P O Box 21
April 25, 2000
Mr Mervyn Bryn-Jones
Institute of Actuaries
Staple Inn Hall
LONDON WC1V 7QJ
Tel: +44(0)20 7632 2100
Fax: +44(0)20 7632 2111
Dear Mr Bryn-Jones
Subject: South African
Actuarial Practice - Financial Services Board
The Chief Actuary of the
Financial Services Board (South Africa) has responded to my letter of
April 7, 2000, addressed to you. I have inserted Mr Andrew’s response
and need to make further comments. Again, as with my first letter, I
would appreciate you passing this letter on to the President of the
Bases of Dispute:
Combining Investment Reserves with Actuarial Surplus
Mr Andrew’s statement in
paragraph 5.6 of his letter sums up the basis of our complaint against
actuarial practice in South Africa. The statement refers to combining
actuarial surplus with investment reserve as one item, simply referred
to as surplus:
“The Pension Funds
Adjudicator has found that there is no difference in principle between
various amounts that make up this total amount: all represents surplus
and the members have no right to this surplus”. ... J Andrew
Anthony Asher, professor of actuarial science and mathematics at the
University of Witwatersrand disagrees with the Adjudicator’s view. In
the March edition of The Actuary, Professor Asher states:
“ ..... His principal
complaints were that his retrenchment payout was based on a reduced
actuarial reserve. First, it did not include the investment reserve
(being the difference between the actuarial and market value of assets),
These practices are
common, although - in my opinion, for what it is worth - the reasoning
that defends them belongs to the category of propaganda; believed only
for its frequent repetition.
Where actuarial values
differ from that of the market, they cannot be traded for cash; if the
investment reserve is required to stabilise the contribution of the
employer, it is similarly necessary to preserve the reasonable benefit
expectations of the members (these are written into our Pension Act).
Payment of actuarial reserve should, therefore, be at market value.
British Actuarial Practice
differs from South African Actuarial Practice -
Ignoring definitions and terminology for the moment, the underlying
difference between British and South African actuarial practice is
determined by GN11. This Guidance Note requires transfer values to take
into account the Market Value of Assets on transfer.
South African Pension
Funds Adjudicator, The Actuarial Society of South Africa and the Chief
Actuary of the Financial Service Board say transfer values need not
take into account the Market Value of Assets on transfer.
Chief Actuary Andrew goes
one step further; he states (paragraph 1.1 of his letter) that if the
trustees want to pay more than an amount equal to the liability of the
member, the trustees may have to change the rules of the fund!
Separately, Mr Andrew has
stated that there are no definitions qualifying these various aspects of
pension calculation. According to Mr Andrew, lawyers are currently
working on developing appropriate definitions.
Note: It was discovered shortly after this letter was written that
Regulation 15 (2) (d) of the Pension Funds Act [provision is made is
Section 16 of the Act for Regulations] describes precisely what
constitutes a surplus.
If the definitions being
developed bear any resemblance to current application of South African
actuarial terminology used determine transfer values, I would urge the
Institute to use its influence to persuade South African actuaries to
reassess their stance and to bring their practice more in line with
In the absence of formal
definitions for South African actuarial practice, I would have expected
South African actuaries to use British actuarial definitions. Certainly,
the GN 26 definition for actuarial surplus is unambiguous, as is the
Funding Ratio definition.
In the absence of formal
definitions, South African actuaries should be persuaded to use common
English language interpretations. This would go a long way towards
preventing English speakers from misunderstanding the interpretations of
the language as applied by some South African actuaries.
is generally understood to mean "an amount left over when requirements
have been met".
on the other hand, generally means and amount put aside or kept back for
a later occasion or special use.
inter alia, “the amount of money or goods for which a thing can be
exchanged in the open market; purchasing power”.
the state of being liable; or what a person is liable for - debts or
Simple, but clear
South African Actuarial
Interpretation of English -
South African actuarial terminology in common use (Ac-speak) totally
ignores the standard English meanings of some words:
An “actuarial reserve”
means “actuarial or accrued liability”.
An “investment reserve”
The South African
“surplus” or “excess funding” consists of a combination of reserve and
It includes the
actuarial surplus - defined in reports as amounts above the required
100% funding requirement;
It also includes
the investment reserve, or the portion of assets set aside to
accommodate fluctuations in the stock market and which forms part of the
funding requirements to achieve a 100% funding level - as declared in
valuation reports to the Financial Services Board.
Actuarial Definitions can
change - The
definition of Funding Level, according to GN 26, means the ratio of the
Actuarial Value of Assets to the Actuarial Liability.
According to South African
Pension Funds Adjudicator, there are two definitions for Funding
The first is the
same as British definition set out in GN26;
definition, according to Murphy, is if you view the funding from a
market value of assets perspective, then the funding ratio takes into
account the market value and not actuarial value of assets.
So, if the Actuarial
Liability is $100m and the Actuarial Value placed on the Assets is $100m
and the Market Value of Assets is $120 million, then the “funding
Ratio”, according to Murphy’s Law, is 120%.
I understand that if the
market value of assets equaled the actuarial value of assets, it may be
120%, but this was not clarified in the example given by the adjudicator
in one of his judgments. This second definition is also the version
applied once the investment reserve is defined as a surplus.
Actuarial Inconsistency -
In summary, sometimes a South African surplus consists of a portion of
the value of assets forming part of the 100% funding level, and a
portion of the value of assets above the 100% funding level.
But this only becomes the
case if a member is moved from one pension fund to another. If the
member stays in a fund, the actuarial surplus remains a surplus, and the
investment reserve remains part of the actuarial value of assets set
aside or reserved to fund the members’ benefits.
Financial Services Board
to monitor Transfer Values -
In Mr Andrew’s letter, paragraph 1.4, he makes note that if the actuary
weakens assumptions in order to minimize transfer values, the Registrar
has the right to refuse to sanction the transfer.
In practice, and confirmed
by Mr Andrew, I have found no evidence of the Registrar rejecting any
transfer values where the actuarial assumptions, as reported in the
valuation reports, have been weakened.
- In the case of the
Revlon Fund, which I was party together with actuary Peter Theunissen,
to referring to the Pension Funds Adjudicator, the actuary claimed he
reduced the valuation base at the request of the client. The Registrar
approved the transfer.
In the Unilever case, pensioners were offered a 10% pension enhancement
if they agreed to their pensions being transferred to a financial
But on transfer, the post
retirement interest rate assumption was adjusted from 4.5% to 5.5%. No
explanation of the long-term disadvantage of this weakening of the
valuation base was given to the pensioners. (I now need to establish if
this is an isolated case!)
In the Sentrachem Group Pension Fund, on transfer the assumptions used
to value the liabilities were not adjusted, but the assumptions used to
value the assets were implicitly weakened.
At the time the transfer
was negotiated, the actuary placed a value on the assets of some 36
percent lower than the current market value. The actuary wrote as
“Accordingly, it is
considered appropriate to bring equities into account at 64 percent of
the market value at 1st March 1992. At the valuation date, the dividend
yield on the JSE All Share Index was approximately 3,2%. By bringing
equities into account at 64% of market value we are effectively assuming
that the underlying long term value of an equity investment is
consistent with a dividend yield of approximately 5.0%. Alternatively,
the discount in equity values at the valuation date can be considered to
be an investment reserve which can be utilized in future to smooth out
fluctuations in equity market values.
All other assets have
been brought into account at the full value at the valuation date in
view of the relatively secure nature of these investments and the
relatively low probability of capital appreciation or depreciation,
assuming there is no material change in long term interest rates.”
Shortly afterwards, when
the transfers took place, the assets were implicitly re-valued,
adjusting the market value assets equal to the actuarial value assets.
No corresponding adjustment was made to the Liabilities.
This implicit revaluing of
the assets was achieved (South African actuarial methodology) by calling
the actuarial reserve a liability and paying a market value of assets
equal to the actuarial liability. This weakening of the valuation base
left behind more than 40% of the original assets set aside for the
funding of benefits.
A 40% reduction in the
valuation assumptions, be they on the side of the liabilities or assets,
is significant. The Registrar approved the transfer.
Common South African
Practice - Many
South African pension fund transfers have followed the route of reducing
the actuarial assumptions used for valuing the assets. This is verified
by Mr Andrew in his letter (paragraph 1.1) where he describes an
actuarial reserve as the liability.
These examples do not
support Mr Andrew’s statement that weakening of the actuarial base on
transfer might not be allowed by his department.
The exact opposite applies
as Mr Andrew actively defends the right of trustees to weaken the
actuarial base on transfer. He goes one step further and states that if
the trustees wished to transfer liabilities with corresponding actuarial
reserve of assets at market value, they may have to change the rules of
Semantics of Liability
versus Reserve -
My understanding of
Professor Asher's view is that an actuarial liability is a theoretical
calculation of a promised benefit as per the rules of the fund. The
value placed on the liability is not a market value of assets; it is a
theoretical unit value or actuarial value.
Likewise, the “actuarial
value” placed on the assets is a theoretical unit value.
The only real value of
assets expressed in the equation is the market value of the assets.
Transferring the “liability”
means transferring the obligation to provide a benefit.
The transfer of the “reserve
value” is the transfer of the assets prudently set aside to fund the
obligation or liability. The value of this “reserve” is dictated by the
formula and assumptions used by the actuary to place a value on the
While not necessarily
expressed in the same way, the end result of transfer values as
described in GN11 is the same as described by Professor Asher.
Therefore, a member cannot
be paid the liability.
A member can only be
paid a market value of assets equal to the liability.
Semantics perhaps, but
failure to be precise about this interpretation has led to millions of
South African pension fund members being left far short of their
original pension funding as a result of transfers.
Actuarial Reserve versus
Accrued Liability -
In the Sentrachem case
that Mr Andrew refers to in his letter, paragraph 1 - Transfer value,
the rules of the Sentrachem Pension Fund, as with most other Funds,
refers not the transfer of the members accrued liability, but to
the transfer of the actuarial reserve value. In the Sentrachem case,
members were advised in writing that their full actuarial reserve value
in the Fund would be transferred.
On transfer, the actuarial
assumption used to value the assets of the Sentrachem Pension transfers
was implicitly reduced by more than 40% and this was approved by the
Financial Services Board.
The R650 million
Sentrachem Fund moved from a deficit position in February 1995 to a R400
million surplus position in 1998, after having transferred out the
majority of its members and pensioners.
Responsibility Funding -
For the record, the complainant in the Sentrachem dispute has referred
the case to the High Court.
Initially, the complainant
could not afford the huge expense of taking the case to the High Court.
Fortunately, however, Swedish social responsibility funding has been
allocated to fight this case.
It is sad that the Swedish
nation has to fund attempts to bring South African actuarial practice in
line with its parent body in Britain!
The Tek Case -
Mr Andrew refers to the
judgment of the Tek case where frequent reference is made to surpluses.
But the judgment does not define what a surplus is nor were the judges
asked to define what constitutes a surplus.
One must assume that the
judges took for granted that surpluses referred to were real or
actuarial surpluses. I would be very surprised if the judges were
referring to assets, originally set aside for prudent funding of
members’ liabilities, as surplus.
This judgment did not
specifically approve the stripping out of investment reserves from
transferring members and renaming these reserves surpluses. The judgment
simply uses the term “surplus” at face value.
It is this very point that
forms the founding argument in the articles written about the alleged
surpluses headed ‘R80 Billion - employers’ surpluses or members’
On the question of Mr Andrew stating that I misrepresented facts, I have
attached copies of my correspondence to Mr Andrew - Annexure 2,
and his response - Annexure 3. Mr Andrew's response must
be read in context with the question, and the reasons for the question,
set out in Annexure 2.
Essentially, I wrote to Mr
Andrew asking him to define what Sentrachem’s rule “reserve value” meant
- see paragraphs 1.5 and 2 of my letter of September 7, 1999. This
question was in respect to the interpretation of the reserve value
transferred, inter alia, on behalf of the union’s mass transfer. Mr
Andrew responded, in writing, that he did not know what the exact
interpretation of the rule was. He said only the actuary of the Fund
could interpret the rule “reserve value”.
I have not
misrepresented the facts on this issue.
Purchase of Annuities -
statement “Pensioners are not entitled in law to a proportionate share
of any investment reserve held with the fund when the fund purchases
annuities on their behalf.”
I have not seen any law
that prevents a fund from transferring the full value of reserves set
aside to fund pensions. On the contrary, British actuarial practice
requires the receiving fund to accept no less than what was provided for
in the transferring fund. Again, GN11 sets out very clearly that the
transfer value must take into account the market value of assets.
Conflict of Interest
- Mr Andrew
refers in paragraph 1.6 that the Registrar is not aware of manipulation
by trustees, acting in collusion with the actuary, to enhance the
residual surplus in the fund for the benefit of the employer.
It is highly likely that
the majority of trustees were totally unaware that their decision not to
transfer the full actuarial reserve at market value may result in huge
surpluses being created in the fund.
What is very surprising,
however, is why more trustees have not queried the sudden appearance of
huge surpluses after members have been transferred.Again, in the
Sentrachem case, R400 million surplus arose in a R650 million fund in a
period of three years. This must have been a very pleasant surprise for
some of the trustees who not only were granted improved pension
benefits, but who enjoyed increased company share values as a result of
the company pension contribution holidays.
So, while Mr Andrew says
it is unlikely that the trustees, with the support of the actuary, knew
they were creating huge surpluses to the resulting financial benefit of
themselves as shareholders, the question which must be asked is why,
after the fact, have they not queried the fairness of the decisions and
tested their impartiality and equity?
Instead many trustees, in
their roles as company executives and shareholders, have applied for
repatriation of surpluses to the companies. So many, in fact, that
special negotiations to agree on how much the companies may take out of
this windfall situation has been ongoing for some time now with Mr
Andrew at the helm of the negotiations.
If trustees are
shareholders of the guarantor companies, they will benefit financially
from these windfall profits. Whether it has been inadvertent or
deliberate actions that have relieved millions of pension members of
reserves originally set aside for them, allocating it to the financial
benefit, inter alia, of the trustees, then the matter of potential
conflict of interest must be investigated!
Mr Andrew’s comments in
paragraph 1.7, that “trustees must have acted in this way because they
believed they were entitled to do so in law and in equity”. That I
believe. But how did they come to this belief. Was it accidental, or
was it advice from their actuaries, who are Fellows of the Institute of
Actuaries and/or Faculty of Actuaries?
British Practice Unknown
in South Africa
- What I found very
surprising was in the Kransdorff versus Sentrachem case, the Adjudicator
noted the applicants evidence of GN11 ...it is a fundamental
requirement.. market value ..etc ..., but stated that his staff were
unable to confirm this practice!
The actuary relied on
British practice for all other assumptions used. The Adjudicator
affirmed these assumptions (with the exception of GN11) to be fair
because they had been approved as fair by the British judicial system.
But using only part of the recipe changes the total nature of the cake.
Vigorous Checking by
Financial Services Board -
I understand from Mr Andrew that as part of the vigorous checking of a
fund prior to approval for liquidation, the Financial Services Board
investigates the fund history up to 12 months prior to the application
This means in the case of
the Sentrachem Pension Fund, where the surplus of R400 million was
created at the direct cost to members over a period some five years ago,
the creation of the surplus will not be investigated. The transfer of
assets at 40% below the valuation has already been approved. Executives
have been granted additional benefits; the company has enjoyed
contribution holidays. Pensioners have been transferred to a financial
institution, again at a transfer value of assets well below the official
valuation of assets at the time of transfer. Pensioners have been
granted inflation adjustments some 6% below the going inflation rate
while R400 million surplus is held in the fund.
Mr Kransdorff is one of
many bemused complainants who has reported the trustees for mal
administration to both the Financial Services Board and the Pension
Funds Adjudicator. He has highlighted the extent of his loss of benefits
and pointed out how executives have been treated more favourably than
On all counts, Mr
Kransdorff’s complaints have been rejected. Mr Kransdorff,
understandably, is one person who has little confidence in the alleged
“checks and balances” of the Financial Services Board. He will now be
relying on Swedish Social Responsibility Funding to correct what, in his
view, is a gross mal administration of not only his own pension fund
benefits, but of the industry.
For the record, the rules
of the Sentrachem Fund require the approved actuary of the fund to be a
Fellow of the Institute of Actuaries of London or The Faculty of
Actuaries in Scotland.
Educating the Public -
In the meantime, I must do all I can to educate the South African public
on the fundamental differences in actuarial practice between the South
African Actuarial Society and its parent body, the Institute of
An estimated R80 billion has systematically been removed from working
South Africans’ retirement benefits. The cost of this enormous problem
will only be realized in the future when the under-funding starts to
impact on millions of workers and as inflation eats away at the
purchasing power of pensions.
It is unacceptable to
argue that past wrongs will be left and only the future funding
requirements will be revised. The current R80 billion, plus all the
company contribution holidays taken, will impact on the future. It must
not be swept under the carpet.
of the Institute and the Faculty -
While accepting that the
Institute and the Faculty have no general jurisdiction over South
African actuaries, I really need to know if the Institute’s code of
conduct and ethics applies to South African actuaries registered as
Fellows of the Institute of Actuaries and the Faculty of Actuaries.
Keep in mind that the
Pension Funds Adjudicator is accepting that British actuarial
methodology was applied in the Sentrachem case. He used the apparent
"fairness" of the methodology as justification for rejecting the
argument that the actuarial reserve value should be transferred at
On the surface, this R80
Billion surplus was created largely under the banner of British
actuarial practice, yet in Britain this practice is illegal.
The Actuarial Society of
The Financial Services
The Pension Funds
Professor Anthony Asher
Actuary Peter Theunissen