Most deferred tax adjustments in financial statements help investors – but not always. The ‘economic value’ of deferred tax assets arising from unused tax losses may be significantly less than the balance sheet figure. However, as a consequence, profit forecasts may be understated, potentially leading to an undervaluation by investors.
We estimate that if the £24bn deferred tax asset of Vodafone were discounted to an economic value then it would instead be closer to £8bn, but forecast profit would rise by about £500m.
We believe that deferred tax assets are often overstated in financial statements and that, as a consequence, post-tax profit metrics used by investors can be understated. For example, Vodafone recognises a substantial deferred tax asset due to tax loss carry forwards that is equal to about 70% of its market capitalisation. This asset is reported in accordance with IFRS and we do not question the figures presented by the company. However, in our view, the application of the accounting standard for deferred tax fails to faithfully represent the true economic value of the asset. We estimate that the Vodafone deferred tax asset reported at £24bn in its March 2019 financial statements has an economic value closer to £8bn. The difference is due to the time value of money – the accounting standard does not permit adjusting for the timing of the tax loss benefit.
The consequence of overstating deferred tax assets is that, when this asset first arose, profit was overstated (or more likely a loss understated). Large tax loss carry-forwards, such as that for Vodafone, often arise due to major asset impairments and associated write downs of investments. Recognising a higher accounting deferred tax asset and associated tax gain at this time would understate the extent of this loss on a post-tax basis. (Actually, for Vodafone, the tax benefit of the write down was only partially recognised when the loss occurred, with more being recognised some years later due to estimate changes – we explain more about measurement uncertainty below.)
Measuring the deferred tax asset of Vodafone at economic value could increase reported earnings by £500m
However, the effect of this initial overstatement is that profit is understated in subsequent periods when the deferred tax asset is reversed and charged as an expense. For Vodafone the current analyst consensus net profit forecast for 2020 is about £2.6bn. We estimate that this could be higher by as much as £500m if the deferred tax adjustments were to more faithfully represent the underlying economics.
We will return to Vodafone below, but first let’s start with an explanation of deferred tax asset accounting and why we think there is a problem for investors.
Understanding deferred tax assets
Deferred tax often seems to be a source of great confusion for investors, but it should not be. Deferred tax accounting is in principle relatively simple. It is, in effect, the application of accruals accounting to the tax expense. The current period tax charge reflects what is taxable in that period. But, for some transactions, gains and losses reported in financial statements in one period affect taxable profits in a different period. Without any adjustment for this effect the tax charge would be difficult to interpret and post-tax measures of performance potentially misleading.
Deferred tax adjustments include those related to unused tax losses and allowances. A deferred tax asset arises because, when an accounting loss is recognised in the financial statements, the tax benefit of this loss may not be immediately available. The tax loss is carried forward to be offset against future taxable profits. The deferred tax adjustment results in a tax gain being reported in profit and loss in the period the loss is made. In addition, the company recognises a related asset which represents the future cash tax saving that will be obtained once the loss is utilised. The benefit of the loss is reported in the period the loss is ‘earned’, i.e. accruals accounting.
This is all fine. The problem is how deferred tax assets are measured in financial statements and, in particular, the treatment of uncertainty and the time value of money.
Probable, but no probabilities
Under IFRS, specifically IAS 12, deferred tax assets may only be recognised in the balance sheet when their recovery is probable. In other words, the asset is only recognised if the company assesses that it is probable there will actually be a future tax saving. The word ‘probable’ is interpreted as ‘more likely than not’ under IFRS, therefore a deferred tax asset could still be recognised even if there is some uncertainly over its recovery.
The use of probability in deferred tax asset measurement acts as an on/off switch. If realisation of a future tax saving is ‘probable’ then the asset is recognised in full at the amount of the available losses multiplied by the tax rate. This applies even if the likelihood of being able to fully utilise those losses is less than 100%. If a company judges that the future utilisation of a loss is unlikely then the deferred tax balance is zero even though there may still be a possibility of a future tax saving.
A portion of the potential asset could still be recognised. If, for example, a company expects that it will only earn sufficient profits to utilise 60% of the available tax losses this determines the amount recognised. However, the calculation is still an on/off switch in respect of that amount, based on whether probable or not.
In economic terms, and for the purpose of equity valuation, the probability of tax loss utilisation does matter. A business where it is certain that a deferred tax asset can be realised as an actual cash tax saving is clearly more valuable than if the probability of realisation is only, say, 60%; however, the deferred tax amount in the balance sheet would be the same. In practice, this could result in the balance sheet amount being less or more than the economic value.
No time value of money
Deferred tax assets arising from tax loss carry forwards are not discounted. If realisation of those losses takes place over several years, the economic value of the deferred tax asset may be significantly below the accounting figure in the balance sheet – which we believe is the case for Vodafone
We should note that the lack of explicit discounting does not affect all deferred tax amounts. While no deferred tax balances are explicitly discounted under IAS 12, for some deferred tax discounting is implicit. This arises where the deferred tax is based on a balance sheet amount that is itself a present value. For example, pension liabilities often result in deferred tax assets and, because the pension liability a present value, then automatically so is the deferred tax asset. Due to the implicit discounting the deferred tax asset for pensions would reflect the economics, unlike that arising from tax loss carry forwards.
Uncertain measurement
The measurement of deferred tax assets and the application of the ‘probable’ criterion is one of the key judgements in producing financial statements. The issue for investors is that, in spite of the clear definition of ‘probable’, it is inevitable that there will be some diversity in how these judgements are applied. For example, while there is no cut off specified in IFRS, we are aware that some companies only recognise deferred tax assets if the losses are expected to be utilised within a certain number of years, on the basis that it is too uncertain forecasting further out.
Vodafone does not impose such a restriction and has deferred tax assets recognised in respect of some losses it says will be utilised over a period up to 60 years. We are not saying that this is wrong, we think it perfectly possible to assess recovery is ‘probable’ even if over an extended period. Our point is that, in practice, the degree of conservatism applied to deferred tax asset measurement varies and this can have a material impact on metrics used by investors.
The deferred tax valuation problem – and a solution
If a deferred tax asset is fully recognised (because its realisation is deemed probable) then the full tax benefit of that loss is reflected in the period the loss is incurred. In subsequent periods there is a cash flow benefit as cash taxes are reduced, but not a profit and loss benefit due to the reversal of the deferred tax asset. This means that where valuations are driven by post tax accounting profit, such as in a PE ratio analysis, there is the potential to understate value.
The solution is to measure these deferred tax assets at something closer to their economic value. This means allowing for the probability of utilisation and discounting to a present value. The consequence would be to recognise an ongoing benefit in the income statement equal to the accretion in order to unwind the discounting effect. This income statement benefit captures the economic value of the tax losses in post-tax profit measures and improves their relevance.
How does this apply to Vodafone?
Vodafone is a good illustration of the challenges of deferred tax accounting for investors. The 2019 balance sheet includes a deferred tax asset of £24.8bn, of which the vast majority, £24.2bn, relates to unused tax losses. The analysis of this asset is clearly significant for a company with a market capitalisation of about £35.0bn. Vodafone provides good and comprehensive disclosures about its tax position, including the tax losses and associated deferred tax asset. Nevertheless, the reporting must be in accordance with IFRS and, as we explain above, the problem for investors is that the balance sheet amount for deferred tax asset may not reflect the true economic value.
Firstly, there is the issue of the probable recognition criterion. As can be seen in the disclosure below, the full potential tax saving due to unused tax losses is £32.4bn, but of this £8.2bn is assessed by the company as not meeting the probable criterion and is therefore not recognised, which leaves the balance sheet asset of £24.2bn. But this is not a true expected value because the amount not recognised probably does not have a value of zero and, conversely, the amount recognised is probably not 100% likely to be realised. In economic value terms it is the expected value of the tax saving that matters, which could be higher or lower than the balance sheet value. The problem is that because IAS 12 and the related disclosures operate solely based on a ‘more likely than not’ on/off switch it is not possible for investors to identify the true expected value.
Vodafone analysis of deferred tax balances
But it is the second problem we identify, the lack of discounting, that we think is probably more significant in the case of Vodafone. Much of the tax saving that the company will realise is expected to take place many years away. The reason is that the tax losses are held in subsidiaries (remember that it is individual companies and not groups that pay tax) where the level of profitability is low relative to the accumulated losses available to offset those profits. In some cases, the company believes that recovery is unlikely, particularly where the loss carry-forward is time limited, but most of the losses are expected to be utilised. Unlike some companies Vodafone does not impose an artificial cut-off in terms of the time period over which recovery can be assessed as probable.
Recalculating the deferred tax asset to allow for the probability of recovery may be impossible. However, we can estimate the effect of discounting and produce a present value that should approximate the true economic value of the tax loss asset. Our calculation is still an estimate because, although the company discloses the approximate recovery period for the major components of the deferred tax asset, we don’t know the pattern of that recovery. We also have to estimate a discount rate, although we get some help in this from the Vodafone disclosures accompanying their asset impairment tests.
Our estimate of the economic value of the deferred tax asset arising from deductible tax losses at 31 March 2019 is £7.7bn, compared with the balance sheet carrying value of £24.2bn. To calculate the economic value we assumed that tax losses are utilised on a straight line basis over the period specified by the company and applied an estimated weighted average cost of capital of 6.5%
Deferred tax asset overstated but, perhaps more importantly, earnings understated
Although the use of undiscounted deferred tax leads to an overstatement of assets and shareholders’ equity, it also leads to an understatement of net income. Under existing accounting rules, the benefit of the available tax losses and the future cash tax savings is entirely omitted from post-tax profit measures if the related deferred tax asset is recognised in full. But if deferred tax were discounted then the income statement would include an ‘accretion’ of the asset due to the discount unwind. This is a very real economic gain and should, in our view, be part of overall performance.
Based on the assumptions we use in our present value calculation, the accretion that should be added to the first-year forecast earnings of Vodafone for 2020 would be about £500m. Current consensus earnings for 2020, which would exclude this benefit, are about £2.6bn. Therefore, including the accretion gain from the deferred tax asset would increase consensus forecast earnings to about £3.1bn and reduce the prospective PE for Vodafone from 13x to 11x.
In our opinion, these revised figures better reflect the economic value of the tax losses available to Vodafone. Of course, we are not necessarily saying that the market is omitting the tax loss benefit entirely. Investors should be well aware of the value of tax loss carry forwards and there are other techniques for including them in any analysis.
Insights for investors
- Remember that, under accounting rules, deferred tax assets are not necessarily a good reflection of the economic value of future tax savings. The amount could be understated but is probably more commonly overstated, particularly where recovery is over an extended period.
- Be aware that overstated deferred tax assets would have resulted in an overstatement of profits in the past (probably an impairment was worse than actually reported) but produces an understatement of profits in the future.
- Where material, consider adjusting deferred tax assets by discounting expected future cash tax savings. Also adjust forecast net income for the discount unwind.
- Look out for the effect of estimate changes. Effective tax rates and profit trends are best analysed by excluding the impact of deferred tax catch-up adjustments. Make sure that forecasts are based on the underlying tax and not actual reported tax that is affected by those catch-ups. Many companies provide useful disclosures related to alternative performance measures (non-GAAP measures) that help.
- Remember that the degree of conservatism applied in deferred tax asset measurement has a knock-on effect on future profitability. The uncertainty in deferred tax asset measurement also provides earnings management opportunities for companies so inclined.
- If you use pre-tax enterprise value multiples such as EV/EBIT or EV/EBITDA then you should separately consider taxation effects. Tax losses are best regarded as a separate ‘non-core’ asset and deducted from enterprise value in deriving these multiples.
- For DCF analysis, because the focus is purely on cash flows, there is less of a problem as deferred tax is non-cash. However, a forecast of tax cash flows requires the modelling of the utilisation of tax losses. In this regard the disclosures arising from deferred tax accounting are still important.