Continental drift: the EU's transition to international reporting standards is well under way. David Tyrrall analyses the change in four countries--France, Germany, UK and also the US--by examining its effects on the annual reports of four prominent firms.

These are interesting times for accountants, as the most ambitious attempt ever to harmonise accounting practices in the EU, and indeed the world, gathers steam.

Early attempts at harmonisation by the European Commission didn't get far, because the member states weren't truly willing to drop national practices. The compromise position between the countries was "you accept my financial statements and I'll accept yours". Meanwhile, the US Financial Accounting Standards Board was pursuing its own projects separately, although with increasing co-operation from the UK, Australia and New Zealand. But the inexorable rise of global capital markets led governments to review this separatist approach and seek greater convergence. As a result, all EU member states agreed to impose international financial reporting standards (IFRS) on listed consolidated financial reports from January 1 this year. These changes will affect all users of financial reports, but the degree of adjustment is likely to vary across Europe. The process is well under way: reports prepared under IFRS are now steadily appearing and receiving press comment.

Before examining the effects of IFRS on the annual reports of four European firms moving from German, French, UK and US Gaaps respectively (see panels), let's consider the main accounting differences between the nations. In France and Germany the state has played a major role in the regulation of accounting, making creditor protection a central concern. In both countries there has been a tendency towards historic cost accounting modified by a prudent attitude that understates profits and assets. This approach contrasts with IFRS, which favours fair values. Such tendencies have been reinforced by the close relationship between accounting profit and taxable profit.

The state has taken a much more hands-off approach to accounting in the UK and US. Independent boards established accounting standards in both countries and they have tended to focus on shareholder information rather than creditor protection. Accounting in the Netherlands, on the other hand, has tended to occupy a halfway house between the Franco-German and Anglo-American traditions.

The International Accounting Standards Board (IASB) has developed in an Anglo-American (mainly Anglo) environment, so it's not surprising that it believes the main aim of financial statements under IFRS should be to provide relevant information for investors.

It's widely believed that Anglo-American accounting standards are more relevant than Franco-German standards are to investors. If this view is correct, you might expect financial statements prepared under IFRS or US or UK Gaap to be more "value relevant". Putting it differently, you might expect a closer correlation between the accounting figures in the financial statements and the market capitalisations (share prices) of companies using Anglo-American standards than that between the accounting figures and market capitalisations of companies using Franco-German standards. You might also expect that, if companies shift from using Franco-German standards to Anglo-American standards, the value relevance of their annual reports will increase.

Normally we can't conduct experiments in economics--or accounting, for that matter--but Europe's transition to IFRS is giving George Kontopoulos at Cass Business School an unusual opportunity to find out whether it will increase the value relevance of accounting information or not. He is using a regression model to examine the value relevance of accounting information for a sample of 200 companies (50 each in France, Germany, the Netherlands and the UK) across the transition period of 2003-06. Obviously, the post-transition results aren't available yet, but the preliminary findings for the pre-transition period have been surprising. Contrary to expectations, financial statements in France and Germany seem to be more value relevant than those in the Netherlands and the UK.

It's early days yet and much research remains to be done, so let's turn instead to two early IFRS adopters--BMW Group and Nokia--and two firms that changed over in January--Faurecia and AstraZeneca--to examine the practical effects of the shift on each firm's financial statements. These are summarised in the panel at the bottom of page 14. The companies were not chosen to achieve the more or less stereotypical results shown, but you could be forgiven for believing that they were. For BMW, profit and equity rise under the shift from German Gaap to IFRS. Faurecia's profit also rises under the shift from French Gaap, but the effect is smaller. For Nokia the shift from US Gaap and AstraZeneca the shift from UK Gaap give apparently random increases and decreases of similar amounts+ indicating their prior proximity to IFRS.

The large changes between AstraZeneca's US Gaap and IFRS results look like an outlier in all this, since we might have expected little change. But this was not a scientifically selected sample and we should be wary of drawing bold conclusions. Most EU firms have still to announce their IFRS results and it may take several years for IFRS to become embedded and its full effects to be realised. A more cautious interpretation would be to emphasise just how variable the results announced on switching to IFRS can be. It would be nice to be able to say that the main causes of the differences are the prime suspects--financial instruments, goodwill, pension liabilities and share-based payments. These are certainly guilty parties, but deferred tax, depreciation, development intangibles, inventories, proposed dividends and provisions are all active accomplices.

Although the conversion has had very variable effects on Nokia's results, perhaps we might take hope from Maijo Torko, the firm's senior vice-president and corporate controller, who said: "IFRS has helped international investors to better understand our financials and has, therefore, enabled us to grow our international shareholder base." But how all this will pan out in share prices remains to be seen.

GERMAN GAAP BMW GROUP

BMW adopted IFRS from its 2000-01 financial statements. As can be seen from the comparative balance sheets, the effect of the shift from German Gaap to IFRS was what we might have expected: large changes in total assets and Liabilities, which combine to almost double the equity, although there are major offsetting items within the net increase.

The 4,536m [euro] increase in equity resulted from the following items:

* Intangible assets: + 2,035m [euro].This increase resulted from the write-back of R&D costs on vehicle and engine projects. Under German Gaap, such spending would have been expensed as incurred. The policy under IAS38 is to capitalise R&D intangibles if key criteria relating to the certainty of future income flows and their related costs are met.

* Property, plant and equipment: + 885m [euro]. The increase mainly resulted from the shift from reducing-balance to straight-line depreciation, which tended to slow down asset depreciation.

* Inventories: + 898m [euro]. Most of this increase was because of the addition of production overheads to inventory values (inventories under German Gaap were measured using direct costs of production only), and because BMW had been recognising higher inventory write-downs (German prudence) than would be normal under IFRS.

* Deferred tax assets: + 922m [euro]. Under German Gaap, BMW recognised all deferred tax liabilities but not all deferred tax assets. Under IFRS, these extra assets were recognised. Deferred tax liabilities also increased by 199m [euro] because they could no longer be offset against deferred tax assets.

* Provisions:--1,252m [euro]. About half of this was reclassification, but 673m [euro]of the decrease was because of the relaxation of German prudence. The recognition of deferred expense provisions (eg, repairs) as permitted by German Gaap is not allowed under IFRS.

* Liabilities and deferred income: + 26,152 [euro]. Most of this was reclassification, leaving a 1,074m [euro] reduction in equity owing to the fair valuation of financial instruments.

The large offsetting movements were mostly related to leased products (+ 7,876m [euro]). Under IAS17, BMW had to reclassify its leased products on both the asset and the liability sides. On the assets side, this can be seen as the shift of 17bn [euro] out of assets from sales financing mainly into receivables from sales financing. On the liabilities side, there is a shift of 16bn [euro] out of liabilities from sales financing into liabilities and deferred income. This Leaves a net increase in the balance sheet total of around 7bn [euro] with the recognition of [eased products as an asset and a corresponding liability in deferred income, both of which were previously off balance sheet.

BMW's balance sheets at
December 31, 2000                      German Gaap    Change      IFRS

                                          [euro] m  [euro] m  [euro] m

Intangible assets                              103    +2,035     2,138
Property, plant and equipment                5,710      +885     6,595
Financial assets                               950       -78       872
Leased products                                 --    +7,876     7,876
Non-current assets                           6,763   +10,718    17,481
Inventories                                  2,809      +898     3,707
Assets from sales financing                 17,578   -17,578        --
Receivables from sales financing                --   +17,082    17,082
Other current assets and prepayments         8,184    +1,423     9,607
Current assets                              28,571    +1,825    30,396
Deferred tax assets                            541      +922     1,463
Total assets                                35,875   +13,465    49,340

Equity                                       4,896    +4,536     9,432
Provisions                                   8,173    -1,252     6,921
Liabilities and deferred income              6,636   +26,152    32,788
Liabilities from sales financing            16,170   -16,170        --
Deferred tax liabilities                        --      +199       199
Total equity and liabilities                35,875   +13,465    49,340

BMW's profit history (excludes Rover restructuring exceptional)

Year                  1996     1997     1998     1999     2000
Gaap                German   German   German   German   German

PAT ([euro] m)         420      638      462      663    1,026
Equity ([euro] m)    4,636    5,240    6,445    3,932    4,896
ROE                    +9%     +12%      +7%     +17%     +21%

Year                  2000     2001     2002     2003     2004
Gaap                  IFRS     IFRS     IFRS     IFRS     IFRS

PAT ([euro] m)       1,209    1,866    2,020    1,947    2,222
Equity ([euro] m)    9,432   10,770   13,871   16,150   17,517
ROE                   +13%     +17%     +15%     +12%     +13%

FRENCH GAAP FAURECIA

Faurecia is 71 per cent owned by PSA Peugeot Citroen. It makes interior parts, mainly seats, for vehicles. Although it has 160 plants worldwide, the European market represents 84 per cent of its 11bn [euro] total sales in 2004.

Faurecia has presented a preliminary comparative statement on its transition to IFRS. Two main items account for the net 309m [euro] decrease in net assets and equity from 1,893m [euro] to 1,584m [euro] in the 2003 balance sheet:

* An increase in its pension liability of 126m [euro] under IAS19 on employee benefits.

* A reduction in goodwill of 135m [euro] as a result of restating the 2001 acquisition of Sommer Allibert as if IFRS3 had applied. This allowed the firm to reclassify a contractual customer relationship as a separate intangible asset from goodwill. The reclassification obviously had no effect, but Faurecia then retrospectively amortised the item to zero by January 1, 2005, thereby pre-empting prospective future impairment charges on it.

Like BMW, Faurecia also had some large offsetting items. It was forced to bring some items that were off balance sheet under French Gaap on to the balance sheet, recognising offsetting assets and liabilities of 295m [euro] in respect of a receivables factoring arrangement. Unlike BMW, Faurecia had already been capitatising development costs (423m [euro])--into inventory. These have now been reclassified, perhaps more appropriately, as intangible assets.

In contrast with the net reduction in the balance sheet, the effects of the transition on the P&L are positive--an increase from 95m [euro] profit to 142m [euro]in 2004--as they were with BMW. The main items were:

* A reduction in depreciation (15m [euro]) on intangible assets, as goodwill is no longer amortised, and on tangible fixed assets owing to changes in depreciation policy.

* Gains on the recognition of deferred tax assets (26m [euro]) that were not recognised in previous years.

The reduction in net assets combines with the increase in net profit to raise return on equity from five per cent to eight per cent.

Faurecia's preliminary comparative transition statement

[euro] m                          2002    2003    2004

Equity             French Gaap   1,955   1,893   1,957
                   IFRS                  1,584   1,701
Net profit         French Gaap     -43      23      95
                   IFRS                            142
Return on equity   French Gaap     -2%      1%      5%
                   IFRS                             8%

UK GAAP ASTRAZENECA

Swedish/UK pharmaceuticals firm AstraZeneca published its annual statements under UK Gaap but denominated in US dollars and with a US Gaap reconciliation because of the importance of its US listing. It didn't adopt IFRS as early as BMW or Nokia, but was one of the first UK companies to publicise the impact of IFRS on its results by issuing a comprehensive preliminary statement (see "Ready reckoner", February). Its CFO, Jon Symonds, observed that the overall effect of the new standards on the financial statements was "modest", although there were some larger offsetting movements.

The $162m transitional reduction in net profit from $3,813m to $3,651m in 2004 resulted mainly from recognising a share-based compensation expense (IFRS2) of $167m and losses on financial instruments (IAS32/39) of $128m, partially offset by gains of $49m from ceasing goodwill amortisation (IFRS3) and $66m on income taxes owing to the recognition of increased deferred tax assets (IAS 12).

It's arguable that AstraZeneca has enjoyed net gains to its P&L from the transition to IFRS. The losses on share-based payments and financial instruments would have happened even without the transition, because the UK Accounting Standards Board adopted these standards largely unchanged into UK Gaap.

In 2004 increases in equity owing to financial instruments ($28m), goodwill ($106m) and income tax ($128m) were swamped by losses on employee benefits (IAS19) of $1,417m, which the firm chose to recognise in full through reserves using principles similar to FRS17. But the combination of these with a reversal of the dividend declared ($1,061m) left equity almost unchanged throughout.

As a result, the balance sheet is more interesting for the changes that did not happen as a result of the transition, rather than those that actually did occur.

Although there was a major effect on BMW resulting from the capitalisation of development expenditures, the effect of this on AstraZeneca has been strangely negligible. The UK SSAP13 (research and development) has very similar criteria for the capitalisation of development expenditures to those of IAS38 (intangible assets). But, where IAS38 specifies that expenditure meeting the criteria must be capitalised, SSAP13 gives firms a choice between capitalising or expensing such expenditure.

US Gaap mandates expensing almost all R&D. Like many UK companies, including all in the pharmaceuticals industry, AstraZeneca expenses all R&D, so it might have been expected that a development intangible would appear on the AstraZeneca IFRS balance sheet, also causing a reconciling item in the US Gaap balance sheet. AstraZeneca incurred $388m of phase IV development costs in 2004. Given that phase IV trials take place after a drug has been shown to work, has been granted a licence and is already available to the public, the case for capitalising such expenditure might seem to be quite strong at first sight. AstraZeneca neatly resolved all these problems by reclassifying the $388m out of R&D costs and under the "Selling, general and administration" heading, which enabled the firm to continue its previous policy of full R&D expensing.

The startling difference in the AstraZeneca balance sheet is not between UK Gaap and IFRS, but between both of these and the US Gaap amounts. This $20bn difference is almost entirely a result of different accounting treatments of the 1999 combination of Astra (Sweden) and Zeneca (UK). Under UK Gaap it's recorded as a merger of equals (pooling of interests), while under US Gaap it's deemed to be an acquisition of Astra by Zeneca.

Acquisition accounting added $14bn of goodwill plus $8bn of fixed assets less $2bn of deferred tax liability to the 2003 US Gaap balance sheet. Like US Gaap, IFRS3 bans merger accounting, so AstraZeneca could have chosen to eliminate the difference between IFRS and US Gaap by restating its 1999 figures under IFRS3. Instead it had used the transitional arrangements in IFRS1 to freeze all prior combinations at the January 1, 2003 merger accounting values. Given that future differences between IFRS and US Gaap figures for share-based payments, pensions and financial instruments should remain small, that the treatment of proposed dividends is now consistent, and that the problem of capitalising a development intangible has been neatly sidestepped, why should AstraZeneca maintain this huge difference? Perhaps it's to avoid the prospect of a large future impairment of goodwill? There is no impairment of assets in the current year. Meanwhile, there will be small ongoing gains in the P&L from the cessation of goodwill amortisation on other acquisitions. Perhaps IFRS1 offers companies some opportunities for accounting arbitrage.

AstraZeneca's preliminary comparative transition statement

$ m                      1999    2000    2001    2002    2003    2004

Net profit    UK Gaap   1,297   2,277   2,906   2,836   3,036   3,813
              IFRS                                      3,014   3,651
              US Gaap  -3,539     865   1,397   2,307   2,268   3,051
Equity        UK Gaap  10,263   9,389   9,586  11,172  13,178  14,418
              IFRS                             11,145  13,120  14,425
              US Gaap  33,735  29,707  27,402  30,183  33,654  35,314
Return on     UK Gaap     13%     24%     30%     25%     23%     26%
equity        IFRS                                        23%     25%
              US Gaap    -10%      3%      5%      8%      7%      9%

US GAAP NOKIA

Finnish telecoms firm Nokia is Listed on the NYSE. It reported under US Gaap until 2001, when it became another early adopter of IFRS. It provides a US Gaap reconciliation in its annual report. The main differences between its figures under IFRS and US Gaap have arisen from the treatment of R&D costs and goodwill.

Both IFRS and US Gaap permit the capitalisation of software development costs, but US Gaap does not allow the capitalisation of some of the development costs that IAS38 would capitalise. This difference gave rise to an increase in profit from 1998 to 2002.

There are also differences in the amortisation and impairment of development costs. While US Gaap assesses impairment against net realisable value (NRV), IFRS assesses impairment against the higher of NRV and value in use. Higher impairment charges on development intangibles meant that development costs were higher, and profit lower, under IFRS in 2003-04.

There was little difference between the goodwill figures from 1998 to 2001, but differences in amortisation and impairment of goodwill increased US Gaap profit by [??] 100m to [??] 300m per year from 2002 to 2004. The differences in amortisation should disappear in future, because both the US Financial Accounting Standards Board and the International Accounting Standards Board now impaur, rather than amortise goodwill. But the difference in the way in which impairment is computed is likely to leave ongoing differences. In 2003-04, differences in goodwill accounted for almost all of the differences between the two balance sheets.

Differences between Nokia's key figures under US Gaap and IFRS

* m                      1997      1998      1999      2000

US Gaap net profit      1,076     1,689     2,542     3,847
IFRS net profit         1,053     1,750     2,577     3,938
Difference                -23        61        35        91
US Gaap equity          3,633     5,102     7,384    10,871
IFRS equity             3,620     5,109     7,378    10,808
Difference                -13         7        -6       -63
US Gaap ROE               30%       33%       34%       35%
IFRS ROE                  29%       34%       35%       36%

* m                      2001      2002      2003      2004

US Gaap net profit      1,903     3,603     4,097     3,343
IFRS net profit         2,200     3,381     3,592     3,207
Difference                297      -222      -505      -136
US Gaap equity         12,021    14,150    15,437    14,576
IFRS equity            12,205    14,281    15,148    14,238
Difference                184       131      -289      -338
US Gaap ROE               16%       25%       27%       23%
IFRS ROE                  18%       24%       24%       23%

This feature was compiled by F Adam, T Damgaard, G Kontopoulos, N Nohamed, M Rajan, I Stanoeva and D Tyrrall ([e-mail]) at Cass Business School, City University.