This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Oxley Act in the US). The merits of such flexibility are thought to lie in its ability to encourage companies to adopt the spirit of the Code, rather than the letter, whereas a more statutory regime would lead to a \box-ticking" approach that would fail to allow for sound deviations from the rule and would not foster investors' trust. Therefore the \Comply or Explain" model ultimately would lead to better governance and its basic premise has been adopted by several other countries (like Austria and Germany). This paper takes stock of the Combined Code's achievements in the UK; in particular, it asks whether it led to too much compliance and too few explanations.
We find that the Code fostered compliance, especially in areas not covered by its forerunner, the Cadbury Code. For example, such provisions include the appointment of a senior independent non-executive director or 12 months service contracts for executive directors. It is encouraging to see that more than half of the non financial constituents of the FTSE350 were fully compliant with the Code at the end of 2004. In addition, we found that on average less than 10% of all companies were not complying with a given single provision.
However the picture looks less rosy when looking at those companies that did not comply with the provisions of the Combined Code. We find that the companies that did not comply with the Code often did a very poor job explaining them. Even worse, in almost one in five cases of non-compliances, companies did not explain their non-compliances at all. When an explanation is provided, most of time it fails to identifies specific circumstances that could justify such a deviation from the rule. Companies that do not comply tend to stick with the same (poor) explanation until they directly jump to compliance. Once compliant, either a company remains compliant or if it ceases to, does not provide good explanations as to why this is the case.
This suggests that companies do not use the flexibility of the Code to fine-tune their governance to their changing circumstances. Rather, companies often seem to make a fundamental choice between compliance and non-compliance.
Interestingly, shareholders seem to be indifferent to the quality of explanations, while the
provision of specific explanations is in fact a way to identify good investment strategies, better than simply focusing on compliance. Returns on a portfolio of compliers do not exceed, significantly, those of non-compliers. In contrast, returns of non-compliers differ significantly according to the quality of explanations.
In light of those findings, we identify areas where the Code could be strengthened with greatest potential benefits:
Provisions pertaining to the minimum percentages of non executive directors (at the
time, one third) and independent non executive directors have the highest frequencies of compliance, above 95% in 2004. At the same time, in relation to these provisions, noncompliant companies are the most likely to give no explanation and the least likely to identify specific justifications. The latest revision of the Combined Code chose to make those provisions more stringent. This will presumably impact on compliance levels, whether the quality of justifications of non-compliance will improve remains to be seen. An alternative route would have been to make those provisions compulsory for the remaining 5% of companies not already complying (and not explaining much).
Ways to foster shareholder's attention to explanations have to be found. Section 2
of the revised code contains main and supportive principles demanding: "Institutional shareholders should consider carefully explanations (and) should give an explanation to the company in writing where appropriate if they do not accept the company's position. They should avoid a box-ticking approach." If possible, the inclusion of a provision in relation to this principle could have potentially significant benefits. The two overall message to be conveyed is that full compliance may not be desirable and that therefore explanations have to be analysed to identify the circumstances where non compliance is in fact a superior way to govern a company.
2. The Combined Code
The Combined Code was in operation from 31st December 1998 to 31st October 2004. Its
provisions are summarised below.
Chairman and CEO (A.2.1) (CEO/COB): There are two key tasks at the top of every public company - the running of the board and the executive responsibility for the
running of the company's business. There should be a clear division of responsibilities at the head of the company which will ensure a balance of power and authority, such that no one individual has unfettered powers of decision.
Senior Non-executive Director (A.2.1) (SNED): Whether the posts (Chairman and CEO) are held by different people or by the same person, there should be a strong
and independent non-executive element on the board, with a recognised senior member other than the chairman to who concerns can be conveyed.
Non-executive Directors (A.3.1) (NEDs): Non-executive directors should comprise
not less than one third of the board.
Independent non-executive directors (A.3.2) (Ind. NEDs): The majority of non executive directors should be independent of management and free from any
business or other relationship, which could materially interfere with the exercise of their independent judgment.
Service contracts (B.1.7, B.1.8) (Ser. Cont.): There is a strong case for setting
notice or contract periods at, or reducing them to, one year or less.
Nomination committee (A.5.1) (Nom.): Unless the board is small, a nomination
committee should be established to make recommendations to the board on all new board appointments. A majority of the members of this committee should be non-executive directors.
Remuneration committee (B.2.1, B.2.2) (Rem.): Companies should establish a formal and transparent procedure for developing policy on executive remuneration and for fixing the remuneration packages of individual directors. No director should be involved in deciding his or her own remuneration. Remuneration committees should consist exclusively of independent non-executive directors.
Audit committee (D.3) (Audit): The board should establish an audit committee of at least three directors, all non-executive, a majority of whom should be independent,
with written terms of reference which deal clearly with its authority and duties.
Director's re-election (A.6): All directors should be required to submit themselves for re-election at regular intervals and at least once every three years.
Pay-linked to performance (B.1): Levels of remuneration should be sufficient to
attract and retain the directors needed to run the company successfully, but companies should avoid paying more than is necessary for this purpose. A proportion of executive directors' remuneration should be structured so as to link rewards to corporate and individual performance.
Internal Control (D.2.1): The directors should, at least annually, conduct a review
of the effectiveness of the group's system of internal control and should report to shareholders that they have done so.
The Combined Code was reviewed by Sir Derek Higgs in his January 2003 report that suggested changes to some provisions and more stringent criteria for others. The changes incorporated into the Revised Combined Code, came into force in November 2004. Examples of stringent criteria are the additional eight contingencies defining the notion of independence and new provisions that require boards to review their own performance.
It was also reviewed in 2006 and suggested few changes to existing provisions and introduced the new provisions. The new provisions incorporated were guidance on liability of non-executive directors and disclosure of corporate governance arrangement.
The data comes from Arcot and Bruno (2005a). They analysed 245 UK non-financial companies, belonging to the FTSE 350 index as of 31st December 2003 between 31st December 1998 to 30th June 2004 the period the Combined Code was in operation. Financial companies were excluded because the overall regulatory environment for those companies differs significantly from that of non financial companies. It is felt that those regulations, although not part of the Combined Code, may well interact with its provisions, and have implications for corporate governance and performance evaluation.
The dataset contains all relevant information from the corporate governance statements and directors' remuneration reports, included in the annual reports of each company for each year. This constitutes a sample of 1; 470 (6 times 245) potential company-year observations of which 1; 286 are available. In turn this gives a maximum number of 10; 288 (8 times 1; 286) principle-company-year observations that are available. The missing observations are either due to annual reports not being available or companies not listed at the beginning of the sample. The dataset provides detailed information with regard to the company's compliance with each of the eight provisions of the code and the actual explanations in case of non-compliance. Arcot and Bruno's dataset also contains information about the number, category and independence of directors, number of meetings and membership of various board committees and the number of pages (and number of words in some cases) of the corporate governance and remuneration reports.
The Level of Compliance.
The maximum number of instances of compliance in the sample is 10; 288 and we actually
find 8; 712 cases of compliance (resulting in an overall frequency of compliance of 84:7%). We first look at the evolution of overall compliance over time and per company. Figure 1 shows that compliance increases every year over the time period, even though full compliance by all companies is not achieved.
Figure 1(to be provided later)
There are 44% of companies which do not comply in 2004 on at least one of the principles
compared to more than 90% in 1998. In 2004, there are 14% not complying on more than 2
principles and 5% not complying on more than 3.
Which is the code provisions driving non-compliance? Figure 2 shows compliance levels for
each principle at the beginning (December 1998-June 1999) and at the end (July 2003-June
2004) of the sample.
Figure 2(to be provided later)
On average, overall compliance per principle increases from 76:7% in 1998 to 91:4% in 2004.
Compliance increased for all principles, except for that relating to the composition of audit
committee, where it remained approximately the same3. Two out of eight principles have the maximum increase in compliance over that period: service contracts (from 35% to 80%) and senior non executive director (from 57% to 92%). Interestingly, these are the two principles which were either not present in the Cadbury code (SNED) or present at different levels (3 year length of service contract instead of 1 year).
Finally it is worth mentioning that no industry pattern can be detected but those levels of
compliances are higher for FTSE100 companies.
The Quality of Explanations.
In this section we investigate the quality of explanations. This requires some qualitative
judgement on our part and necessarily contains some subjectivity. In order to maximise
objectivity and clarity, we distinguish three categories.
The first category is the most objective and easiest to verify: we identify the companies that
do not comply and provide no explanation. This is the worst category if the quality of an
explanation is to be judged by its ability to justify why non-compliance may be warranted.
The absence of any explanation is certainly not in spirit of the code, and it is not even in
accordance with the letter because the code requires a narrative explanation to be given in
case of non-compliance. Surprisingly, the absence of any explanation is quite common in the
sample. For the entire period, almost one in five instances of non-compliances (on average
17%) are not explained. This percentage remains relatively constant throughout the period,
with a slight decrease towards the very end, as illustrated in Figure 3.
Figure 3(to be provided later)
There is some variation in the frequency of the lack of explanations across principles (arrived at by taking averages over time). Figure 4 highlights that two contrasting trends can be seen.
First, three principles (those requiring one third of non executive directors, a majority of independent non executive directors, and remuneration committee consisting only of independent non executive directors) have more than 25% of their non-compliances not explained. Second, deviations from two principles (those requiring a majority of independent non executive directors on audit committee and nomination committees) are explained in more than 90% of the cases. However, all principles suffer from the absence of explanations to some extent.
Figure 4(to be provided)
We now discuss what constitutes good explanations and here some subjective judgement comes into play. We distinguish between good and bad explanations by searching for the presence of verifiable and specific elements relating to the company's circumstances in the narrative statements. The following examples illustrate our typology.
Examples of explanations that we consider to be of good quality are for instance:
The Board has not at present formally appointed a senior independent director, other than the Chairman, to whom concerns can be conveyed. Three new non-executive directors have been appointed within the last 12 months, and it is considered that the Board should be given time to settle into its new composition prior to taking such a decision (BBA 1998).
In determining its overall policy in respect of service contracts, the Committee aims to balance the costs associated with any early termination provisions with the need to protect GlaxoSmithKline's intellectual property rights. The Committee maintains a close watch, through its advisors, on trends in contractual terms amongst other
companies in the competitor panel and in the wider market place.
It is committed to ensuring that, in achieving this balance, its processes are fair, while limiting as far as possible the scope for `rewarding failure'. The Committee has considered the recent guidance produced by the Association of British Insurers and the National Association of Pension Funds in the UK. It will take this into account, alongside market practice, when reviewing contractual terms.
Executive Directors are employed on service contracts under which the employing
company is required to give 24 calendar months' notice of termination and the Ex-
executive Directors are required to give 12 calendar months' notice.
Executive Directors' service contracts contain `garden leave', non-competition, non-
solicitation and confidentiality clauses.
The Remuneration Committee currently believes that one year contracts would not
be in the best interest of GlaxoSmithKline with regard to offering a globally compet-
itive overall remuneration package and securing maximum protection for its intel-
lectual property rights.
The Remuneration Committee believes that the current termination payments due
under Executive Director's contracts are justified because they represent fair and
reasonable compensation in the event the contracts are terminated, given market
practice and the associated restrictions arising from the need to protect intellectual
property. (GlaxoSmithKline, 2002).
In contrast examples of explanations that do not identify special circumstances are as
The Board has not identified a senior independent non-executive director, as specified by the Code, because it considers such an appointment to be unnecessary at present (Reuters 1999).
The board believes that this arrangement (i.e. service contracts superior to 12 months) is in the best interests of the company (Rentokil Initial 1998).
Although Mr. Wilson has the combined role of Group Chairman and Chief Executive, the Board considers that the requirements of the Code are satisfied and that the combination of these roles does not work to the disadvantage of the Company or its shareholders (Wilson Bowden plc 2001).
The company ensures that it recruits to the board only individuals of sufficient calibre, knowledge and experience to fulfil the duties of a director appropriately. There is no formal training programme for directors (A.1.6).
The company does not have any non-executive directors on the board (A.2.1, A.3.1,
A.3.2, A.6.1). The directors are mindful of the provisions of the Combined Code in
this regards and regularly reviews the situation.
The company's nomination committee is made up of the chairman and managing
directors. There are no non-executive members on the committee (A.5.1).
The company does not have a formal remuneration committee (B.1.1-3, B.1.9-10,
B.2.1-6, C.2.3) but the emoluments of the directors are the subject of appraisal by
the chairman and the managing directors taking into account individual
performance and market conditions.
The company does not have an audit committee (C.2.3, D.3.1, D.3.2) but the board
as a whole regularly monitors internal controls and also ensures that an objective
and professional relationship is maintained with the auditors. (W.M. Morrison2004).
Our classification of explanations is simple to implement and easy to replicate. We do not make any judgement as to whether the explanations provided are valid from a business perspective.
So, in that respect, our identification of good explanations can be termed optimistic. This
classification method gives us two further groups of explanations that we refer as "specific
There is a noticeable time trend that can be observed in the evolution of the quality of
explanations as illustrated in figure 5.
Figure 5 (to be provided later on)
There is an inflexion point after the beginning of 2001. From then onwards, we observe a
decrease in general non specific explanations and an increase in specific explanations. This
could reflect the fact that after the fall of the market and the end of the technology boom,
increased attention was paid to governance issues. This may have been further reinforced by the collapse of Enron and WorldCom in 2001-2002 creating more pressure for better quality explanations. However, as compliance also increased significantly over the period this possible improvement could be the result of a decreasing number of non-compliances.
The incidence of specific explanations (containing specific and verifiable circumstances) varies significantly when compared across principles. Figure 6 provides the average percentages of specific, general and no explanations for each of the eight principles.
Figure 6 (to be provided later on)
Explanations for non-compliance with the senior non executive director provision perform the best in the sample. Overall, they have the lowest percentage of general explanations (45%) and one of the lowest percentages of no explanations, so that in more than 40% of the cases of non-compliance the explanations are good. For the remaining principles, the frequency of good explanations is roughly about 20% except for nomination committee where good explanations appear only in 14% of the cases. These results have to be judged, in light of the fact that there are fewer companies which are not compliant with provisions relating to the audit, nomination and remuneration committee as compared to the provisions relating to the senior non executive director or to the length of service contracts.
We believe that one important determinant of the quality of explanations is their diversity.
For instance, in the case of the designation of a senior non executive director, companies offer a variety of circumstances to justify non-compliance: some companies point to the risk of division in the board, others to the existence of a strong non-executive presence on the board etc.
In contrast, explanations regarding composition of remuneration committee many simply
assert that this would not be in the company's interest or claim that the Chairman or the
CEO should serve on the committee (without stating further reasons). The role of the diversity of explanations is also important in our view to assess the overall value of the "Comply or Explain" approach as we will argue later.
In summary, when analysing any one of the eight principles we find that out of one hundred
company-year observations, roughly ninety comply. Out of the remaining ten cases, two do
not provide any explanation and with the exception of the senior non executive director and
audit committees, there are then six instances of unconvincing explanations. Pertinently, we
identify approximately two cases of genuine explanations. Furthermore there is possibly a
positive time trend with regard to the quality of explanations as the percentage of specific
explanations improve after 2001, but on diminishing non-compliances.
The Shareholders' Attitude
The fact that non-compliance is low, or that average explanations are poor does not imply
those explanations are irrelevant. The quality of explanations does have a significant economic impact. We demonstrate this by performing the following simple exercise: we form a first portfolio in any one year comprising of companies that fully comply or, if they do not, provide specific explanations. Similarly, we form a second portfolio consisting only of non-compliers that either do not give any explanation or provide general explanations. The first portfolio on average returns 5% more than the second one on an annual basis.
More strikingly, Arcot and Bruno (2005b) perform the same exercise, where in addition adjustments are made for differences in risk characteristics. They first group stocks on the basis of company market capitalisation and book to market ratios (i.e. book value of equity divided by market value of equity). Then, for each of those groups the difference in returns,
between a portfolio comprising the compliers or the non-compliers with specific explanations, and a portfolio of non-compliers with general or no explanations, exceeds 12% annually. This difference is statistically significant5. Further, Arcot and Bruno (2005b) find that even among non-compliant companies, investing in those giving specific explanations generates higher returns than investing in companies that fail to provide any explanations. In fact, a portfolio of non-compliers giving specific explanations has returns similar to a portfolio of complying companies. It is therefore quite clear that the quality of explanations is driving the difference in returns.
In our opinion, these results are indicative of two things. First, there is a clear difference in
how those two groups of companies are governed. Poor explanations can potentially be a way to identify companies with poor governance. Second, the market has not yet fully picked up those differences. Investors that would buy the portfolios of compliers and non-compliers with specific explanations, and sell the portfolio of non-compliers with poor or no explanations, would drive the stock prices of the first group up and depress the stock prices of the second one. This would result in stock price movements that would reduce returns discrepancies. A difference of 12% annual is hardly compatible with the idea that such investment strategies are taking place on a large scale.
This suggests that an important dimension in the assessment of the "Comply or Explain"
approach is shareholders' attention to explanations. After all, the enforcement of the code is left to shareholders as the preamble to the Combined Code clearly states: " companies
should have a free hand to explain their governance policies in the light of the principles,
including any special circumstances applying to them which have led to a particular approach. It must be for shareholders and others to evaluate this part of the company's statement."
If shareholders put no pressure on companies to provide better explanations, they are perfectly entitled not to do so6.
We are by no means the only ones to express the view that shareholders might not be exerting pressure towards better explanations. As recently as November 2nd, 2005, Richard Hopper, Chairman of Inform PLC, in his letter to the Financial Times, argued: "Explanations of non-compliance by companies are at best ignored by some shareholders organisations or at worst the company is criticised for not achieving a full set of ticked boxes."
We have statistical evidence suggesting that indeed little attention is paid to explanations,
as opposed to compliance itself. Firstly, the quality of explanations at company level does
not improve over time. Most companies in our sample start with some type of explanation
and stick with that explanation until at some point they jump to full compliance. In other
words, we do not observe companies providing general explanations moving towards specific ones before ultimately complying. This evolution is not in contradiction with the fact that the average quality of explanations increases after 2001: the poor explainers became compliers at a quicker rate than non-compliers with better explanations. Put together these observations suggest that, if any pressure is put on non-compliers, it is directed towards compliance rather than quality of explanations.
Second, companies that cease to fully comply do not provide good explanations. In the whole sample, there is only one instance of a company which stops to comply and then provides a specific, verifiable explanation that was not purely transitory (e.g. the company does not have a separate CEO and chairman of the board as a result of the CEO's resignation). In fact, most companies that stop complying on some principle give no explanations. There could be two possible interpretations of this fact. It could be that those companies are unable to provide good explanations, because they do not have valid reasons to cease compliance. Or, it could be that they do have some, but do not take the step to state them in their annual reports. The first interpretation suggests that there does not seem to be set recurrent circumstances under which, in the words of Sir Derek Higgs, "a valid exception to the sound rule" is warranted. The second interpretation is that there are some such circumstances, but can hardly be used in support of the "Comply or Explain" approach: the outcome would be the same without any regulation. The absence of explanations, while there could be some, suggest that shareholders do not pay much attention to those so that managers do not feel compelled to justify non-compliance.
Moreover, if there is shareholder pressure on corporate governance issues, this relates to compliance rather than explanations and usually takes place after periods of bad performance.
The case of Morrison can be used to illustrate this form of shareholder activism. Morrison
has always been not compliant with most (i.e. six out of eight) provisions of the code and
either no explanation was offered or a poor quality explanation was provided by the company.
Shareholders apparently did not raise this issue as long as the performance of the company
was good. In 2004, Morrison completed the takeover of Safeway, after which (in July), Morrison announced its first profit warning in its 106 year history. This was followed by three more warnings in quick succession, which led to shareholder pressure and the appointment of David Jones as its first non-executive director in March 2005. In its annual meeting in May 2005 the company revealed its inability to forecast the financial position for the coming year. Shareholder pressure further intensified which led to the appointment of three more independent non-executive directors in July 2005 and a fourth in September 2005. In the meantime the CEO of the company, Bob Stott resigned and Sir Ken Morrison stepped back from operational responsibilities. It is easy to check that although until July 2004, the stock price performance of Morrison was in line with the market. After that date Morrison significantly under performed the FTSE100 index to the extent of nearly 40% upto July 2005.
We believe that Morrison's case illustrates some features common to many companies. In
particular it demonstrates that the intervention by shareholders in matters of corporate governance is usually not pre-emptive but typically occurs post bad performance. This highlights a significant cost of the flexibility offered by the Code, in that shareholders' incentives to take pre-emptive actions are not fostered. Moreover this form of enforcement does not generate better quality explanations but instead invariably pushes towards full compliance.
7. The Value of Flexibility
Our final comments suggest some policy recommendations, with a caveat which is the exclusion of financial service companies from our sample. The great achievement of the Code is in the very high number of compliances. But in light of our analysis, one might ask whether a less flexible and a more statutory approach should be considered, at least in some areas.
As compared to a statutory regime a flexible system like the Combined Code adds value if
there are conditions under which one size does not fit all. If there is full compliance, or if no
meaningful explanations are observed (in case of non-compliance), the "explain" part of the
code is ineffective. The relative benefit of flexibility, relative to a statutory regime, must be
therefore commensurate to the number of good explanations.
As compared to a statutory system, flexibility also introduces the possibility that some companies do not comply without having any good reason for doing so. The option not to apply a principle in a flexible system is thus put to bad use. This is most likely to be the case for companies that neither comply nor explain. The cost of flexibility, relative to a statutory regime, must be thus commensurate to the number of non-compliances which are not explained.
In light of the above discussion we compile the ratio of the frequency of specific explanations to the frequency of no explanations, for the each of eight provisions. Using this ratio we can evaluate the relative value of the Comply or Explain approach compared to a statutory approach. Figure 7 summarises our findings.
Figure 7 (to be provided later on)
This suggests that the value of flexibility is the lowest for the provisions relating to one third
of non executive directors, majority of independent non executive directors and composition of the remuneration committee. There are indeed more instances of no explanation than specific ones for these provisions. Given that the high level of compliances for the first two provisions (i.e. 98% and 95% respectively) not much would be lost if those provisions were made compulsory. In contrast, the number of specific explanations significantly exceeds the number of no explanations for the provisions pertaining to senior non executive director and composition of the audit committee. As noticeably, those two provisions are those with the highest frequency of non-compliances, there the flexibility, to comply or explain, might be of greater value.
Our above measure of the value of flexibility might be imperfect. There could be at least two
sources of noise in the calculations. It could be that there are some companies not complying for genuinely good reasons without explaining those circumstances. The puzzle there would be why they could not provide explanations of those circumstances. Symmetrically, there could be some explanations that we classified as good and are in fact not. In particular, the lack of diversity in explanations for certain principles might suggest that those explanations might not be so informative. Moreover, one could argue that a compulsory regime which would recognise some well identified set of circumstances where companies would not be obliged to adopt a principle would not penalise those companies. So the relative benefit of the "Comply or Explain" approach measured by those explanations may overestimate the true contribution of this approach. But if those two sources of errors are not subject to systematic bias, i.e. more frequent for some unobserved companies' characteristics than for others, this ratio would still contain some meaningful information.
Finally, we have ignored what would happen to the company's currently complying if there
was a move towards a statutory regime. Presumably nothing would change for those, as they would be complying under both regimes. There are two possible counter-arguments to this view.
First, their conditions might change in the future. As already mentioned, the fact that we do not find companies ceasing to comply and providing genuine explanations is suggestive that there may not be significant recurrent circumstances where ceasing to comply is needed
(a part from unforeseen circumstances). Second, one could take the view that in the current
state of affairs, too many companies complies. This argument is supported by our belief that
shareholders' pressure is essentially targeted towards compliance, not explanations. Admittedly, a move towards a statutory regime can only make things worse in this perspective.
However, if one thinks that compliance has reached excessive levels, one also needs to recognise that the current approach suffers from some shortcomings: how to modify shareholders' attitude? The revised code suggests that institutional shareholders spell out in writing their dissatisfaction, if they are dissatisfied, with some explanations of non-compliers. This is likely to induce companies to provide better explanations and we believe constitutes a step in the right direction.
The market seems to have chosen compliance as the rule. However, there seems to be some misunderstanding as to what compliance entails. The Combined Code asks companies to either apply its provisions, or to explain why they do not. To be precise, both are valid ways of complying. It would therefore be more accurate, we feel, to refer to the approach as "Apply or Explain", as this is what companies are asked to do. This admittedly minor change of terminology may help to promote the view among shareholders that greater attention has to be paid to explanations. It would recognise explicitly that companies that do one or the other de facto comply. One could then identify the non-compliers" as the companies that do neither.