Acquisition premiums and the private benefits of control
In 2014 the management and strategies of several large publicly listed firms were the subject of high-profile public debates between activist shareholders and the firms’ managers. For example, in September 2014, Starboard Value, a US-based hedge fund, publicly criticised the management of Yahoo! and suggested that it explore a merger with AOL. A month later, Third Point, a US investment fund, stepped up its criticism of Dow Chemical’s performance and strategy. In Japan, Sony resisted calls from Third Point to spin off parts of its operations, until Third Point sold its stake in Sony.
These public debates have taken place against a backdrop of increased shareholder activism. In total, The Economist estimates that since 2009, 15 per cent of the members of the S&P500 have come under ‘attack’ from activist shareholders.1
The adversarial stance adopted by many activist investors has has caused more attention to be paid to the benefits and value associated with corporate control and how the benefits of that control are split between owners and managers. In modern finance a distinction is drawn between ‘ownership’ and ‘control’. Ownership refers to the proportionate right to a share of the net assets held by a company and the income generated by those assets. Control, on the other hand, refers to the ability to direct the company’s operational, strategic and capital allocation decisions.
For many assets, ownership confers control; ownership of a car grants the owner control over how that car is used. However, in corporate contexts the relationship between ownership and control is often less clear.
At one end of the spectrum, a family business might be 100 per cent owned and managed by a single founder or family. In this case the family both owns and controls the company. Such tightly held, family-orientated ownership structures are a common characteristic of Asian corporations. In certain cases, a family might achieve control of a company, or a group of companies, through a complex structure of stock pyramids and cross holdings. They might also enjoy voting rights disproportionate to their rights over the corporation’s cash flows.
On the other hand, in the US and Europe, large companies are typically publicly listed and their ownership is dispersed across a large number of shareholders, each of whom may only own a small fraction of the company’s equity. The management of these companies is delegated to a team of managers who do not own a significant part of the company’s equity and may, or may not, have contracts to align their interests with those of the company’s shareholders. Although, as activist hedge funds have shown, it is possible to wield considerable influence over a company’s strategy even in the absence of a controlling stake, giving rise to a degree of de facto control over the entity.
Between these two poles lie a diverse range of ownership-control structures. In some cases, a degree of control might also be held by third parties. For example, debt covenants can restrict the set of actions available to the company and its management in respect of, inter alia, the payment of dividends, the assumption of additional debt, or the disposal of major capital items. These covenants transfer some elements of corporate control to debt, rather than equity, investors (or their agents, the management).
A complete taxonomy and discussion of the many ways in which control may be distributed among different classes of investors and stakeholders, who may in certain circumstances include suppliers or customers, is too vast to cover in this paper. In the interests of simplicity, this paper assumes that the controlling investor holds a majority of the company’s equity and voting rights and is able to exercise full control over the actions of the company’s management.
The value of control
The value of a commercial asset is a function of the amount, timing and risks of the cash flows that it is expected to generate. In theory, the value of control should be no different – to be valuable, control must allow its beneficiary to: (i) earn additional cash flows; (ii) bring forward the realisation of expected positive cash flows; (iii) push out the realisation of expected negative cash flows; or (iv) reduce the risks associated with the expected cash flows.
In practice, there are many ways in which an investor might realise value from control. These fall into two broad categories. First, control allows the investor to organise the affairs of the company in ways that maximise the value of the company to the benefit of all shareholders. These general benefits of control might arise from steps taken to:
- improve the company’s profit margin;
- increase the efficiency with which the company uses its assets;
- reduce the company’s debt interest rate;
- optimise the capital structure of the company; or
- reduce the company’s tax burden.
Within the DuPont framework, each of these actions would increase the return on the company’s equity and, in theory, its share price.2 In each case, if these strategies are successful they will increase the value of all of the company’s equity, benefiting not just the controlling investor, but also any non-controlling investors. As observed in the Financial Times:3
Companies that are doing poorly are the ones that tend to get targeted [by activist hedge funds]. A well thought out business plan for a company that has become somewhat entrenched in its thinking can result in all shareholders over time becoming substantially enriched.
However, not all of the benefits of control are necessarily shared, or shared proportionally, across all equity investors.
Control might also allow an investor to generate private benefits, possibly at the expense of non-controlling investors. For example, the controlling investor might set the company’s dividend policy to suit its cash flow needs rather than those of other investors. Alternatively, the controlling investor might seek to extract value from the company at the expense of non-controlling investors through ‘self-dealing’, which might include actions such as:
- withdrawing assets from the company at below market value;
- inducing the company to enter non-arm’s-length contracts with parties related to the controlling investor; and
- restructuring the company’s ownership rights to further reduce the influence of, or ‘squeeze out’, non-controlling investors.
Even if the controlling investor does not engage in self-dealing, it enjoys (implicitly valuable) protection from the potential self-dealing activities of other investors.
The extent to which a controlling investor can use these, or other, strategies to realise value from its control is dependent on the specific characteristics of the controlling investor, the asset and its management team.
The availability of general benefits depends to a large degree on the quality of the company’s current operations and incumbent management. If the company in question is already well managed, uses its assets wisely and invests optimally then the additional general benefits that can be generated through control may be relatively small, in the absence of improvements in the efficiency of the company’s capital structure.
Alternatively, if the company has poor management, or is failing to keep up with its competitors, a controlling investor may be able to generate additional value by having the company invest in new projects or undertake cost-cutting programmes to improve its margins. In this case, the value of control will be a function of the expected additional cash flows and the ease with which the controlling investor can implement the necessary changes.
The industry in which the target company operates likely affects both general and private benefits of control. It stands to reason that the potential benefits of management change would be greatest in industries where a company’s performance is particularly sensitive to the strategic actions of management. Dyck and Zinglaes (2004)4 found that more competitive, lower margin industries are associated with lower private benefits, because companies operating in those industries are less likely to survive the inefficiencies introduced by self-dealing.
In many jurisdictions non-controlling investors enjoy some statutory protection from self-dealing. For example, in Singapore, shareholders are protected by well specified accounting standards and disclosure requirements that make it easier for the non-controlling investor to verify the income earned by the controlling investors and identify instances of self-dealing. Shareholders can also seek remedies from the courts if they feel that their interests have been unfairly prejudiced by the actions of the company.
Additional protections might also be enshrined in the company’s articles of association or shareholders’ agreement as super-majority clauses; veto rights or buyout rights.
These protections reduce the controlling investor’s ability to generate private gains at the expense of the non-controlling investors, reducing the value of control. Variations in the protections granted to non-controlling investors, and the enforcement of those protections, suggest that the value of control should vary between jurisdictions; lower values of control should, in theory, be observed in jurisdictions that grant non-controlling investors extensive protections, as compared to jurisdictions with fewer protections for non-controlling investors.
Empirical research bears out this prediction. Both Dyck and Zinglaes (2004) and Nenova (2003)5 found that the premiums paid for control are greatest in countries offering fewer protections to non-controlling investors. Dyck and Zinglaes (2004) also found that the extraction of private benefits was also curbed by extra-legal factors, such as social norms and the ease with which public opinion is mobilised and expressed.
Estimating the value of control
Despite the importance of control to investors, it is the subject of relatively little empirical investigation. This lack of evidence is, in part, a function of the concealed nature of the benefits conferred by control, particularly the private benefits. A controlling investor will only value the private benefits of control when the extraction of those benefits cannot be frustrated by the non-controlling investor. If the extraction of private benefits can be observed, and demonstrated in court, a non-controlling investor may more easily resist a controlling investor’s efforts to appropriate those benefits.6
To overcome this difficulty, researchers have attempted to infer the value that investors place on control by observing the prices paid in transactions that confer control, rather than attempting to estimate the value of control directly. Researchers have identified two ways in which the value of control might be inferred from publicly observable data.
The first method focuses on companies with multiples classes of traded shares conferring different voting rights. Shares with voting rights tend to trade at higher prices than shares without voting rights. This price differential reflects the value of a vote. Researchers assume that a shareholder competing for control would be willing to pay the minority holders of voting shares a premium, over the price of a non-voting share, that reflects the value of the private benefits the controlling shareholder expects to earn.7
This approach suffers from certain weaknesses. Most significantly, at any given date, relatively few companies are the subject of a contest for control. However, the minority holder of a voting share can only realise the value of the vote when control is contested. Therefore, for those companies not subject to an active contest for control, the price differential between voting and non-voting shares will include, among other things, an assumption about the probability that a contest for control will arise.
In addition, many jurisdictions prohibit multiple share classes, which limits the number of countries that can be subjected to empirical analysis. Even where companies are permitted to issue multiple classes of shares, those that elect to do so generally represent a small minority of listed companies. The self-selected nature of the sample raises the concern that the results of such studies may not be applicable to the wider universe of companies that do not issue multiple classes of shares.
The second method draws on the prices paid in acquisitions of controlling stakes in publicly listed companies. Transaction prices are sourced either from privately negotiated acquisitions of controlling stakes from a single vendor, or from instances of control being acquired through an offer made to all shareholders. The value of control is assumed to be captured by the difference between the transaction price per share and the traded share price before the transaction was announced (the transaction premium).
As a purely hypothetical example, suppose Emptor Ltd acquires a controlling stake in Caveat Plc for US$13 per share (the ‘transaction price’). Before the acquisition, the shares of Caveat Plc traded at a US$10 per share (the ‘pre-acquisition market price’), and rose to US$12 per share after the transaction (the post-acquisition market price). In this case, it might be assumed that:
- the total value of control to Emptor Plc was US$3 per share, namely the difference between the transaction price and the pre-acquisition market price;
- Emptor Plc’s control of Caveat Plc was expected to generate US$2 per share of general benefits accruing to all shareholders, namely the difference between the post-acquisition market price and the pre-acquisition market price; and
- Emptor Plc expects to be able to extract US$1 per share of private benefits, namely the difference between the transaction price and the post-acquisition market price.
This approach is also not without potential problems. Most fundamentally, it has been suggested that the premiums observed in the acquisitions of controlling stakes are simply due to systematic overpayment or evidence of a winner’s curse.8 Between 1992 and 2006 BCG estimates that 58 per cent of M&A deals destroyed value for the acquirer’s shareholders.9 Implicitly, positive average ex ante control premiums might conceal average ex post control premiums of close to, and perhaps below, zero.
Dyck and Zinglaes (2004) reject this overpayment hypothesis. They argue that if the observed premiums were attributable to overpayment, one would expect the acquirer’s share price to fall following the announcement of the acquisition. Since, in those cases where data is available, no statistically significant fall in the acquirer’s share price is observed, Dyck and Zinglaes conclude that the premiums paid in acquisitions are not attributable to systematic overpayment.10
Even where overpayment is not a significant issue, it is unclear that transaction premiums can be used to arrive at widely applicable estimates of the value of control. The difference between a company’s pre-acquisition market price and the transaction price might be driven by a number of things unrelated to the value of control.
In some cases the quoted market price of shares in thinly traded companies, may not reflect the contemporaneous price at which those shares would trade on the market. For example, in many small quoted companies a large percentage of the outstanding shares are held by insiders, who are neither buyers nor sellers in the market. In these cases, shares in the company may trade infrequently on the market.
Since the quoted share price reflects only the price at which the company’s shares were last traded, this price may no longer reflect the company’s market value. The transaction premium observed in the acquisition of the company might, therefore, include an ‘update’ to the market price that is not entirely attributable to control.
In addition, many transactions are motivated by factors beyond the target company’s cash-generative capacity. PwC’s 2014 M&A Integration Survey Report found that in 2013 ‘access to new brands, technologies or products’, ‘access to new markets’ and ‘operational synergies’ were each reported as ‘very important’ strategic goals by over 40 per cent of survey respondents.
These strategic goals may generate significant additional cash flows for the purchaser. However, the value of these cash flows will be heavily dependent on the specific characteristics of the acquirer. For example, following its acquisition of F&N, TCC Assets was expected to realise significant synergies from the sale of F&N’s products through the distribution network operated by its subsidiary, ThaiBev. These synergies would not have been available, or at least not to the same extent, to an alternative acquirer.
As a result of perceived synergistic or strategic benefits, the value of the asset to an acquirer may exceed the market price for the asset. An acquirer who is able to derive value from an asset that would not be available to other market purchasers is sometimes referred to as a ‘special purchaser’ who places a ‘special value’ on the asset.11
This gives rise to a distinction between the ‘value’ of the asset to the acquirer and the ‘price’ at which it is willing and able to acquire the asset. Typically, the value of the asset to the acquirer sets an upper limit on the price it is willing to pay for the asset.
The potential vendor may also derive special benefits from ownership of the asset. A transaction will only occur if the value of the asset to the potential acquirer exceeds its value to the owner. The gap between these two values creates a space in which the transaction price can be negotiated. The agreed transaction price will then be a function of the relative bargaining positions of the two parties.
This suggests that the size of an observed transaction premium may be very sensitive to the unique characteristics of the acquirer, and to some extent, the vendor. Care must therefore be taken when applying the results of analysis based on such premiums to valuations in which strategic benefits may, or may not, be present.
More generally, Nath (1990, 1994 and 2011)121314 has argued that in the majority of cases companies trade at close to their control value and therefore the premiums paid in acquisitions do not reflect the benefits of control. He identifies three main reasons for this:
- first, only 3–4 per cent of publicly traded companies are subject to takeovers in any given year. This is much lower than would be expected if gaining control of public companies enabled the acquirer to realise significant additional value;
- second, the premiums observed in acquisitions are a function of the laws of supply and demand. The presence of an investor seeking to purchase a large number of shares represents an increase in demand for those shares. This results in an increase in the share price, which is unrelated to control. It is, however, unclear why a short-term excess of demand for a share would increase the price of that share above the value of that share over the long term; and
- third, to gain control of a public company, it is necessary to convince the holders of the majority of the company’s shares to sell. The price required to convince the majority of the company’s shareholders to sell may exceed the traded share price, which represents only the price at which the marginal shareholder would be willing to sell. Again, this effect is unrelated to the value of control.
It remains debatable whether Nath’s observations imply that the trading price of a share already includes a premium for control, which should be backed out when valuing a minority stake (as Nath suggests), or whether the applicable control premium is small and therefore the value of a share without control is approximately equal to the value of a share with control (as others have suggested).15
A final complication, common to both of the approaches used to estimate the value of control, concerns the related concepts of liquidity and marketability.16 Individual shares in publicly listed companies are generally assumed to be both liquid and marketable. Listing shares of a company provides a mechanism through which the shares can be easily and quickly traded, as well as access to a large pool of potential purchasers.
Generally speaking, large blocks of shares are assumed to be less marketable than small blocks, since fewer purchasers will be able to afford the larger stake, reducing the pool of potential purchasers. In transactions involving large tranches of shares a discount may be applied to the transaction price, to reflect the reduced marketability of the asset.
While a large stake in a company is able to command a premium for control, it is also subject to a discount for lack of marketability. Therefore, unless these two effects are separated, the analysis based on the transaction prices of large blocks of shares may, in fact, understate the value of control.
However, the relationship between control and marketability is not necessarily so simple. A full account of the potential interactions between control and marketability is beyond the scope of this article. However, as a simple example, a 45 per cent stake in a company might lack both control and marketability, since few investors would be willing to make such a significant investment in a company without the protections conferred by control.17
Increasing that stake to 51 per cent would, in most cases, grant the investor control. Since control is a desirable characteristic, the presence of control might make such a stake more attractive to potential purchasers. Thus, in some circumstances, moving from a large, but non-controlling stake in a company to a controlling stake may result in increased marketability as well as increased control.
Control premiums in valuation
The empirical challenges that arise when seeking to estimate the value of control have not prevented valuation practitioners from applying large control premiums, and discounts for lack of control, to their valuations.
Whether an adjustment for control is applicable depends, in part, on the valuation approach adopted. Two of the most frequently used methods of estimating the value of an asset are income approaches, such as discounted cash flow (DCF) modelling, and market-based approaches, usually based on comparable companies’ multiples of (say) profit or sales.
Income approaches to valuation assume that an asset’s value today is equal to the value of the cash flows that the asset is expected to generate in future, discounted at a rate that reflects the risks inherent in those cash flows. In a DCF approach, the company’s expected cash flows are projected over a discrete period, followed by an assumption of a mature state from which a ‘terminal’ or ‘continuing’ valuation is derived.
An advantage of DCF models is that they allow for valuations to be performed using multiple scenarios. Thus, the expected value of changes to the company’s capital structure, investment plans, revenue streams or costs can be explicitly included in the analysis.
In many cases the cash-flow projections and discount rates used in DCF models are assumed to represent the optimal use of the company’s assets and its ideal capital structure. The DCF value, therefore, includes, at a minimum, many of the general benefits of control. Applying an additional premium predicated on the availability of general benefits from control would double count these benefits.
In cases where DCF analysis is used to value a non-controlling stake in a company, it is common to apply a ‘discount for lack of control’ to the pro rata value of the non-controlling stake. This discount reflects that the controlling investor may extract private benefits that reduce the value of the non-controlling investors’ stake to a fair market value that is below its pro rata value.
The applicability of a ‘discount for lack of control’ also depends on the context in which the valuation is performed. It may not, for example, be appropriate to apply a discount for ‘lack of control’ in cases involving the unlawful expropriation or wrongful taking of an asset.
Mathematically, the applicable discount for lack of control is often assumed to be equal to:
1 – (1/(1 + control premium))
However, this relationship potentially fails to distinguish between general and private benefits of control. Since the discount for lack of control may only be applicable in cases where the controlling investor is able to extract private benefits at the expense of the minority, the above formula may overstate the discount for lack of control if the control premium includes general benefits of control.
In practice, many jurisdictions require mandatory offers to be made for all of a public company’s stock once the acquirer’s stake exceeds a certain threshold. The presence of these requirements further complicates the analysis of the relationship between control premiums and discounts for lack of control. The acquirer of a large stake in a public company might be unwilling to pay for the perceived private benefits of control, if there is a risk that it will be obliged to purchase all of the company’s outstanding shares. This might reduce both the control premium and the minority discount implied by the transaction price.
More generally, since discounts for lack of control reflect, in part, the risk that a controlling investor might seek to extract private gains at the expense of the non-controlling investor, they might be more significant in valuations where there is a controlling investor capable of extracting private benefits, or where it is likely that a single investor will gain control of the company.
A market-based approach relies on the market prices of shares in companies sufficiently comparable to the target. There are two commonly used sources of pricing data:
- the prices at which shares in publicly listed comparable companies are traded on the stock market (trading prices); and
- the prices at which blocks of shares in comparable companies are acquired in publicly announced, arm’s length transactions (transaction prices).
Market-based valuations that use trading prices are often assumed to exclude any value of control. Since the ownership of small parcels of shares does not, generally, confer control, it is assumed that valuations performed with reference to the prices of small parcels of shares will also not include the value of control.
However, when applying a control premium to valuations performed using trading prices, care must be taken to ensure that the benefits of control are not double counted. If the comparable companies used to value the target asset are well run, applying an additional premium predicated on some improvement in the target company’s efficiency or growth may double count the benefits of control.18
To date, Asia has seen relatively little public criticism of managers by shareholders. However, there are signs that hedge funds are increasingly interested in the gains that could be made from taking activist positions in Asian corporates. For example, Third Point has recently used its stake in Funuc, the world’s largest robotics company, to call for a share buy-back. In February 2015 Singapore saw the launch of a new hedge fund, EVA Capital SP, whose explicit strategy is to take activist positions in small and medium size construction and engineering companies.
These developments suggest that public battles for corporate control may become increasingly common in Asia. As this occurs, the ability to accurately estimate the value that might be gained from control will become increasingly important.
Attempts to estimate the value of control have been hindered by the difficulties of directly observing the value placed on control by investors in situations where the researcher cannot observe those investors’ plans or expectations. As noted above, there are legitimate concerns that control premiums estimated with reference to the prices paid in acquisitions may systematically overstate the value of control.
Ideally, to overcome these difficulties, the incremental cash flow and risk reduction benefits associated with control should be projected and valued explicitly, with reference to the specific characteristics of the target asset and its acquirer. Regardless of the valuation approach adopted, care must be taken to ensure that the application of a control premium does not double count these benefits.
Where the benefits of control are not susceptible to direct observation, the applicable control premium is left within the judgement of the valuer. The valuer must therefore decide where the company lies within the range, from companies that could command a significant control premium to those that might command no premium. This assessment should be guided by the characteristics of the company itself.
- The Economist, ‘The new masters of the universe’, 9 February 2015.
- CFA Institute, 2007. Financial Statement Analysis, CFA Program Curriculum (2007) Level 1 (Volume 3) (Volume 3). Edition. CFA Institute; p227.
- Financial Times, ‘Shareholder activism: Battle for the boardroom’, 23 April 2014.
- Dyck and Zinglaes (2004): Dyck, Alexander and Zingales Luigi. ‘Private Benefits Of Control: An International Comparison’, Journal of Finance, 2004, v59(2 April), 537-600.
- Nenova (2003): Nenova, Tatiana, 2003, ‘The value of corporate voting rights and control: A cross-country analysis’, Journal of Financial Economics 68, 325–351.
- It should be noted that this resistance may be unsuccessful, particularly where courts are reluctant to interfere in decisions taken by a company’s management, except in the most extreme cases.
- See Nenova (2003) and Dyck and Zinglaes (2004) (Noted 5 and 4, above).
- A winner’s curse can occur in auctions where participants have incomplete information concerning the target asset and the other potential acquirers. In such cases, the asset’s eventual acquirer cannot be sure that it has not overpaid for the asset, since all the other participants in the auction, by definition, considered the asset to be worth less than the acquirer paid.
- BCG Perspectives: ‘The Brave New World of M&A’.
- This analysis relies on the assumption that privately owned acquirers are no more susceptible to overpayment than publicly listed acquirers.
- See, for example, the definitions provided by the International Valuation Standards Council.
- Nath 1990: Nath, Eric, 1990, ‘Control Premiums and Minority Interest Discounts in Private Companies’, Business Valuation Review v9(2), 39-46.
- Nath 1994: Nath, Eric, 1994, ‘A Tale of Two Markets’, Business Valuation Review v13(3), 107-112.
- Nath 2011: Nath, Eric, 2011, ‘Best Practices Regarding Control Premiums: comments regarding the appraisal foundation’s proposed white paper on control premiums’, Journal of Business Valuation, 2011, v2, 25-30.
- See Bolotsky (1991) for a discussion of both positions.
- Although the terms are sometimes used interchangeably, in this article ‘liquidity’ is defined as the ease and speed with which an asset can be converted in to cash at relatively little cost. Marketability on the other hand refers to the size of the potential pool of acquirers for the asset.
- In practice, and depending on (inter alia) the profile of the other shareholders, a 45 per cent stake might confer a degree of de facto control, despite the absence of a majority of the company’s voting rights.
- In practice, one might assess whether the comparable companies are well run by comparing their growth rates, margins and other operational efficiency metrics to industry averages.