Terry Smith is pictured during an interview at the Financial Times building in London.
© FT

In August 1992, my book Accounting for Growth was published. It exposed how companies used accounting trickery to flatter their reported performance. Nowadays, there are brokers such as Muddy Waters and Iceberg who specialise in revealing these practices, but in 1992 such research was most unusual so it caused quite a stir.

My then employer tried to stop its publication, which of course only made people want to read it, sending it to the top of the non-fiction charts. I was fired, which sent my career off in a more entrepreneurial direction, and several of the companies named in the book got into serious difficulty or simply went bust.

A second edition was published in 1996 and I’ve often been asked to reprise the subject with another book.

One of the reasons I haven’t is that publication of the book roughly coincided with the start of a successful campaign by the Accounting Standards Board, led by Sir David Tweedie, to stamp out many of the abuses in company accounting.

Another is that I am not sure many investors or analysts study company accounts any longer. Instead, they seem to rely upon management presentations using “adjusted”, “core” or “underlying” earnings or profits.

One sector in which my Fundsmith Equity Fund does not own any stocks is pharmaceuticals. This seems to surprise some commentators, who think that drug companies would represent exactly the sort of dependable returns we seek. After all, such stocks benefit from seemingly inexorable growth in demand for healthcare, especially among the ageing populations of the developed world, and margins that are shielded from competition by patents.

One reason we don’t own them is that the sector has become rated on the basis of “underlying” earnings. Beginning in about 2010, many major pharmaceutical companies started a switch to reporting what they term “core” earnings. This switch was allegedly to smooth out exceptional items from reported earnings and make trends more recognisable.

So what is excluded from earnings based on generally accepted accounting principles (GAAP) to get to “core” earnings?

1. Restructuring costs, although they seem to be a recurring item in the accounts of a number of companies. GlaxoSmithKline, for example, has not had a quarter without any since 2008.

2. “Exceptional” legal charges. Once again, given the nature of the industry involving patents, patent disputes, regulation and product liability, it seems inevitable that significant legal expenses will be a more or less constant feature of a pharmaceutical company’s profit & loss (P&L) account. So it is hard to see how they are by nature exceptional.

3. Intangible asset amortisation and impairment. When pharmaceutical companies buy a drug from another company or buy another drug company — and they have been doing a lot of that — they create intangible assets that represent the amount they paid over and above the tangible or hard assets acquired. This is usually the vast majority, if not all, of the cost and GAAP requires it to be both amortised by a charge to the profit and loss account, usually over the life of the drug patents, and written off in the event that a drug fails its trials — a not infrequent event.

Some will argue that it is acceptable to exclude these intangible charges as they are “non cash”, but doing so turns the P&L account into a hybrid of accrual accounting and cash flows. If you are interested in a company’s cash flow, and you should be, the place to gauge that is in the cash flow statement, not a doctored P&L account which excludes some non-cash items. AstraZeneca, for example, has over £16bn of these intangibles on its balance sheet which would cause an annual charge of some £1.6bn, but this is not reflected in its “core” earnings.

Moreover, excluding these intangible items means that the cost of acquiring new drugs and biotech companies does not appear anywhere in these “core” earnings. In light of this, is it any surprise that the pharmaceutical sector has been on a binge of buying biotech companies, spending $80bn in 2014 alone?

All of the adjustments have one thing in common — they make the reported “core” earnings higher.

Faced with this opportunity to flatter the earnings it is also no wonder that management incentives have been remodelled to take advantage, with some or all of management remuneration in the sector based on “core” earnings.

Unsurprisingly, since 2010 the GAAP earnings per share (EPS) in the sector have decreased significantly as a percentage of “core” EPS — in the case of AstraZeneca from 84 per cent to just 23 per cent by 2014. In other words, as their pay has come to depend upon “core” earnings, more and more bad stuff has been excluded from the calculation.

For example, AstraZeneca reported GAAP EPS of $5.60 in 2010 but courtesy of “core” earnings this became EPS of $6.17. By 2014 the GAAP EPS of just $0.98 had become “core” EPS of $4.28.

The net result of all this is that the ratings which many pharmaceutical stocks appear to trade on bear little resemblance to reality based on their GAAP earnings:

Pharmaceutical imbalance
GlaxoSmithKline 22.513.5
Bristol-Myers Squibb52.634.2
Eli Lilly39.631.7
Source: Fundsmith

In my view, if you are an investor in pharmaceutical stocks this should worry you a lot. If you want to know more, broker Baden Hill Sanlam recently published a critical — and in my view good — research report on “Big Pharma”. The irony is that in order to discover what is really going on you do not need to be a sophisticated financial analyst. All you need to do is get the GAAP EPS number from the accounts.

Hence my view that there is no point in companies engaging in any form of accounting chicanery, when their legerdemain seems to have the entire market looking somewhere other than the accounts.

Terry Smith is chief executive of Fundsmith LLP

Get alerts on Major Pharmaceuticals when a new story is published

Copyright The Financial Times Limited 2022. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article