18 May 2018

Evidence from market reaction to sudden deaths suggests that levels of CEO pay are not as mad as they seem

The current AGM season is shaping up to be a lively one here in the UK for pay votes. Companies are being put on the Investment Association’s naughty step (for companies receiving 20% or more vote against a resolution) at three times the rate of last year. Many of the votes against are because of pay increases or bonus outcomes that are viewed to be unacceptable. The role of proxy voting agencies appears to be particularly influential this year, with the rate of AGAINST recommendations on remuneration resolutions by ISS, the dominant agency in the UK, also being around three times last year. Anything that looks like an attempt to increase pay is getting short shrift.

The tougher stance of investors and their advisers is being fuelled by the political controversy stoked up around high pay. Polly Toynbee captured the public mood in the Guardian this week: “I go to company AGMs to stare at masters of the universe paying themselves eye-watering, breathtaking sums. In the one day a year on which they are on public display, I seek a glimmer of shame, any blush or blench. But when some eccentric shareholder questions the size of their swag these masters never bat an eyelid, unabashed in their brazen effrontery. Will the day ever come when Britain looks back in incredulity at such shameless fatcats.” 

It now goes almost unquestioned that CEOs are faceless bureaucrats benefiting from a back-scratching pay culture entirely divorced from merit. After all, what impact can one person have on an organisation with thousands of employees. Surely most companies largely run themselves? Executive pay packages, which are asserted to be increasing rapidly, are deemed unjustified and immoral. 

It’s a myth that executive pay continues to increase at a rapid rate: in fact pay of FTSE-100 CEOs has risen even less than average national earnings since the financial crisis. But is it just too high full stop? Median pay in the FTSE-100 is around £4m per annum, approximately 150x the average UK wage of £27,000. This ratio is often taken to be manifestly absurd, providing prima facie evidence of a racket in executive pay.

So what is a CEO really worth, and can that ever be £4m a year? Two rather morbid papers shed light on the question by looking at the market reaction to sudden CEO deaths. Many methods for looking at the impact of CEOs on company performance suffer from the problem that it can be almost impossible to untangle the various factors in play that affect performance. By contrast, the Grim Reaper runs a randomised trial, with sudden deaths reasonably considered exogenous to company performance. It seems callous, but as a result these sad cases do provide insight as to how shareholders value CEOs. 

Both papers look at the Cumulative Abnormal Return (CAR) over a short period around the time of the CEO announcement to quantify the impact that shareholders ascribe to the CEO’s death. That is, the amount the share price moves above or below what would reasonably have been expected without the death.

The first paper is by Quigley, Crossland, and Campbell and was published in the Strategic Management Journal, a top-rated management journal, in 2016. The paper finds that the CAR can be positive or negative. It is more often negative, suggesting that investors normally view the sudden death of a CEO as bad news, but when positive the CAR tends to be larger. They find that the absolute size of the CAR (whether positive or negative) has increased markedly over the last sixty years, from an average magnitude of around 3% in the period 1950 to 1970 increasing steadily to around 8% for the period 1990 to 2010. 

The second paper is by Jenter, Matveyev, and Roth. This paper is not yet published in a peer reviewed academic journal, which normally suggests the results should be treated with caution. Indeed the authors themselves present the paper as preliminary and incomplete. But the paper has been presented, debated, and scrutinised at the European Winter Finance Conference, and given the reputation of the authors, we can have confidence in the findings. These authors also looked at the CAR around the time of a CEO death, and found a small, but statistically significant, average negative impact of around -2%. However, they found that this average disguised a wide degree of significant variation depending on the CEO characteristics. In particular, the death of younger, shorter tenure CEOs led to larger negative impacts, of around -4% or slightly more. The impact doubled to -8.8% for young founders.

By contrast the older CEOs had a positive average CAR of +3.6%, with the impact again being magnified for founders to +5.3%. This suggests that shareholders saw benefits in the removal of more entrenched CEOs, particularly if they were founders, but with the converse applying for younger more recent appointments. No one can accuse financial markets of being sentimental.

These papers contain many insights that I can’t do justice to here. But taken together they point to the fact that shareholders ascribe a value difference of broadly +/-5% to a good versus bad CEO. That may not sound like a lot, but with the typical FTSE-100 company worth around £10bn, this means that for shareholders the difference between a good and bad CEO is around £1bn. At current long term interest rates that’s enough to pay the typical CEO pay package of £4m per annum in perpetuity and still leave change of £750m. Even at £10m pa, a good FTSE-100 CEO easily pays for themselves. 

I’ve written elsewhere about the body of evidence that suggests that CEO pay is much more explicable, in terms of economics, than is commonly assumed. These two papers reinforce that view. So is it rational for investors to be stepping up their activism against CEO pay in the way they are? It’s hard to argue against scrutiny of pay, but what we’re seeing is a fairly undiscriminating approach to clamping down on CEO pay awards, driven by the analysis of governance specialists and proxy agencies. This is proving effective in bearing down on pay inflation in aggregate, but it’s a blunt instrument. And as with all blunt instruments there’s collateral damage. 

What these two papers show is that while some CEOs are hugely valuable to their company, others are holding them back. A governance-driven approach to pay levels, that ignores firm and CEO-specific factors, does not discriminate. Things could get worse. The BEIS Select Committee has executive pay in its sights again. There’s growing momentum for harsher rules on pay votes, with penalties for votes below 75% or even 80%. But such reforms run the risk of putting yet more decision-making power in the hands of people who aren’t qualified to make judgements about CEO worth and the CEO’s fit with the company or its strategy. Applying a single rule-book across the market will have hidden economic costs.

But this is politics not economics. Shareholders, and companies, cannot ignore the public mood – or at least do so at their peril. Business needs to show restraint to claw back some of the trust lost over recent decades. But amid all the sound and fury, it is worth remembering that levels of CEO pay are not as mad as they seem. And in terms of economics, good CEOs are cheap at twice the price.