.

Two otherwise equivalent employees (same job, same direct pay) will have widely different total remuneration, including the value of the DB promise: an older employee will have more than a younger one; a female will have more than a male; an employee with more rapid pay increases will have more than one who has plateaued and, in the presence of an automatic survivor pension, the partnered will have more than the unpartnered.  And what about those who haven't joined?  Are they compensated for the missing pension-based remuneration?

PensionReforms suggests that employers should instead focus on the total and let employees decide whether, how and how much to save for retirement.  That means getting rid of future DB promises, not because they can be costly and uncertain but because they are distortionary and unfair.  Moving to DC arrangements will also simplify matters for employers, their owners as well as tax and regulatory authorities.  (File size 324 KB; 31 pp) 324