Join our community of smart investors

Financials in focus

Companies will always present their figures in the most flattering way possible which is why investors need to be on their guard. Stephen Wilmot explains the tricks and tactics used
October 31, 2014

The business of being a residential landlord is, on the surface, pretty simple. You own a property and you rent it out. Your profits depend on your rental income minus costs on the one hand, and on the change in value of the property on the other. The value of the company is the house price minus the mortgage. It's all highly intuitive.

You'd think pinning a value on Grainger (GRI) Britain's largest listed residential landlord, would be correspondingly intuitive. Sadly not. The corporate accounting equivalent of the house-price-minus-mortgage calculation is 'net asset value' (NAV): assets minus liabilities. But Grainger publishes two figures for NAV, neither of which are close to the NAV calculated under International Financial Reporting Standards. IFRS NAV was 122p a share as at 31 March, but the company prefers to cite 'gross' NAV of 272p and 'triple net' NAV of 228p. Its shares are currently priced at 184p. So are they cheap or expensive? Clearly the answer depends on what you include in NAV.

Grainger is not a lone case. Although its spread of NAV figures is unusually broad, valuing most property companies depends on working out what measure of book value is most realistic. And property companies aren't the only ones. Companies valued on the basis of earnings multiples routinely adjust those earnings to exclude 'exceptional' IFRS costs they feel distort their 'underlying' performance.

How valid are these adjustments? The Financial Reporting Council's annual review in October found 'a good level of corporate reporting by large public companies, particularly FTSE 350 companies'. But it noted that reporting by smaller companies, particularly on Aim, was of patchy quality. In response, it has set up a project aimed at 'driving a step change in the overall quality of reporting by companies in this part of the market'.

The adjustment of earnings is a key focus. The FRC is particularly concerned about:

• Lack of, or poorly described, accounting policies for exceptional or similar items

• Inconsistent application period-on-period

• Recurring and/or immaterial items often reported as 'exceptional'

• Lack of symmetry between debits and credits, i.e. including 'bad' news but not the related or equivalent 'good' news as exceptional

• Lack of comparative information.

The question of whether stocks offer value is fundamental to investing. But value can only be understood relative to either profits or NAV. Both are heavily adjusted in ways that differ from company to company; there are currently no guidelines on how to adjust profits (the property industry does use a standard methodology for adjusting NAV, as we will see).

Whether the FRC's efforts lead to greater standardisation remains to be seen. Finding norms that suit everyone is notoriously difficult. Private investors would therefore be wise to rely on their own understanding of accounts rather than a regulatory response - useful as that would also be. The only way to understand a company's accounts is to read them in depth over various reporting cycles. But the following primer should be a useful guide to the most common adjustments.

 

The earnings statement

'Adjusted', 'underlying', 'normalised' or 'headline' profit figures strip out various IFRS costs to give a more accurate - or sometimes just more flattering - picture of a company's operational performance. The rationale for removing costs is typically a combination of the following: a) they are 'one-off' items that won't recur and have little to do with the company's commercial operations; or b) they are non-cash 'accounting' charges.

The most common instance of a) is restructuring costs. These mainly consist of redundancy packages for edged-out staff. Restructuring is usually presented as a fresh start, so the logic for excluding the associated costs is explicitly that they will not recur. This makes sense, but investors shouldn't forget that the costs are still real: they still come out of shareholders' funds.

Moreover, some companies restructure year after year. This is mainly by accident rather than design, but an exception is engineering group Melrose, which specialises in buying underperforming companies, sifting through their costs to improve their margins, and then selling them on. Every year it posts 'exceptional' restructuring charges and costs associated with the purchase and disposal of companies. True, these corporate costs have nothing to do with the operational performance of the businesses in Melrose's portfolios - and investors are interested in that operational performance. But labelling the costs as exceptional is a bit misleading.

Other 'one-off' - but often in reality recurring - costs typically stripped out of earnings figures include legal fees when companies face litigation; fees paid to advisers in merger and acquisition talks, whether successful or aborted; and one-off losses (or gains) made when a business or asset is sold for less (more) than its book value.

The classic example of b) is the 'amortisation of acquired intangibles'. When companies make acquisitions, they typically pay a premium to book value that is accounted for as 'good will'. That good will may then be reduced through gradual write-downs: so-called amortisation. This is not a cost that involves any cash - or, more precisely, the cash was paid upfront in the acquisition rather than in the year of the expense. Companies that grow through bolt-on purchases of technology companies will typically amortise substantial amounts in this way. Health and safety specialist Halma is a good example.

This is perfectly valid and predictable - except when investments go wrong. Sometimes companies are forced to make big 'impairments' to good will: unexpected write-downs when acquisitions fail to live up to expectations. The most notable example in recent business history is the $8.8bn write-down Hewlett-Packard took in its 2012 annual results, following its $11bn acquisition of Autonomy in 2011. This risk applies not just to acquisitions but also to spending more generally on intangible assets like patents or software. Any spending that is capitalised on the balance sheet rather than expensed as a routine cost will have to be amortised (or, if the resulting asset is tangible, depreciated).

Another instance of b) is the exclusion of unrealised gains or losses on a portfolio of assets or liabilities. Real-estate investment trusts (Reits) routinely strip property revaluations out of their income statements to reveal an underlying earnings stream from rents. This is useful for investors, because the paper profits from property revaluations cannot be easily distributed as dividends, making rental earnings a more useful guide to dividend prospects.

But investors should be concerned when companies count asset write-downs as 'exceptional' and then write-backs as normal. This has been the case for many house builders. As the FRC documents in its October report, they wrote down land values in the financial crisis, excluding these so-called provisions from their adjusted earnings figures as being exceptional. But when they built houses on the land, sparking the release of these provisions, the resulting super-profits were not deemed exceptional.

 

The balance sheet

Many companies that are valued on the basis of their balance sheet will report an adjusted NAV figure that strips out or - more often - adds in items censured under IFRS. For property companies, the European Public Real Estate Association (EPRA) publishes standardised adjustments. There are three crucial ones that transform IFRS NAV into EPRA NAV.

First, trading assets are marked to market. Most of a property company's assets are held on the books as 'non-current assets', which means they are revalued under IFRS and the gains or losses put through the profit and loss statement. But some assets - typically development sites or assets ear-marked for disposal - are accounted for as 'current assets', which are required to be held at cost. These are (usually) written up to market value under EPRA guidelines.

Second, the revaluation of both current and non-current assets on a property company's balance sheet creates a capital gains liability. EPRA NAV strips this liability out. This makes very little difference for Reits, which don't pay tax on capital gains. But for non-Reits, such as developers, the liability can be substantial.

Finally, property companies that bought hedges before the financial crisis to protect themselves against further interest-rate increases, only to see rates collapse, are required to account for the extra interest-cost burden, discounted into the future, as a liability on their balance sheet. This so-called 'fair value of derivatives' gets stripped out under EPRA guidelines, the rationale being that the liability will disappear if the debt or hedges are held to maturity. In reality, however, many companies have chosen to pay off the hated hedges to reduce their finance costs and boost profits. McKay Securities is a company whose EPRA NAV has diverged significantly from its IFRS NAV for this reason (rising long-term interest rates have reduced the difference this year).

EPRA NAV is often described as the 'going concern' value of a property company - its value if it can continue trading. The more conservative 'break up' value is EPRA NNNAV or 'triple NAV'. This still accounts for trading assets at their market value, but includes deferred tax and the fair value of hedges in the same way as IFRS.

The reason Grainger's three NAV measures are all so different is that most of its properties are so-called regulated tenancies, held for sale rather than for rental income. This means they count as current assets and are held at cost on the books. Revaluing them takes the company's IFRS NAV of 122p to triple NAV of 228p. Grainger then has 11p per share of costly hedges and 33p per share of deferred tax, stripping out which pushes NAV up to 272p. With the shares trading at 184p, however, you thankfully don't have to use the over-bullish EPRA valuation measure to conclude that Grainger is undervalued.

Confusingly, property companies adjust their balance sheets in different ways, despite apparent standardisation under EPRA. For example, the developer Development Securities publishes an 'EPRA' NAV that - very conservatively - does not mark its trading assets to market. Its more flamboyantly run sector peer Helical Bar publishes an 'EPRA' NAV that does. Despite all appearances, the figures are not comparable - a point crucial to comparing sector valuations.

 

Like-for-like figures - like?

Some companies give their accounts a positive spin by hiving under-performing operations off into a 'non-core' division ear-marked for sale or closure. In the case of the banks, these non-core divisions have become known as 'bad banks'. Barclays' 'core' operations generated a return on average equity of 11 per cent in the first half. But add the bad bank in and that return fell to 6.5 per cent.

Like-for-like sales figures are on the face of it more reasonable. They are based on the principle of stripping out operations that have contributed sales to only one of the two periods necessary to calculate a growth figure. For retailers, pubs and restaurants, these are recently opened outlets; for other companies, they are typically recent acquisitions or disposals.

On the face of it, the figures should be hard to game. But a public spat opened between long-term rivals Tesco (TSCO) and J Sainsbury (SBRY) in January over the measurement of their Christmas trading figures. Sainsbury's investor relations team sent an email to City retail analysts accusing Tesco of being 'a bit disingenuous' in reporting 1.8 per cent like-for-like sales growth, because the number included sales made using Clubcard vouchers. Apparently this is not compliant with IFRIC - the IFRS Interpretations Committee. Of course, we have since learned that this was not the only area where Tesco's reporting was disingenuous.

Some companies also come up with their own versions of like-for-like. The serviced-office provider Regus, for example, reports on a 'mature' portfolio of offices it has had open at least two years. Self-storage group Big Yellow has an 'established' portfolio that was fixed as far back as 2007. As ever, the key is to read the asterisks that explain the figures, rather than simply be dazzled by headline numbers.

Companies also often adjust their growth figures for the impact of currency movements. This was a notable feature of the recent mid-year results season, as many groups' first-half results were hit by the pound's rise in the second half of last year. The constant-currency figures clearly give a more accurate sense of business performance. But it is cash earned in sterling terms that funds dividends, so investors would be foolish to ignore the reported growth rates. Moreover, the calculations are usually simplistic figures based on what profits would have been last year if exchange rates had been at the same average levels. Companies with hedging arrangements in place don't typically adjust their adjusted figures to take account of any benefit they accrue from forex moves.

 

Unadjusted or undiluted?

A common source of confusion is that 'basic', when applied to per-share figures like NAV or earnings per share, can have two meanings: unadjusted or undiluted. Basic or undiluted EPS is post-tax earnings (potentially adjusted) divided by the average number of shares in issue during the period under review. This can be diluted by adding to the number of shares those that could be vested under employee stock options. Mercifully, the difference is usually not great.

 

How we adjust figures at Investors Chronicle

The figures in the tables that accompany our results articles are almost always unadjusted and undiluted - in other words, they are the 'basic' or 'statutory' figures published under IFRS. This is to provide consistency and a degree of neutrality. We make very rare exceptions where the basic figures are highly misleading (Grainger being one), which we will always flag in an asterisk.

In the body of the articles, however, we will often refer to adjusted figures if we feel they are more meaningful. The earnings multiple typically given in the IC view will also usually be based on an adjusted forecast provided by an analyst. We try to make this clear wherever it is no too cumbersome. Moreover, we will often use adjusted figures in the tables given beneath tip articles. This is because we are not trying to be neutral: we are arguing a partisan case. We will explain adjustments wherever necessary.

 

Life assurers: an unresolved debate

Life insurance companies are so adamant that IFRS norms don't do justice to their business model that they publish a whole different set of accounts on an 'embedded value' basis. There are two key principles behind embedded value accounting. First, it is meant to smooth profits by stripping out investment returns, which can lead to wild swings in earnings under IFRS. Second, it boosts the IFRS book value by bundling up future earnings from long-term savings and investment products (discounted in the usual way). The resulting adjusted book value is considered the key valuation measure for life insurers.

But the details of embedded value accounting have proved controversial. The European Insurance CFO Forum, a talking shop for insurance finance directors, set norms a decade ago, and since then these have gradually been tightened. In March 2009, in the depths of the financial crisis, Aviva (AV.) and Old Mutual (OML) embraced a shift to a more conservative set of standards called Market Consistent Embedded Value (MCEV), even as Legal & General (LGEN), Prudential (PRU) and Standard Life (SL.) clung to the old European Embedded Value (EEV) standards. The divide lives on, with no uniform embedded-value accounting methodology. Solvency II rules, due for implementation in 2016, may finally force the sector to find agreement.