It needs someone who has studied the cabala to divine why two of the world's largest (re)insurance operations apparently produce well below par investment earnings. Trevor Petch explains.

By revenue, company A is in the top three in the world in its field of operations, with operations in 60 countries. Company B is the largest in the world in its sector.Once upon a time, they were famous for providing a career structure almost as comfortably paternalistic as their Japanese counterparts, if not quite a Chinese-style “iron rice bowl”. Times have changed, but either can probably still take its pick of graduate recruits annually. Their investment departments are staffed with specialists in each class of instrument in which they invest.

Then consider the following proposition:
Company A and Company B produce a return on their own investments which appears to be amazingly low by comparison with their international competitors. Not only that; they also appear to have been consistently unable to make anything like the return on investment which they would get by investing in their local stock market index, or to achieve the average return on government bonds.

It seems intuitively improbable - but it is the sort of thing which the investment community is frequently invited to believe about Allianz and, especially, Munich Re.Part of the explanation lies in the opacity of German insurance company accounting. Compiling the financial statements of a German insurer is an arcane profession ideally suited to those who, in previous incarnations, practised cabalistic studies or divined the future by examining the entrails of sheep.German insurance accounts have to be legal, but not completely “true and fair” in the Anglo-Saxon auditor's sense. Investment income is one simple example. Insurers' investment earnings appear low because the investment earnings stated in the profit and loss account are low. But “investment earnings” and “investment earnings released to the profit and loss account” are not the same thing. And this can only be regarded as sensible, when earnings which contribute to pre-tax profit will be taxed at a marginal rate of over 55%.

The German insurance industry holds most of its investments in equities through intermediate vehicles. These take two basic forms. One is tax efficient investment companies, which typically hold stakes in industrial companies and banks dating back to the early days of German post-war reconstruction. At the end of last year, these and most directly-held stakes (such as Munich Re's holdings in Allianz) have been transferred to limited partnerships to maintain flexibility ahead of a tightening up of regulations on taxation of exchanges of participations this year.

More important are the spezialfonds, in effect single owner investment trusts, which are used for most short term equity investments and some bond holdings as well. In 1996, for example, German life insurers held 77% of equities through these vehicles, compared with 17% directly and 6% through participations (equity stakes of over 20% or held through investment subsidiaries). The owner will receive income from these companies in the form of a dividend.

Spezialfonds may net off capital gains and losses within the fund. Current income accrues to the fund, but the owner pays the tax arising immediately. One effect is to transfer an unidentifiable part of current investment income to valuation reserves.

There is another reason to doubt that Allianz and Munich Re have investment returns as low as is sometimes suggested: Allianz will produce IAS accounts (almost identical to US GAAP) for 1998, and Munich Re probably for 2000. Were either to reveal levels of investment earnings massively below those of international peer companies, such a step would hardly be under consideration.

Half of the surprise has already been removed by the European Union requirement for disclosure of market as well as book value for equities, bonds and real estate. Under IAS, a fair value will also have to be stated for schuldscheindarlehen, the privately issued mortgage or municipal debt-backed securities in which German life insurers invest heavily, but the relative impact will be small.

Restated earnings will be far more interesting. IAS tends to boost relative earnings compared to continental accounting by requiring full discounting of acquisition costs and of goodwill (goodwill acquired since 1994, strictly). Equalisation reserves have to be unwound.

Hannover Re has already published its 1997 accounts restated into US GAAP. Shareholders equity is 3.4 times higher than under German GAAP, or DM2.56 billion against DM748 million. Net income is 4.1 times higher, or DM394.5 million against DM95.7 million. The simple return on equity is 15.3% against 12.8%.Net investment income in the German accounts (including investment allocated to the technical account) is DM1.086 million, a return of 6.7% on the investment portfolio of DM16.098 million at market value. Under US GAAP, net investment income and realised capital gains total DM1.315 billion, on investments of DM15.456 billion, a return of 8.5%.

Can we apply the same kinds of multiple to Allianz or Munich Re? No, because their valuation reserves and investment portfolios are far more extensive and more diverse, for one thing. But it can confidently be predicted that Allianz's return on investments will be higher than the 7.7% it stated in its 1997 accounts, based on “released” investment return/average book value.

How high? My guess is high enough to nudge the reported overall RoE of 12.4% after tax for 1997 under German GAAP to 14% for 1998 under IAS.

Trevor Petch is an insurance analyst with Robert Fleming Securities Ltd.