EPS growth: Demergers and special dividends

Differences in adjustments to the share count related to special dividends and demergers can impair the comparability of earnings per share. Under IFRS, EPS growth depends on whether a stock consolidation accompanies a distribution. However, stock consolidations, by themselves, have no economic impact and should not affect performance metrics.

In Vivendi’s recent distribution of shares in Universal Media Group, the lack of an accompanying stock consolidation resulted in a discontinuity in per share metrics. However, in a similar distribution by GSK, a stock consolidation produced a very different outcome. We explain the problem for investors and how you can adjust to ensure comparability.

Whenever a company makes a distribution to shareholders, the resources that remain available to the business are diminished and therefore profit is likely to be lower than it otherwise would be. This applies to any distribution, but clearly the effect is most pronounced where the distribution is more than merely a regular payment out of current period profit. Share buybacks, special dividends, demergers and spinoffs can result in a significant reduction in resources and a similar reduction in profit. However, the effect on earnings per share and EPS growth also depends on what happens to the share count.

For share buybacks the share count following the buyback falls by the number of shares repurchased, which means that EPS growth is largely unaffected. EPS can still rise or fall depending on how the change in resources affect profit relative to the change in shares. For example, borrowing or using surplus cash to fund a buyback is often a source of EPS accretion (enhanced EPS growth), considering that the effective price earnings ratio of debt or cash (the reciprocal of the post-tax interest rate) is generally greater than the company’s pre buyback price earnings ratio . Although any accretion should be interpreted after considering the change in leverage, EPS and other metrics pre and post the buyback are comparable.

No automatic reduction in the share count for special dividends and demergers

For special dividends, spinoffs and demergers there is no automatic reduction in the number of shares, which means that these transactions would normally be EPS dilutive and reduce EPS growth.

An extreme example is the spin-off of Universal Media Group (UMG) by Vivendi. In 2021 Vivendi distributed to its shareholders most of its 70% shareholding in the UMG subsidiary. While the company retained a 10% holding in UMG, the transaction represented a significant distribution of resources. Indeed, at the point of distribution, the value of UMG transferred to Vivendi shareholders represented about 67% of the market capitalisation of Vivendi.

Because Vivendi did not do a stock consolidation1A stock consolidation has the effect of reducing, by a fixed proportion, both the aggregate number of shares outstanding and the shareholding of each individual shareholder. There is no accompanying cash payment to investors. to coincide with the distribution, there was no related change in the share count. The small reduction in average shares shown below is attributable to unrelated share repurchases.

Vivendi earnings per share disclosure

Vivendi 2021 financial statements

The effect on Vivendi EPS growth is not readily apparent in the 2021 financial statements, considering the unusual nature of the gains and losses in 2021, including the gain related to the UMG disposal and the fact that the UMG contribution to profit in both periods has been reclassified as ‘discontinued operations’. Nevertheless, it is clear that future EPS for the much smaller group will be significantly different from what it would have been without the distribution and not comparable with prior periods.

However, this confusing change in per share metrics at the time of a significant distribution can be avoided if the distribution is accompanied by a stock consolidation. Vivendi chose not to apply a stock consolidation, but many other companies in a similar situation have done so.

Stock consolidations

Usually, a change in share capital that is not accompanied by a change in resources, including stock splits, scrip issues (including the effective scrip issue component of a rights issue) and stock consolidations, necessitates retrospective adjustment to the share count.

For example, if a company does a 1 for 1 scrip issue and doubles the number of shares, future EPS and other per share metrics will automatically be one half of what they otherwise would be. This reduction in EPS makes no sense given that a change in share count without any consideration has no real economic effect.

To obtain meaningful comparisons, the share count for the period before the date of the scrip issue, including that used for prior period per share figures, should be doubled and prior period per share metrics previously reported halved by applying the scrip factor.

{ \sf Scrip \, factor = \dfrac{Existing \, shares \, ± \, Change \, in \, shares \, for \, zero \, consideration} {Existing \, shares}}

For stock consolidations the effect is reversed and the reduced share count going forward results in higher per share metrics. The adjustment is to retrospectively reduce prior period number of shares by assuming the consolidation had taken place at the beginning of the earliest period reported. However, under IFRS, reducing the prior period number of shares is not done if the stock consolidation accompanies a significant change in resources.

If a capital distribution such as a special dividend or a demerger combined with a stock consolidation has the “overall effect of a share repurchase at fair value”, the normal retrospective adjustments for stock consolidations are not applied. This has the effect of treating the combination of the distribution and stock consolidation as though it were a share repurchase. The result is a reduction in the share count commensurate with the reduction in resources.

Here is exactly what it says in IAS 33.

IFRS – Stock consolidations accompanying a distribution of resources

IAS 33.29: A consolidation of ordinary shares generally reduces the number of ordinary shares outstanding without a corresponding reduction in resources. However, when the overall effect is a share repurchase at fair value, the reduction in the number of ordinary shares outstanding is the result of a corresponding reduction in resources. An example is a share consolidation combined with a special dividend. The weighted average number of ordinary shares outstanding for the period in which the combined transaction takes place is adjusted for the reduction in the number of ordinary shares from the date the special dividend is recognised.

IAS 33 paragraph 29

Incidence of stock consolidations accompanying demergers varies

In the case of Vivendi, the value of the UMG distribution was about 67% of the pre-distribution market capitalisation. If Vivendi had applied a 33 for 100 stock consolidation, and thereby reduced the number of shares by the same 67%, the combined effect of the two transactions is a share repurchase at fair value. Because the reduction in resources is commensurate with the reduction in number of shares, the earnings per share (and other per share metric) comparability problems would have been avoided.   

To be clear there is nothing wrong with Vivendi’s accounting – it complies with IAS 33. Nor is their EPS growth necessarily misleading as long as investors remember the demerger related distribution effect in 2021. However, we think EPS growth would be much more meaningful and easier to interpret had the company done the stock consolidation.

Pharmaceutical company GSK did such a stock consolidation in a similar situation when earlier this year it demerged its Consumer Healthcare business Haleon. The amount distributed to shareholders, and the impact on GSK profits, was lower than in the case of Vivendi; the value of Haleon being only about 20% of the pre demerger value of GSK. Nevertheless, without a stock consolidation adjustment, per share metrics would still have been distorted. Accordingly, the company did a 4 for 5 stock consolidation, effective on the ex-div date for the distribution of Haleon shares to GSK shareholders.

GSK explanation of the Haleon related stock consolidation

GSK press release 18 July 2022

Companies may also apply stock consolidations to special dividends that are either a one-off payment or potentially a significant, but temporary, increase in regular payments. Without the stock consolidation (and the lack of the normal retrospective adjustment to the share count) these distributions may also result in misleading EPS and other per share growth statistics.

UK food retailer Tesco is a recent example. Like Vivendi and GSK, their distribution also related to a business disposal, except in this case their business in Thailand and Malaysia was sold for cash, with the cash itself distributed to shareholders in the form of a special dividend. The dividend represented 21% of the pre distribution value and was accompanied by a 15 for 19 stock consolidation. This resulted in a reduction in share count from 9,629m in 2021 to 7,658m in 2022, as shown in the note below.

Had the normal retrospective adjustment been applied for the stock consolidation, the prior period share count would have been reduced by the factor 15/19 to 7,601m and the reduction in resources due to the distribution would have diluted EPS. However, in accordance with IAS 33, because the combined effect of the distribution and consolidation is effectively a share repurchase at fair value, no retrospective adjustment was applied for the stock consolidation. EPS growth is consequently higher than it would have been without applying the consolidation (although, like Vivendi, the IFRS metrics for this period are affected by the business disposal gain).

Tesco special dividend and stock consolidation

Tesco financial statements 2022

As with demergers, there is no requirement to combine a special dividend with a stock consolidation. Therefore, for other companies you may well see an EPS effect arising from a special dividend.

Although less obvious, regular dividend distributions also distort per share metrics

It is not just special distributions that can cause comparability issues, many companies have regular dividend payments that are significant, and which inevitably affect EPS growth. Many mining and insurance stocks, for example, currently have very high dividend yields, some of which are in double digits. These same companies could pay lower regular dividends and periodically pay special dividends to produce the same overall return of capital. If the special dividends were combined with stock consolidations, a materially different EPS picture would emerge.

Improving per share metrics for investors

Different approaches used by companies to return capital to investors, combined with the varied use of accompanying stock consolidations, produce very different results. IFRS accounting, in the form of IAS 33 paragraph 29, tries to help companies by enabling them to structure a special dividend or demerger as though it were a share buyback for EPS purposes. However, that only really helps investors if all companies actually apply the stock consolidation, which is certainly not the case. Even if consolidations are always applied there is still the problem of a different treatment for regular and special dividends.

IAS 33 paragraph 29 is not as helpful at it might appear

A stock consolidation has no economic effect. Halving the number of shares, for example, with each worth twice the amount than previously, does not change anything.2In some situations, the price per share may matter if the price is very high due to the adverse effect on liquidity. However, this only really applies in extreme situations, and in any case, this is an argument for stock splits not stock consolidations. Therefore whether a distribution is accompanied by a stock consolidation or not makes no economic difference to investors. However, it does make a difference to the reported EPS figure.

A difference in financial statement data should mean there is a difference in the underlying economics; if there is not, then what is the point of the accounting? In our view, having different per share metrics depending on whether or not a company applies a stock consolidation makes little sense.

There is an argument that IAS 33 paragraph 29 should simply be deleted with no special treatment for stock consolidations that accompany special distributions. As a result, the financial effects of special distributions would be the same as regular dividend payments.

Of course, this approach means investors would have to interpret all per share metrics after considering the effect of the distribution, which adds a layer of additional complexity. However, this is already required where comparing companies with different regular distribution policies, for example through the use of total return rather than simply focusing on stock price changes.

However, we think a better approach would be to apply the IAS 33 adjustment to all ‘special’ distributions, irrespective of whether an actual stock consolidation is applied, so that historical trends are more meaningful. This ‘buy-back equivalent’ approach could even be applied to all distributions, including regular dividends, to further aid comparability.

Buy-back equivalent earnings per share

In a buy-back equivalent earnings per share (and other per share metrics), combining the distribution with a stock consolidation is applied to all distributions, including those for which there is no actual accompanying stock consolidation. To do this the distribution is combined with a hypothetical stock consolidation to produce the equivalent buyback, with an offsetting hypothetical stock split applied to offset the consolidation.

There is no retrospective adjustment for the hypothetical stock consolidation, in accordance with the approach in IAS 33. However, the offsetting hypothetical stock split is treated in the normal manner, where shares in issue and per share metrics prior to the distribution date are retrospectively adjusted. The hypothetical adjustments have no effect on the share count, but they facilitate the necessary retrospective adjustment to ensure comparability in per share growth metrics.

A buyback equivalent EPS

The Footnotes Analyst

Applying this to the Vivendi distribution we discussed earlier, the hypothetical stock consolidation is 33 for 100 to reflect the distribution of 67% (approximately) of pre-distribution value. The hypothetical scrip issue is therefore the reverse 100 for 33. No scrip factor adjustment is applied for the stock consolidation, but prior period shares are increased by the scrip factor of 100/33 to retrospectively recognise the hypothetical stock split. This increases the share count for the period prior to the distribution and consequently reduces per share metrics for that same period. The result is comparability over time.

Amending IAS 33 paragraph 29 would help investors

In our view, it would help investors if paragraph 29 in IAS 33 were amended so that all special distributions were treated as share buybacks, irrespective of whether an actual stock consolidation is done or not. This would result in more comparability and resolve the problem of some companies doing consolidations and others not, as illustrated by our analysis of Vivendi and GSK.

We also think that investors should consider applying the buy-back equivalent approach to per share metrics for all distributions, including regular dividends. Obviously, its application to all distributions is more burdensome, but if you are analysing a sector with high but also varied dividend yields it may be worth the effort.

Stock prices and other per share metrics

The problem we identify for EPS also affects other per share metrics such as dividends per share, book value per share and even the share price itself. Like EPS, the problem does not affect current (post distribution) metrics but rather the comparability with historical data and hence the relevance of growth.

Because IFRS does not have any measurement requirements for per share metrics other than EPS, whether and how adjustments are made for so-called ‘stock events’, including stock consolidations, is up to individual investors and particularly the data providers. We believe that data providers follow the IAS 33 approach for changes in shares unrelated to a change in resources, including applying the same retrospective adjustments. However, it seems that data providers make different choices in the case of Vivendi where the significant distribution was not accompanied by a stock consolidation.

We have not investigated all the per share metrics for Vivendi but have focused on historical share price data to illustrate the divergence in practice.

Data providers have made very different adjustments to Vivendi historical stock prices

Some data providers do not seem to have made any retrospective adjustments at all. Consequently, the historical stock price chart for Vivendi shows a significant fall on September 21, 2021, the ex-div date for the UMG distribution. The following chart available on NASDAQ.com illustrates the ‘no adjustment’ approach.

nasdaq.com – Vivendi 5-year historical stock price chart


However, other data providers do make adjustments to historical prices prior to the UMG ex-div date. This chart provided by Euronext covers the same 5-year period as above.

EuronextVivendi 5-year historical stock price chart


In the Euronext chart, the stock price immediately before the ex-div date has been restated from the €31.53 that was originally reported (and shown in the NASDAQ chart) to €13.03, a reduction of 58%. Surprisingly this still results in a price movement on the ex-div date where the market opening price is €10.35. While a price difference is to be expected, we question whether this scale fairly reflects the economics and hence whether the adjustment for prior periods accurately reflects the UMG value distributed. We do not believe that the fall in price from €13.03 to €10.35 reflects an economic change in that overnight period or that the market is so inefficient. We can only assume that the adjustment factor applied is incorrect.

Indeed, other data providers who adjust Vivendi stock prices before the UMG ex-div date have applied very different adjustment factors. For example, Bloomberg, Barrons, and the Wall Street Journal, all make much larger reductions. In the case of Bloomberg, the price immediately before the ex-div date (was reduced to €9.17, a 71% reduction, with this adjustment factor also applied to all prior prices. However, for the Wall Street Journal and Barons the equivalent adjusted price is €6.28.

In all three cases a positive price change (which we also do not think is realistic) is reported on the ex-div date and the historical price chart looks very different. We show the Bloomberg chart below.

Bloomberg – Vivendi 5-year historical stock price chart


Different data provider approaches to prior period per share adjustments, due to special distributions such as demergers, is confusing and problematic for investors. For example, incorrectly adjusted historical prices can result in misleading signals for momentum investors and incorrect beta factors where these are based on historical stock price regressions. Just one severely mis-stated daily price change can lead to a distorted historical beta.

We know from experience that managing stock events in a database of per share metrics is challenging, but at least for most such events the adjustments that are required are clear. Data providers can simply follow the rules laid out in accounting standards for EPS. However, in the case of demergers and special dividends, it seems to us that the accounting is deficient, which results in data providers adopting their own approach. Applying our ‘buy-back equivalent’ method, as we explain for earnings per share, would help in achieving consistency for all per share metrics. Even better would be for the IASB to mandate this approach in IFRS.

Insights for investors

  • Significant distribution of resources can have a distorting effect on per share metrics, including EPS and the stock price.
  • Stock consolidations combined with a distribution create the overall effect of a share repurchase at fair value. Such consolidations are exempt from the retrospective adjustment that is otherwise applied.
  • Companies can choose whether to apply a stock consolidation to offset the negative per share impact of distributions which leads to a lack of comparability.
  • A buy-back adjusted EPS and other per share metrics can be calculated by using a technique that involves a hypothetical stock consolidation and scrip issue. Buy-back adjusted metrics are more comparable.
  • A buy-back adjusted approach can be applied to all special dividends and demergers or can be applied to all distributions, including regular dividends.
  • Beware of historical per share data given by data providers, including stock prices. These may be calculated very differently – particularly where special distributions are not accompanied by a stock consolidation.

Don’t miss new articles published by The Footnotes Analyst – subscribe here:


No advertising, no spam, just new articles. Unsubscribe anytime.

Send us a question or comment about this article:

    For related articles select subject tags or see the suggestions below:

    Print or save article as a pdf: