Elegant Letter

2 1
P E R T H P H 1 1 R G
6 September 2010
International Accounting Standards Board
30 Cannon Street
United Kingdom
Dear Sirs
I write to provide comments on the Exposure Draft ED/2010/3, “Defined Benefit Plans, Proposed
amendments to IAS 19”.
Personal background
I have been a professional trustee of occupational pension schemes in the UK for 23 years. I am
current chairman of the industry-wide Railways Pension Scheme and a former DTI-appointed
trustee of the Mineworkers’ Pension Scheme.
I am generally opposed to the proposed amendments to IAS 19 and in particular the change from
using an expected return on plan assets to applying a net interest discount rate on the net defined
Overall response to the exposure draft
Not only can I not see reasons for change at this time but I would go further and say that such
changes as are proposed will further distance the artificial financial accounting reporting from the
reality of the legal and commercial relationships that have been developed between fiduciaries of
defined benefit pension schemes and their sponsoring business employers. I live today in a world
of scheme-specific pensions regulation, not one attempting uniform, one-size-fits-all which is,
frankly, not fit for purpose.
The logic, if indeed IASB members have any, appears to be that by invoking the relative idea of a
“high quality corporate bond” this infers something objective and relevant that is somehow better
than using a scheme-specific “estimate” of the scheme’s expected returns, which have been
reported on by actuaries and others qualified to do so. Financial reporting includes many, many
instances where judgment and/or estimates must be used.
Corporate bonds are a relatively recent phenomenon, but even so it has been apparent in recent
years that the mis-classification of some bonds as being of a high quality, based on rating
agencies’ opinions, has been only in relative terms to other bonds, has been subject to market
manipulation, varying levels of liquidity and, consequently, varying levels of volatility, with such
mark-to-market accounting often bearing no stable relationship to the underlying quality,
recoverability or expected cash yield of the said bonds.
The proposed change in financial reporting also does not seem to grasp that many defined benefit
schemes can and do achieve their expected returns over reasonable funding periods, although of
course never in any one particular year, nor should that be expected. While the proposed change
may be said by you to target a perceived abuse by those preparers of financial statements who
Derek Scott BAcc (Hons) CA
Linda M. Scott BAcc CA
are considered consistently to overstate their expected returns, it would do so at the expense to
those many schemes and their hard-working fiduciaries who properly attempt to reflect the
expected investment return characteristics and the agreed terms of the defined benefits payment
contract entered into with the employers and their employees, current and former.
I believe you would do better to concentrate both the preparers of financial statements and their
auditors on whether the expected returns chosen by particular companies are consistent with two
levels of evidence: backward-looking in terms of historical achievements, and forward-looking in
terms of prudent estimates of what should be achievable from the balance sheet date,
notwithstanding general uncertainty and liquidity-based volatility in the investment markets.
Accounting for investments within defined benefit schemes is an area that many users of financial
statements seem to find difficult to understand. I would hope that, over time, users have improved
their understanding of IFRS and IAS 19, in particular, and are better placed to interpret financial
statements prepared in accordance with IAS 19. Given there is no single, theoretically correct
answer on how to account for defined benefit plans, however, it is my view that the
understandability of defined benefit plan accounting and the objectivity derived through the
relevance and consistency (rather than an objectivity derived through attempted and misplaced
uniformity) of accounting over time should be the prominent considerations.
My greatest objection is to your proposal to eliminate the existing IAS 19 concept of the expected
return on plan assets. No clear rationale is given as to why your proposed approach is suddenly
more correct than the existing approach, and the disadvantages, the so-called “unintended
consequences”, of making the change would appear to outweigh any advantages. Indeed, based
on my reading of the IASB’s vast and extensive literature on these proposals and also after
participating in the one-sided IASB webinar presentation run by some of your staff, your principal
reasoning would seem to me to be to stamp out a perceived abuse by some, presumably in a
minority, preparers of financial statements who intentionally overstate the expected rate of return
on assets to flatter profit. While I would accept this abuse must exist to an extent, the present
disclosure requirements of IAS 19 if properly applied should provide users of financial statements
with the opportunity to challenge entities, even if their auditors fail to do so first, where they
consider that excessive expected rates of investment return have been assumed.
A net approach to measure market-derived interest rate costs is also inconsistent and downright
inefficient compared with what many defined benefit schemes’ actually do with their investment
strategies. Defined benefit schemes invest in assets that their fiduciaries expect to deliver
sufficient quality and higher returns in order to reduce the long-term cost of the schemes which
are otherwise funded inefficiently (not to mention issues of tax relief and public affordability) by
employer and employee contributions. In many cases, the mark-to-market performance of these
assets will not reflect anything like the more stable patterns of income and other cash flows from
planned realisations to fund scheme liabilities as they fall due. These income and other cash
flows can be budgeted. Market volatility, despite claims by some to the contrary, cannot even be
modelled for any reasonable length of future time. That is in the realms of pseudoscience.
In the context of the IFRS framework, I am informed that an appropriate distinction between profit
or loss and other comprehensive income has yet to be determined. It therefore seems
inappropriate to alter the presentation of elements of defined benefit cost between profit or loss
and other comprehensive income until that distinction has been properly refined and agreed as
part of the IASB’s ongoing work on performance reporting.
I wonder from time to time why accounting standard setters do not appear to see it as being within
their remit to consider the wider public interest. Accounting standards, however, should be
neutral. In this case, even the IASB should acknowledge that the proposed changes are likely to
further discourage the provision of higher retirement benefits by employers to employees. On
your conscience be it.
My responses to some of the specific questions raised in the Exposure Draft are below, although
these should be viewed in the context of my overall questioning of the need to make further,
partial changes to IAS 19 at this time.
1. I agree with the proposal that an entity should recognise all changes in the
present value of the defined benefit obligation and in the fair or intrinsic value
(which is not necessarily the same as market value) of plan assets when they
2. I am content with a proposal that entities recognise unvested past service
cost in the period when the related plan amendment occurs.
3. The components of defined benefit cost as defined in current version of
IAS 19 are in my view appropriate.
4. I agree that any service cost component of defined benefit cost should
exclude changes in the defined benefit obligation resulting from changes in
demographic assumptions.
5. As previously stressed, I strongly disagree with the proposals for
determining the finance cost component of defined benefit cost and consider it
more appropriate to retain the existing components of expected return on assets
and interest on liabilities at this time.
6. As previously explained, I believe the expected return on assets (as
determined in accordance with the existing IAS 19) should remain a component
of profit or loss.
I am surprised to see the IASB proposal to prescribe the captions within
which individual components of defined benefit cost should be presented.
Generally and elsewhere in IFRS, income statement classification is usually not
so prescribed.
7. I see no need to alter the accounting requirements for settlements and
Questions 8 to 12
The body of standards developed by the IASB has resulted in ridiculously
lengthy financial reports that are not widely read and consequently poorly
understood by investors, even by some so-called analysts. Financial reports
should not be required to disclose what appears to be every piece of information
from an aggregated wish list of what each user of financial reports say they
would like to have.
The IASB seriously needs to reconsider how it canvasses the views of users.
Inevitably a user will state that he or she requires more, rather than less,
information. It does not follow that significant numbers of users will actually use
such information. Taken as a whole, the financial statements become far, far,
too long, difficult to understand. Important information must be lost within less
important information.
I would urge the IASB to re-initiate concerted attempts to reduce the
disclosure burden in financial statements, to make financial statements more
readable and understandable. I believe any new disclosure requirements for
defined benefit schemes should be developed with this in mind. At the same
time, standards of auditing need to be stepped up to address perceptions of
abuse in the use of expected rates of return which have no credibility by
considering both backward-looking and forward-looking evidence.
14. I agree that in many situations a defined benefit multi-employer plan does
not have a consistent and reliable basis for allocating the plan obligation, plan
asset and cost to the individual entities participating in the plan, and it may be
appropriate for individual entities to account for such a plan as a defined
contribution plan. Last man standing obligations need to be understood,
however, and if the allocation of obligations is not made at least the disclosures
should explain the ultimate underwriter(s) within multi-employer plans.
15. I believe that any changes should be applied retrospectively.
16. For the reasons explained above, I totally disagree with the IASB’s
assessment of the benefits of the proposals. While I do agree it is appropriate to
eliminate some of the options in IAS 19, as with other IFRS, I do not believe that
further fundamental changes are required or have been theoretically justified at
this time.
17. Where I would personally take further issue is with the use of "fair value"
within accounting standards. Fair values, which are typically equated to "efficient
market" values, are used to justify mark-to-market accounting. IASB seems to
assume that markets are "efficient" and can therefore be used to obtain "fair
values" as if that is a sign of their objectivity, of their strength even as a basis of
financial reporting.
But if the assets I hold as a fiduciary are indeed "fairly valued" then I should
probably have sold them or be selling them as they have achieved what they are
most likely to achieve. I should of course to be "really efficient" then reinvest the
proceeds in "undervalued" assets for future income and profit. But accounting
standard setters seem blind to recognise such possibilities within their treatment
of long-term monetary investments. (They seem to allow for it in the
capitalisation of acquisition prices, including goodwill, for future income streams
expected from subsidiaries or associate company investments, and also with
other fixed assets.)
I also think pension scheme fiduciaries and scheme sponsors who believe the
asset allocation is the paramount consideration are playing unwittingly into the
hands of accounting standard setters like IASB. Standard setters demonstrate a
poor understanding of the operation of liquid markets, of the pro-cyclical nature
of their measurement bias, and the real components of long-term investment
returns. But such a one-dimensional analysis of a pension scheme's financial
position in financial reporting by companies (and by fiduciary trustees) leads
inevitably to mark-to-market measurement, placing uniformity of market prices
above relevance and objectivity.
By focussing mainly/exclusively on the balance sheet structure at a point in time,
rather than the cash flows and realisable income generation over time, the
financial reporting of pensions obligations which results is, in my view, of limited
credibility and limited overall usefulness.
Finally, I also think what results fails any reasonable interpretation of "a true and
fair view". But of course the accounting standard setters seem to have fixed the
"true and fair view" nowadays to mean something entirely different, something
which merely conforms ("is in accordance") with UK or EU Accounting
Standards. That approach seems to me designed to suit the professional firms
who profit from auditing and related services, and who are able to limit their
professional liability to an expensive form of compliance-type reporting which
avoids them having to apply more in the way of judgment and forward-looking
skills, at precisely the time when those judgments and skills are needed by users
of financial statements.
Yours faithfully
Derek Scott