Canary in the coal mine

16 October 2020

Executive pay practices can signal governance troubles ahead. Investors should take note.

Pay as a governance signal

On June 26th boohoo group plc announced a new Management Incentive Plan. This could pay out up to £150m to participants if the market capitalisation of the business increased by around two-thirds over three years. Most eye-catching were that each of the co-founders, Mahmud Kamani and Carol Kane, would receive one-third of this award each, or up to £50 million. A similar award was made to the Chief Executive John Lyttle when he joined in 2019. Within a fortnight, the company was embroiled in a scandal relating to working conditions in its supply chain, wiping nearly 50% off its market capitalisation.

Were the two stories connected?

Critics of corporate greed frequently claim that high levels of executive pay encourage venality. I’ve tended not to agree. Large-scale evidence suggests that typical levels of CEO pay are readily explicable by reference to the increasing value and complexity of our largest companies. Evidence from sudden deaths of CEOs suggests that the difference between good and bad CEOs far outstrips the cost in terms of their pay.

But this doesn’t mean that pay is irrelevant. Some shareholders have long suspected that excessive pay, defined as pay that is way outside the norms for the type and size of company, may be an indicator that the board has lost control of the Chairman or CEO. Shareholders were concerned about Martin Sorrell’s pay at WPP mostly because of what it told them about board governance and succession. Similarly, the pay debacle at Persimmon was not viewed in isolation but led to fundamental governance changes with the resignation of the Chairman, Chair of the Remuneration Committee, and eventually the CEO.

Boo-boo at boohoo?

So what about the pay scheme at boohoo? There were two courses for concern:

First, the overall level of pay was large for a company of boohoo’s size. Although a two-thirds increase in market capitalisation over three years sounds a lot, it broadly equates to expected upper quartile performance, so is not any more stretching than normal LTIP targets. In this context, an LTIP opportunity of £50m over three years is very high given the size of the company, even accepting it has been growing very rapidly.    

As so often in these cases the pay award was justified by reference to the proportion of value created. The announcement describing the scheme showed that the total pay-out at maximum would be less than 5% of the value created for shareholders (c. 1.7% for each of the founders). Surely such a small percentage is a price worth paying for great leadership? But this is not how competitive pay markets work. Shell’s market capitalisastion increased by 75%, or $100bn, in the three years from January 2016. By this logic the CEO Ben van Beurden might have deserved $1.7bn in pay for 2018 (he received $24m). 

Use of “share of value” arguments to justify pay should always be a red flag, requiring further challenge. The logic can be used to justify, based on small percentages, pay packages that are way outside competitive norms. But what counts in a competitive market is the extent to which the level of pay is consistent with the norms offered to the same post holder in other companies. Any large deviation from the competitive level needs to be subject to sceptical enquiry.

The second concern was that pay package was relatively short term. Full payout could be received after three years with no subsequent phased holding or vesting periods. Performance measures based on share price alone are not themselves a cause for concern over the long enough term. Indeed the share price integrates many dimensions of performance, including ESG, over time frames of five years or more. However, three years is arguably too short. Andblock vesting events create behavioural distortions and short termism as research by Alex Edmans has shown.

In boohoo’s defence, retail is a sector where private equity has been very active. And private equity pay schemes are more inclined to be designed on a share of value basis, although the share awarded to management is very dependent on deal size and driven by competitive opportunities on similar sized transaction. We also cannot see inside the company to assess the criticality of the management team, retention concerns and so on. As Smith & Nephew found, the market can be unforgiving when talent departs. Following the resignation of their CEO, Namal Nawana, over pay in October last year, the share price fell by 9% or around £1.5bn. This loss must vastly exceed the cost of the pay award required to retain him. No board wants to be accused of failing to retain a talented leadership team. 

But the boohoo scheme had two design features that would have given cause for valid shareholder concern and challenge. But of course boohoo is AIM listed, rather than being on the main market, and as such they are under no obligation to obtain prior shareholder approval for executive pay schemes. boohoo did nothing illegal or unusual here, and investors who bought shares in boohoo would have known they would have no decision rights over pay.

Anecdote or systematic issue?

My own experience as a consultant was that excessive pay demands were often the sign of a CEO believing their own publicity and a board that had lost discipline and oversight. I remember one case of being asked to design an exceptional Top Up plan for a long-serving and clearly narcissistic CEO. The company became embroiled in scandal before the plan could even be put to shareholders for approval. But this is just an anecdote - as are the cases of boohoo, Persimmon, and WPP - and so proves little. What does the systematic academic evidence have to say? There are at least two high quality studies that indicate that shareholders should be watchful about excessive pay.

The first is a paper by Lucian Bebchuk, Martijn Cremers and Urs Peyer. They define a measure called ‘CEO pay slice’ which is the fraction of the aggregate compensation of the top-five executive team paid to the CEO. They reason that this measure could be a measure of the relative importance of the CEO within the top team or of their dominance and ability to extract rents. Any ratio measure suffers from the weakness that it can be affected by the numerator and denominator. But they find some interesting results following a set of carefully controlled investigations, with higher CEO pay slice being associated with a remarkable range of factors including:

• Performance factors such as: lower price-to-book ratio, lower accounting profitability, worse acquisition decision (measured by market reaction); and

• Governance factors such as: more opportunistic (i.e. favourable) timing of CEO option grants, more luck-based CEO pay in the event of positive industry shocks, less CEO turnover (controlling for performance and tenure).

They do not identify direct causality but assemble a sufficient body of evidence to suggest that excessively large CEO pay slice is associated with agency problems and therefore a useful filter for studying governance risk at firms.  

The second paper is by David Yermack and studies personal use of company planes by CEOs. The cost of personal plane usage is high enough reliably to exceed SEC disclosure thresholds and so provides a relatively robust dataset for studying executive perks. There are two broad theories on perks. One is that they are a part of optimal contracting, and that companies can provide benefits to CEOs at a cost and convenience they couldn’t realistically obtain themselves. Under this view, any value from perks is offset against other parts of compensation. The second view is that perks are a result of rent extraction in weakly governed firms and could on this view act as a signal for wider governance failures.

David Yermack studies disclosed personal use of company jets during the period 1992 to 2003. Disclosed personal usage increased nearly five-fold during this period, although the author acknowledges this is likely at least in part due to increasing disclosure standards. Personal characteristics were the main predictors of the extent of jet usage for personal benefit. Older CEOs used than younger CEOs. But the strongest result was golf-club membership. CEOs who were members of Augusta National or another club far from their headquarters incurred two-thirds more flight costs than the average.  

It’s clear that shareholders don’t like CEOs to use company jets for personal use. On announcement of the perk there is a 1% fall in share price, vastly in excess of the capitalised value of benefit in most cases. This suggests the shareholders consider ‘excessive’ perks as an indicator of wider issues. However, the 1% fall vastly underestimates the subsequent underperformance that David Yermack documents. He finds that in the year prior to first disclosed use of a plane for personal use by the CEO the stock outperforms by 5%. But this is then swamped by underperformance of 9% in the year following and 4% pa year thereafter – massive margins. This underperformance coincides with disclosure of bad news in the years following the perk announcement in terms of increased write-offs and negative earnings surprises. Again, causality is not established, but the impacts are material enough to be striking.

A window on the board

Remuneration is an area that has a high potential for emotionally charged conflict within a board. At the same time, it is difficult to think of an area of engagement between the non-executives and executives in which so much information is provided externally, both through the decisions made and the process followed. For these reasons, remuneration practices can provide a window on the wider workings of the board and the nature of the power relationships at play. These have direct implications for governance.

Note that the concern here should not about high pay. Nor should the concern be about different or tailored pay structures, which might reflect a company’s particular strategy. Rather it is about excessive pay, defined as pay that is unnecessarily different in level or nature from established norms for the type or size of company. High pay may simply be the result of rigorous pay processes being applied  taking company circumstances into account in a competitive market for talent. Excessive CEO pay could be an indicator of a CEO that has lost perspective or a board that has lost control. 

It’s not always easy to tell which is which. But when investors are presented with exceptionally high pay, and especially if it’s short-term in nature, they should think carefully about what might be around the corner.

---