EQUITABLE LIFE MEMBERS
EQUITABLE LIFE: PENROSE AND BEYOND
- ANATOMY OF A FRAUD
A paper by Dr. Michael Nassim
Last Updated: Friday, February 11, 2005 09:59 AM
Besides
saying how the money may have disappeared, Lord Penrose has also indicated
where and when it did so. Professor
Smith’s presidential statement in the 1992 accounts includes: ”… The
fundamental philosophy is that each generation of policies should receive
benefits commensurate with the earnings produced during its lifetime.
Beyond the bounds of normal commercial prudence, it would be alien
to our culture to hold back benefits from one generation to build reserves
for a future generation. As
we say in our literature, for new policyholders future bonuses must depend
primarily upon the earnings produced on the investments of the new
premiums. Any deliberate
cross subsidies between generations would not be “equitable”.
Lord Penrose comments: “As
appears frequently in these documents, there was an inherent contradiction
in the statement. If nothing was held back for the future, any over-allocation
had to be recovered from the future.
That had to involve deliberate anticipatory cross-generational
subsidy and to be, in these terms, ‘inequitable’” (P4.82-3;
see also EARW p7). Did
Professor Smith know this?
A
more detailed account of cross-generational subsidy is given in P6.71-3.
Lord Penrose explains: “…While the split between GIR,
reversionary bonus and final bonus might vary depending on the policy
structure, the same total return was credited to accumulated policy values
regardless of duration. This
contrasted with the practices of many other companies with significant
amounts of single premium business, where more flexible bonus systems
allowed for variation according to duration in-force and financial
experience over the relative periods.
This flexibility allowed inter-generational smoothing and enabled
the company to avoid locking in to payout patterns in the way the Society
did” (¶ 71). And:
“More generally, it is unclear on what basis the Society could
ever have expected to have been able to operate without inter-generational
transfers in the absence of flexibility to differentiate by duration.
In With Profits Without Mystery it was acknowledged that each
generation of maturing policies was dependent on capital provided by
succeeding generations (cf. Ponzi in narrative below- MN). Flexibility to
differentiate by duration would have been necessary for such a system to
operate fairly in the absence of an inherited estate.
The Society’s practice did not provide that flexibility, and,
given its full distribution policy, over-payment on older duration
policies was the direct result” (¶ 73).
This, then, was a crucial piece of knowledge for better-informed
policyholders or groups if they wished fully to exploit product
flexibility, and the availability of total policy values from 1987
onwards. It also seems fair
to assume that few individual members without access to well-connected
financial advice would have known of it.
With this explanation in mind it is also instructive to look at PR Table 6.13 and 6.16, which show how terminal bonus became increasingly dominant over the years 1980-2000. The story can be made more complete and the effects of it judged by graphing a Society promotional table showing terminal bonus accumulation over the 35 years preceding 1999 expressed as a percentage of all the guaranteed elements. It forms part of Society leaflet (45J099 WP/UL 2.99) entitled: “Pension and Life Assurance Plans. Annual Statements- Further Information”. When plotted as in the figure below, the data show that the biggest relative annual terminal bonus additions were made over the 1973-86 period, i.e. that spanning the dispersal of the estate during which Barry Sherlock was the Appointed Actuary and subsequently Managing Director. This was also the period of steepest relative accumulation rate of terminal bonus. The graphs also indicate how much terminal bonus was awarded retrospectively to policies in force during the decade prior to its introduction in 1973.
NB:
The cumulation profile illustrated is relative to the guaranteed
elements of individual asset share and guaranteed bonus cumulation (the
cumulation of which is also important for determining the absolute
amount of asset share, which in turn depends upon individual contribution
history). It assumes a constant level of annual contribution, but, given
the structure of the fund, suitably informed policyholders might have done
well to increase their later contributions.
They would also then have had to crystallise their gains at an
appropriate and advantageous time; later policyholders whose policies had
more time to run were perforce left exposed, and had also to meet the
costs of previous over-distributions (cf. Ponzi below).
After 1985/6 the real glory days of terminal bonus and potentially
crystallisable gain were over.
From the sales point of view it all looked very impressive, and so
it is not surprising that the senior management team wished to carry these
figures over to support a new fund in 1987.
The prospect of such performance continuing on the same basis was,
however, entirely without foundation, such that the use of these prior
performance figures was inappropriate, un-representative, and misleading.
The
beneficiaries of “deliberate
anticipatory cross-generational subsidy”, who received the rump of the
estate and more between 1973 and 1987, were thus longer duration
policyholders who were in other ways also selectively favoured (P chap
3.46, 51, 60, 74, 94, 103, 109, 112, 129, 144, 145 items 5 & 8, 152,
161 & 168; chap 4.16 &36; 14.68).
And since the GAR option was withdrawn for policies begun after
1988, this group was also essentially comprised of longer-term GAR
policyholders. Moreover after
1987, “The advertisement of
“policy value” to members who had reached an age at which it was open
to them to take their benefits allowed those with the financial acumen or
appropriate advice to elect for an early maturity date and crystallise
their benefit entitlement at an unduly high value to the disadvantage of
continuing members. In the
early 1990’s, this was facilitated by new flexible products developed by
the Society” (P chap 4.29). The
implications of this are serious, and the more so if they involved fund
switching. One reason for the
selective bonus allocation to policies of particular durations was said to
be maintenance of the Society’s position in performance tables, i.e.
marketing considerations. And yet, as initially explained, the historical
problem was that established members would resist significant increases in
membership because it diluted their own claims on the estate.
Given the Society’s commitment to expansion, it would have been
politic to forestall this by awarding longer-term members a
disproportionate share of the estate. One must also hope that similar
favours were neither sought nor proffered for inward transfers of longer
duration policies during the drive for expansion.
Depending
upon personal and external circumstances, postponement of retirement could
also achieve the same end. Consider,
for example, Headdon’s explanation of why the actual results for 1995
had shown payments to policyholders considerably in excess of those
projected for the year. “
It appears that some clients were delaying retirement until the managed
pension contract became available. A
substantial part of retirement proceeds will, in fact, have been left with
the Society and contributed to the premium income position…”.
In the same paragraph (chap 4.128) Lord Penrose continues:
“The consequences of the managed pension for policyholders was to
become a significant factor in the closing years of the reference period.
Evidence from the independent financial adviser sector indicates
that some at least some of its practitioners appreciated that it was to
the advantage of their clients to take benefits at a time when policy
values were high relative to underlying assets.”
Rough
news on smoothing generates new insights.
If an investment fund holds
free reserves in one form or another, they confer the ability to even out
fluctuations in earnings from year to year.
The greater the reserve, the longer returns and bonuses can be
maintained at what is deemed to be a safe average level in times of
adversity. If there is no
estate or other free reserve, then earnings cannot be smoothed out in this
way. Not surprisingly,
therefore, Lord Penrose was unable to establish that the Society had a
defined smoothing policy despite its officers having repeatedly referred
to the existence of one. It
is, however, possible to discern what substituted for one from the WPWM
paper, parts of Sherlock’s 1989 letter to policyholders, and a revealing
exchange recorded in the PR between Christopher Headdon and Ernst &
Young audit actuary Ian Bannon in 1995. Under the WPWM paradigm there was
no estate or free reserve; in the absence of one the unconsolidated
surplus, i.e. all the accumulated terminal bonus, became what was referred
to as the “investment reserve”. One
of the grey areas in the 1989 WPWM paper was exactly what the investment
reserve would be called upon to do until it went to policyholders.
In concise technical terms it was unclear how such surplus was
hypothecated before it “crystallised” on becoming contractually due in
maturing claims.
First and foremost, in the
absence of any other funds the investment reserve had now to be the surety
for the guaranteed element in policies.
Effectively, therefore, policyholders’ un-guaranteed bonus was
used to underwrite the guaranteed part of their own fund value, and with
hindsight one may well question what sort of a guarantee that was.
The investment reserve had also to finance ongoing with-profits and
other business, and to take up new business strain as the Society’s
membership grew in advance of anticipated premium income.
A bona fide custom of paying out full policy values in a
climate of good faith required the investment reserve also to stand as
guarantor for itself. This in turn inevitably meant that in market
conditions when underlying asset values were depressed, and all the
previously awarded terminal bonus stood at more than market value after
also having to cover any associated shortfalls in guaranteed bonus, the
investment reserve would be negative. Under these circumstances, the only
way individual total policy values could be maintained was to allow
supposedly limited swings of the investment reserve between positive and
negative. And since this
could only work under uninterrupted relative stability in market
conditions, any capital smoothing capacity was more apparent than real. Moreover once the Society had compounded this situation by
attempting to extend it with loans and re-insurance treaties the position
was ultimately untenable. Come the GAR liability payable in good faith and
the situation was hopeless without the bad faith DTBP claw-back.
Whatever the deficiencies in
this situation, it purported to maintain total policy values rather than
to dampen pulsations in the yearly bonus stream.
Supposedly it smoothed capital value rather than any annual
additions to it, which notwithstanding Sherlock’s limited explanation of
it is not a layperson’s ordinary understanding of smoothing.
Moreover, rather than risk thereby reducing it extended
indefinitely, and increased progressively as premiums and dominant
terminal bonuses accumulated. Meanwhile, if a steady stream of annual bonus additions were
to be maintained at a level that satisfied market expectations, these were
a deferred liability that did not require immediate further setting off
against the investment reserve, and allowances for future premium income
such as Zillmer adjustments, subordinated loans or bonds and internal
business loans, all of which anticipated the income continuing could be
swung in to cover them. But in fact most of the Society’s policies were
recurrent single premium ones, such that a steady annual premium flow
could not be taken for granted. Even
so, the PR demonstrates that an inappropriate quasi-Zillmer adjustment was
ultimately extended to cover £900million of the Society’s shortfall in
funds. When the premium stream began to dwindle in 1998/9 the
position became progressively more uncovered.
With these inherent
weaknesses there remained but one means of sustaining the notion of
smoothing, and paying out total policy values in full as they matured or
were surrendered on contractual terms in a manner that did not further
erode statutory solvency margins. As revealed in the Headdon-Bannon
exchange, it was to pay the unconsolidated terminal bonus part of claims
as they arose from a first call on the Society’s current earnings,
always against the aforementioned backdrop of positive-negative swings in
the value of the investment reserve. By contrast the guaranteed part of
maturing claims was struck out from the guaranteed liability reserve.
This overview of smoothing or
the lack of it is worth following because it allows several further
significant insights. Firstly one can recognize that different accounting standards
may contain disparate conventions for hypothecating free assets, let alone
a so-called “investment reserve” of assets which are not free because
they are being held against an eventual liability.
Hence disparities or elective shifts in hypothecation of any
resulting deficit, obscured by the Society’s sophistry-laden business
paradigm, are what allowed key members of the senior management team
successively to elude most of the Society’s directors, virtually all its
members, the organisation of the Faculty & Institute of Actuaries, the
Society’s auditors and the regulator.
Herein lay the usefulness of the three different sets of accounts.
More sinister is the realisation that “first call” payouts from
current earnings, and a bonus stream predicated on future earnings rather
than inherent financial strength, are characteristics of accumulating new
business strain followed by a Ponzi scheme. In this as in so much else,
the many aspects of non-disclosure are paramount, but on the face of it
the Ponzi scheme was legal, because it was all levered off the
un-guaranteed part of policy values.
Even so, the London and Edinburgh discussions of the WPWM paper
show that an informed and interested minority of the actuarial profession
could and did deduce much of what was going on. As yet, history has not
related what other representations they made within their profession, or
how else they may have used their knowledgeable inferences.