EQUITABLE’S MISSING BILLIONS by Michael Josephs

 

The Lord's of Nonsense Trilogy

Without Profits: the Mystery that caused Equitable's Demise

Equitable a Regulatory Disaster

 

The Enigma of Equitable’s Funds

When Equitable Life closed to new business in December 2000 after failing to find a buyer, no-one could understand what was wrong. It was implied that the liabilities to holders of policies with Guaranteed Annuity Rates (GARs) were so great that they threatened the survival of the Society. Gradually, a fuller picture emerged. Equitable’s assets were inadequate to back its policy liabilities, despite a seven year bull market, and bonus declarations that were good but not exceptional.

Ned Cazalet, the insurance analyst, was probably the first to bring this to public attention in his evidence to the Treasury Select Committee. “In fact, the level of financial restoration that would have been required by an outright bidder (which would have taken on the long term liabilities of the Society by means of a demutualisation and transfer of such business) was probably more like £4bn," he said.  This was despite limiting bonuses for the year 2000 to only 3^1/3%.

Later, after Equitable had instituted an internal investigation and cut all members' total policy values by 16%, an industry ‘first’, Vanni Treves, new Chairman of Equitable told the Daily Telegraph: "We corrected the financial imbalance the society had been suffering from for some years and which, of course, was exacerbated by poor investment returns.

"The bonus policy that had previously been adopted was only sustainable in situations which no longer pertained - a substantial inflow of new premium income."

Charles Thomson, the new CEO, was also reported as saying: “Bonuses paid on With-profits policies by Equitable Life were too high, leading up to the group's closure to new business. “

Later, the Independent Actuary Michael Arnold included these words in his report on Equitable’s S425 compromise scheme:
“Aggregate policy values exceeded asset values by approximately 10% as at 31 December 2000." (para 2.2.7) Total assets at that time were approximately £35 Billion, and the shortfall is therefore taken to have been around £3.5 Billion.

It's existence was further confirmed by witness testimony at the case brought by Equitable against its former auditors Ernst and Young in January this year, referring to a forensic investigation into Equitable's finances by Price Waterhouse Coopers in 2001.

How did the "black hole" arise?

Most life assurance companies investing in equities, as Equitable was, had acquired substantial surpluses following the long bull market which hit its peak in April, 2000. It is very surprising that Equitable should have been short of £3.5 Billion, when prudent management might have ensured that it had built up at least £3 Billion as a smoothing reserve; (a scarcely adequate 11% of WP assets). In other words, the Society was short of something like £6 to 8 Billion.

Investors' Association research had already identified that Equitable ran its funds in a unique way - outlined in the paper With Profits Without Mystery, presented by the former Managing Director and Appointed Actuary, Roy Ranson, in 1989, and that this had led to two specific unusual innovations:

 

¨       Entitlement to terminal bonus (TB) was awarded annually, in increments. Life companies normally recalculate notional TB each year, disregarding the previous year’s figures, which are not treated as part of policy value. The incremental approach poses a risk that the funds earmarked the previous year for Non Guaranteed Bonus (Equitable’s term for TB), which do not appear anywhere in the accounts, will be used for something else, unless additional controls are applied.

 

¨       The Society had a stated practice of allowing early surrender of policies at the notional (‘redemption’) policy value, including the terminal bonus without imposing any penalty (known as a Market Value Adjuster - MVA - in the industry)although there was no contractual obligation to do so.

 

It seemed that both of these policies could give rise to substantial costs, particularly if non-guaranteed (terminal) bonuses were being over-declared. Estimating the levels of these apparently over-declared bonuses was the key to the problem.


What the survey found

A financial model was developed using the global flow of With profits premiums and claims to estimate the net amount of terminal bonus in members' total funds at any one time. It incorporated explicit discount factors mirroring those employed by the Actuary when valuing the policy liabilities, and was refined a number of times to better match the actual financial results.


These are the key results of the investigations:

 

¨       There is a good probability that Equitable was short of assets throughout the whole of the 1990s, and that the bonus notifications to policyholders were not fully covered by assets.

 

¨       The typical shortage appears to have averaged around £1 Billion over the period 1990-1996 inclusive, and around £1.8 Billion over 1997-2000.  The increase over the latter 4 years is closely correlated with the onset of the GAR costs.  (The real increase may well have been higher on a consistent accounting basis, but we used the actual figures as published in the Accounts.)

¨       If so, policyholders would have had an excessive expectation of what their policies were really worth.

¨       Any policyholder redeeming, surrendering or switching a With Profits policy during that time frame, would have taken away any excessive valuation included in that policy. The model estimates such ‘over-declaration’ costs at £1.5 Billion.

 

¨       About 25% of the claims in a typical year derived from transfers or surrenders, i.e. early terminations. The model gives an estimated valuation discount year by year, and the early termination (‘free surrender’) cost is taken as the difference between the actual claim cost and the discounted liability. Such early termination costs come to just over £1.1 Billion.

¨       Over the years, such unbudgeted costs would also cause a progressive loss of interest to the Society which , at 7% per annum, would accumulate to £0.9 Billion, making a total shortfall of £3.5 Billion in all.

The model can also be applied to premium and claim flow post-2000, but there were a number of disruptive factors which would interfere with the model assumptions. These include the implementation of the Compromise, the 16% levy on policies in July 2001 and varying levels of MVA and redemption penalty. Also, the liabilities showed distinct instabilities during the period 1998-2001 due to the effects of GAR reserves and other underwriting problems.

¨       However, the model does indicate an unbudgeted cash loss of around £1.3 Billion during the whole of 2001, before and after the 16% cut was applied, which, if confirmed, would bring the total losses to nearly 5bn pounds.

Cross checking the figures

A second model was also used to check on the results of the first. It took a diametrically opposite approach to estimating the ‘hidden’ terminal bonus funds, using a "bottom-up" approach based on pension business only, while the first model uses a "top down" method based on all With profits premium and claims flows. The second model omits about 10% of WP business.  Both models agreed on the basic conclusions.

The second model assumes that the average age of pension policies stays fairly constant with the passage of time, and also that all such policies are subject to constant annual premiums. In practice there are many large single premium policies which would tend to push up the estimate. For at least the two reasons mentioned, we would expect the second model's results to be on the low side. The first model does indeed give results for terminal bonus amounts that are about 10% higher than the estimates from the second model, which is a surprisingly close agreement.

But both models show asset shortfalls at the end of the 1990s, with the second model giving a shortfall 250m pounds lower than the first model's 3.5bn.Both show shortfalls for every year from 1990 onwards, with the exception of 1993 in the case of the second model.  [For historical reasons, the main model is referred to as ‘Model B’ in the charts.]

Implications

Equitable Life's management have refused to reveal their analysis of the figures over the past decade. Our figures show how the specific and very unusual innovations which Equitable adopted from the 1980s would have led to major unbudgeted cash leakages from the Society’s assets.

They also explain the disastrous financial position which the Society found itself in after the closure to new business. Not only -as was well known - did the Society not keep an orphan estate of free assets, but it was, in addition, short of 3.5bn in assets covering the purported value of policies as stated on bonus notices.

 
It was this shortfall, rather than the House of Lords' decision about the Guaranteed Annuity Rates, which we believe brought down the Society. Indeed, we estimate that even if the House of Lords had ruled in the Society’s favour, It would have faced a shortfall in excess of £3 Billion at the end of 2000.  This would have worsened dramatically as share prices dropped around the world, and the Society would still have been forced to close.

It remains to be seen how the Society's accountants, auditors, management and regulators failed to discern and put right a problem which our figures show had been developing over many years.

LIST OF APPENDICES

 

CHART 1 -  Model B estimations of cash losses year by year.

 

CHART 2 - Comparison of  Estimated Shortfalls from Models A and B.

Model B is the primary model used in this paper; Model A is the secondary one.

 

TABLE  1 - Calculation of Estimated Losses, derived from over declarations

and early termination costs.  Losses are based on monies paid out (or transferred) in claims, and are derived from Model B.