Memorandum submitted by The Standard Life
Assurance Company
— The Pensions Commission assumes that
charges under a NPSS should be around 0.3%, as this is the expected
long-term level of charge for the Swedish funded defined contribution
(FDC) scheme. This assumption is flawed. Some functions of the
Swedish FDC scheme are heavily subsidised by the Swedish taxpayer
and, therefore, are not covered by the 0.3% charge.
— Using the Pensions Commission's own
figures, a flat single annual management charge of 0.3% requires
whoever runs the NPSS to accumulate losses for the first 30 years,
with debts peaking at over £3.5 billion. If the government
run the NPSS then MPs need to be aware of this long-term funding
commitment. If the private sector puts up this capital then it
may need to make significant additional interest charges to reflect
the risks inherent in this project and, in particular, the long
payback period.
— The Regulatory Update 64 rule should
be dropped. There is no need to replace this with a similar rule
in the run-up to a NPSS taking effect.
— Providers are unlikely to want to
serve a large part of the employer pension market over the next
four years (until the introduction of a NPSS). The current distribution
model requires payments to be made to intermediaries to establish
and develop new pension schemes. These costs are unlikely to be
recouped if charges fall to 0.3% in four year's time.
— NPSS should be bought by savers without
regard to the current Conduct of Business rules. However, advisers
should take account of the existence of a NPSS in making recommendations.
— The NPSS should fit within the new
simplified tax regime, due to take effect on 6 April 2006.
INTRODUCTION
1. Standard Life is the market leading provider
of group money purchase pension schemes in the UK. Due to the
overlapping nature of the current private pension system and proposals
made by the Pensions Commission, we are keen to participate in
the debate into the possible introduction of a National Pension
Savings Scheme (NPSS).
2. We note that the Committee is not seeking
submissions which are circulated elsewhere. However, bearing in
mind the parallel inquiry that the Work and Pensions Select Committee
is conducting, there are areas of overlap which means that some
of the content of this response is duplicated in our response
to the other inquiry.
3. There are a number of factors that will
affect the operational cost of a NPSS. The main factors are:
— The shape and level of charges levied
against savers' accounts (this is a determinant of the amount
of debt that the system would build up in the early years, the
cost of this debt being an operational expense).
— Take-up rates (a higher number of
joiners will help defray fixed costs).
— How quickly wages, and therefore,
contributions rise (again this is a determinant of debt levels
in the early years).
— How many people leave the system
each year (paid-up accounts).
— Any costs associated with enrolling
people into the scheme, and the extent to which information and
advice is required.
— The level of service provided to
savers (regular communications, access to telephone and internet
help, and so on).
— The regulatory regime imposed by
government and the Financial Services Authority.
4. A main topic of your inquiry concerns
the regulatory and distribution issues connected with the introduction
of a NPSS. In addition, you also ask how government can avoid
the difficulties encountered by Stakeholder Pensions. However,
we note your interest in the operational costs of a NPSS and we
therefore begin our response there.
THE NATIONAL
PENSIONS SAVINGS
SCHEME
5. We have several comments to make about
the Pensions Commission's recommendation to establish a NPSS.
The first of these concerns the level of charges.
NPSS CHARGE LEVELS
6. The target level of charge set by the
Pensions Commission is 0.3%. The Commission bases this expectation
upon the long-term cost of the Swedish funded defined contribution
scheme.
7. However, in Sweden's funded defined contribution
scheme (premium pension), the average charge is currently 0.64%,
despite having now received ten year's worth of contributions.
The charge on the default fund is 0.37%—still above the
Commission's target (1). The Swedes do not expect average charges
to fall to 0.3% until 2020 (by which time, 25 years worth of contributions
will have been invested).
8. Another important fact ignored by the
Pensions Commission is that the account charge in Sweden only
covers part of the cost. The government agency that operates the
premium pension (the Premium Pension Authority or PPM) has a large
part of its expenses met by central government. Specifically,
the PPM shares the cost of collecting contributions with the income
pension (Sweden's main state pension scheme). It also shares the
cost of communicating with savers—including the Swede's
famous "orange envelope" that is sent out annually.
9. The cost of these shared services amounted
to as much as SEK 426 million in 2005 (1). But to date, the PPM
has paid nothing towards these shared services. That means the
PPM's costs have been directly subsidised by the Swedish taxpayer
for the last 10 years. If the PPM were to share these costs equally
with the income pension then the PPM's share of these shared costs
could be as much as SEK 213 million a year in 2005 terms.
10. If shared costs in previous years were
of a similar size (for example, SEK 454 total in 2004), the Swedish
taxpayer has already provided subsidies to the PPM of over SEK
2 billion in 2005 terms. Actual declared debts of the PPM amount
to SEK 1.86 billion (1). Had these shared expenses been borne
by PPM in previous years, then outstanding debts would more than
double the declared level. Historic interest costs on this debt
would also have been much higher, adding to the accumulated deficit.
11. To put the extra ongoing cost in perspective,
the declared operating costs of the PPM (the amount on which account
charges is based) were only SEK 220 million in 2004 (1). This
means that the PPM's costs would almost double if it were to pay
a half share of shared expenses.
12. Yet even when the PPM starts paying
a share in 2005, it will only pay a share in proportion to the
size of the contributions collected relative to the whole. As
the income pension has compulsory contributions of 16% and the
premium pension just 2.5%, the proportion of shared costs it will
meet will only be 2.5/18.5ths.
13. If such a system were put in place in
the UK (the Pension Commission's proposed NPSS is very similar
to the Swedish premium pension), it is far from clear whether
it would be able to share the costs of collecting premiums with
the basic state pension. Her Majesty's Revenue and Customs has
reportedly ruled out this option because the IT infrastructure
would not be able to cope.
14. In that case, a UK version of the premium
pension would have to pay for the full costs of collecting contributions,
unless our government chose to subsidise the account charges with
taxpayers' money.
15. If, in Swedish PPM terms, operating
expenses of SEK 220 million are equivalent to a target charge
0.3% (after 25 years of operation), then the target charge in
2020 would have to be significantly higher if annual operating
expense were instead SEK 646 million in 2005 terms (SEK 220 million
plus SEK 426 million).
16. The Pensions Commission's assumption
that 0.3% is an achievable 25-year target based upon the experience
of the Swedish funded defined contribution system is seriously
flawed for the reasons given above.
17. If the Government is to proceed with
its own scheme then taxpayers ought to be aware of the likely
costs of such a scheme and that taxes will be used to subsidise
the annual management charges as happens in Sweden.
18. We recommend that the Committee carry
out its own investigation to establish the facts. Specifically,
whether and when an annual management charge of 0.3% is attainable.
NPSS CHARGE STRUCTURE
19. Flat annual management charges such
as those proposed by the Pensions Commission do not cover the
substantial investment required in the early years of any new
system. Investment is required to establish the IT and other administrative
infrastructure. However, the major capital requirement comes from
subsidising charges in the early years.
20. Standard Life has calculated that, based
upon the Pensions Commission's own cost and coverage estimates,
the NPSS would not reach break-even until its 30th year. This
analysis makes the following assumptions:
— That six million people save via
the NPSS (Pensions Commission estimate).
— Charges of 0.3% (Pensions Commission
estimate).
— IT and administrative systems establishment
£500 million (Pensions Commission estimate).
— Account set-up expense £90 per
member (Pensions Commission estimate).
— Account maintenance cost £25
per annum (Pensions Commission estimate).
— Paid-up account maintenance cost
£19.50 (Pensions Commission estimate).
— Average annual amount of saving per
member £1,000 (Pensions Commission estimate).
— People retire from the NPSS based
upon a 40-year cycle, ie at the rate of 150,000 per annum (Standard
Life estimate).
— Termination expenses are the same
as account set-up charges of £90 per retiring member (Standard
Life estimate).
— Members retiring from the NPSS take
their funds with them (for example, they buy annuities) and the
0.3% charge on these funds stops.
— People join the NPSS (eg school/university
leavers, immigrants) at the rate of 150,000 per annum.
— Interest on NPSS debts is charged
at a rate of 4% (Standard Life assumption based upon long-term
gilt redemption yields).
— Funds grow at a nominal rate of 6%
per annum, before charges (Standard Life estimate).
— 200,000 people (3.3% of members)
per annum leave the NPSS before retirement and their account becomes
paid-up. For example, they emigrate or join a better quality scheme
that has opted-out of the NPSS (Standard Life estimate).
21. On the basis of this analysis, the NPSS
will run up debts of £3.5 billion by year 18 (assuming cost
and contribution inflation of 4% a year) or £3.1 billion
in year 16 (assuming cost and contribution inflation of 2.5% a
year). See Appendix 1 which illustrates the net capital requirements
under each of these projections.
22. This raises two important financial
questions for members of parliament, if they are asked to support
plans for a government-run NPSS. Firstly, are they prepared to
sanction a new national savings scheme that will require the country
to borrow to subsidise charges on what are effectively private
pension savings? Note that the bulk of this borrowing could not
be regarded as an investment in the infrastructure of the UK.
Secondly, are MPs prepared to sanction a NPSS that will not break
even until 2035 on optimistic assumptions? MPs need to make a
judgement whether this policy will endure long enough to break
even.
23. If government decides that an NPSS-style
savings scheme is desirable, then the only other alternative would
be to ask the private sector to run the scheme instead.
24. However, the economic and structural
costs of an NPSS-style alternative, at a flat annual management
charge of 0.3% per annum, are little different for the financial
services industry than they are for government. The estimated
costs of running such a scheme, and laid out in Appendix F of
the Pensions Commission's Second Report, were derived from discussions
that the Pensions Commission had with the industry.
25. Indeed, the cost of running the scheme
(before any profit margin is taken into account) is likely to
be higher for the private sector than it would be for government.
There are two reasons for this:
i. In addition to supplying capital to finance
start-up costs and subsidise charges until break-even point, the
industry must also hold capital reserves (solvency margin capital).
ii. The cost of capital to the financial
services industry is higher than the cost of capital for government.
If financial services companies raise equity from their shareholders,
those shareholders will typically be expecting a gross return
on that capital of between 10% and 15%; substantially higher than
the rate the government can borrow money at. In the case of this
particular project, shareholders are likely to demand an even
higher return, due to the added risk of a long payback period
and the fact that no previous pension reform has lasted much beyond
10 years.
26. Alternative charging structures would
reduce this capital strain. One example would be a mixture of
a regular monthly administration fee of £2 per month and
an annual management charge of 0.3%.
27. Based upon this charging structure,
the projected net capital requirement is set out in Appendix 2.
This reduces the peak capital requirement to just over £1.1
billion in year four, with break-even reached in year 14.
28. The effect of this style of charging
structure on the saver is minimal. Appendix 3 shows that the net
loss of retirement savings is 6.63% of a hypothetical charge-free
fund for a typical NPSS saver (2) at a flat charge of 0.3%.
29. However, when a £2 monthly administration
fee is added, the net loss increases to only 8.87% of a hypothetical
charge-free fund.
30. To put this into perspective, the net
loss of fund resulting from current stakeholder charges (1.5%
for the first 10 years and 1% thereafter) is 20.76% of a hypothetical
charge-free fund.
31. Appendix 3 also shows the net loss of
charge-free fund for other selected charging structures.
32. In cash terms, the fund value at retirement
(in today's money and assuming future price inflation of 2.5%)
for a typical NPSS saver (2) is £98,654 with a £2 per
month administration fee added, as against £101,088 with
the only charge being a 0.3% annual management charge. The hypothetical
charge-free fund is £108,261.
33. Based upon current stakeholder charges,
the today's money value of estimated retirement savings is only
£85,791. Note that although stakeholder charges are higher
than the charging structures mentioned above (and those laid out
in Appendix 3), these new structures are only made possible by
changes to state pension policy and changes to the regulatory
and legislative environment.
34. At current annuity rates, these funds
would generate monthly incomes of:
— £528 (0.3% annual management
charge only throughout);
— £515 (0.3% annual management
charge throughout plus a £2 per month administration charge);
and
— £448 per month (current stakeholder
charges).
35. If the government does want the financial
services industry to run an NPSS-style scheme, then industry and
government must be willing partners. The industry is not seeking
to make excessive profits from running such a scheme and has suggested
that this is overseen by an independent regulator.
36. We recommend that the Committee explore
alternative charging structures to that proposed and the suggestion
of an independent regulator whose job, amongst other things, would
be to ensure that savers get a fair deal.
NPSS ALTERNATIVE—WHO
SHOULD RUN
IT?
37. In the insurance industry's favour is
the fact that much of the initial capital cost required to pay
for IT and other administrative infrastructure has already been
incurred. The incremental cost to tailor administrative systems
to the needs of an NPSS-style alternative would, therefore, be
relatively light.
38. Other considerations are that the insurance
industry has the experience necessary to administer a national
scheme, for example, we have the systems and processes necessary
to:
— communicate effectively with scheme
members (via telephone, internet and post);
— collect contributions from employers
and reconcile the amount of those contributions;
— report contribution discrepancies
to The Pensions Regulator;
— account for investments.
39. Importantly, the insurance industry
has the capacity to handle the expected volumes.
40. We recommend that the Committee explore
these issues with each of the parties that has tabled a bid to
run an alternative to the NPSS.
REGULATION—ADVICE
41. At the level of charge required by the
Pensions Commission, NPSS pensions will be bought (by default)
rather than sold. There is simply no margin available to pay for
any distribution cost, whether that is persuasion or full regulated
advice.
42. Advisers do need to be aware of a NPSS
and any recommendations they give should take into account its
existence. To this extent, the FSA should recognise NPSS in its
Conduct of Business rules.
43. However, the FSA need not continue with
the Regulatory Update 64 (RU64) rule that required the recommendation
of non-stakeholder pensions to be justified as to their superiority
over stakeholder pensions. Nor should they extend this rule to
cover a NPSS.
44. There are two reasons for this. First,
a clear division is now apparent between non-stakeholder and stakeholder
products. This was not the case when RU64 was first introduced.
However, since then, the design of stakeholder pensions has largely
stood still, constrained by their charge cap. On the other hand,
personal pensions and group personal pensions have changed markedly,
such that they now offer a much wider range of funds and other
options than stakeholder pensions.
45. The second reason is that the Pensions
Commission imposes its own quality condition on schemes that opt-out
of the NPSS. These must (in charges and contributions terms at
least) be equivalent to or better than a NPSS.
46. In the interim—between now and
the advent of NPSS—we are unlikely to see any rush to sell
the current higher charge (relative to NPSS) products. In particular,
any schemes which fail to meet the opt-out conditions will necessarily
convert to the NPSS.
47. So, rather than there being a last minute
rush to sell new pension schemes, the opposite is likely to be
true. Providers are unlikely to want new schemes today that fail
to meet the standards set by a NPSS, or schemes where there was
a likelihood on the part of the employer to "level down"
to NPSS in 2010.
48. The Committee should be aware of this
issue and should consider the impact on employees and employers
if the financial services market finds it impossible to serve
any but the best quality group pension schemes over the next four
years.
REGULATION—PRUDENTIAL
49. We would expect the FSA to maintain
its current role in overseeing providers' financial soundness,
risks inherent in strategy, control mechanisms, corporate governance
and so on.
THE ROLE
OF THE
TREASURY
50. The Treasury's involvement in pensions
is largely related to setting the tax laws within which pensions
operate. In examining how the NPSS fits in, the Treasury should
pay close consideration to the new simplified tax regime that
is to take effect on 6 April 2006.
51. There will be a temptation to regard
a NPSS as being outside of the new tax regime. However, if such
a path is followed, numerous difficulties will arise in the inter-relationship
of two tax regimes. Any new NPSS should, therefore, be part of
the new tax regime that takes effect on 6 April 2006.
52. One potential side effect of the combination
of the new tax regime and a NPSS is that savings (if these are
all the private savings one has) can be "trivially commuted"
up to a value of 1% of the lifetime allowance (£18,000 in
2010-11 terms).
53. Given that a fair proportion of NPSS
savers will be first time retirement savers, it is likely that
these people will be able to commute their NPSS savings at retirement,
rather than buy a pension.
54. If they then spend their commuted fund
values, these people will fall back on the state for support.
This will have an impact on the Pensions Commission's projections
that indicated that the number of people on means-tested benefits
would fall.
55. Assuming that the lifetime allowance
increases in line with price inflation, it is unlikely that anyone
earning less than £20,000 in today's terms, and with no pension
savings other than those from a NPSS, will exceed 1% of the lifetime
allowance before 2020 (the first 10 years of the NPSS).
LESSONS FROM
STAKEHOLDER
56. The key lesson learned from stakeholder
for providers is that single charge products introduce an enormous
business risk, particularly if savings are switched to another
provider before costs can be recouped.
57. The consumer is also badly served by
the design of stakeholder. The level of the charge cap allowed
some inducements to be paid to intermediaries, but not enough
such that full advice could be given.
58. Providers rationally offered intermediaries
as much commission as they could afford within the confines of
the charge cap. However, in doing so, they made optimistic assumptions
about how long that business would remain on their books to justify
these payments. Therefore, instead of charging consumers more
for the significant risks being run in selling stakeholder pensions,
providers sold business on uneconomic terms. It is unlikely such
a situation will persist or be repeated. (3)
59. The design of the NPSS should be such
that it encourages providers to compete for savers' attention
by either providing exceptional service/flexible products/strong
investment returns and/or competes for the saver's attention by
offering the lowest charges.
60. Single charge products produce a cross-subsidy
from long-term savers to short-term savers. And, as noted earlier,
borrowing to subsidise charges may not present those consumers
with the best deal if interest on that borrowing has to be paid
back to shareholders at rates of between 10% and 15%. Therefore,
different charge designs, other than a flat annual management
charge, should be considered (see earlier discussion in paragraphs
26-36).
REFERENCES
(1) Private discussion with PPM officials
in February 2006 and PPM Annual Report 2004
(2) Typical saver based upon saving £1,000
per annum as per Pensions Commission estimates. Fund growth 6%
per annum nominal. Charges and contribution inflation 4% per annum.
Figures quoted are based upon saving £1,000 per annum over
40 years. Net present values arrived at using 2.5% future price
inflation assumption. Source of annuity rates is FSA comparative
tables, highest quote available on 27 March 2006.
(3) E.g. see "Polly put the kettle
on", Ned Cazalet, January 2006.
March 2006
| 
