Back to Community
The CEO Arrogance/Narcissist Index

Hello community of alpha seekers and quants,

I'd like to propose a new index which is to quantify the arrogance - humbleness of CEOs using quantitative scorings to be defined (probably by psychologists knowing their human behaviour stuff).

My thesis is that just like Karen Rubin has shown that female led companies outperform their peers (see https://www.quantopian.com/posts/investing-in-female-ceos-sector-neutral-a-different-benchmark-and-new-data) I would postulate that humble CEOs who are Servant Leaders outperform arrogant ones.

This seems to support such a hypothesis:
https://www.forbes.com/sites/ericjackson/2012/01/11/why-narcissistic-ceos-kill-their-companies/

...and there's probably a lot more of it if we dig plus some researchers such as the mentioned Don Hambrick.

This is NOT a political project but rather I'm just hunting alpha in alternative ways.

Ultimately it seems reasonable that a company's performance compared to its peers would be amongst other things a function of the skills of its staff both professional and human.

It is usually pretty straight forward to pick the best for the job in terms of merits - but what about the human aspect. Here I postulate that a measure of arrogance could be a good proxy for measuring fitness.

What I call for and need:
- anyone knows if similar data exists and where?
- anyone willing to join and create the index (huge undertaking - but probably fun)

On offer:
"(Wo)Men wanted for hazardous journey. Low wages, bitter cold, long hours of complete darkness. Safe return doubtful. Honour and recognition in event of success" -- Shackleton, exploring the Antarctic

17 responses

Hi @Bernino,

Its an interesting project that you have chosen! Having worked for many years in the corporate world in both employee and consultant roles, i have experienced first-hand both the harm and also the good that good or bad CEOs can do, and i have a few comments.

Strong authoritarian leadership and narcissistic leadership might perhaps appear to be similar, but they are not. An excellent example of a well-known past leader who was highly authoritarian but definitely NOT narcissistic was the former and now deceased leader of Singapore, Mr. Lee Kwan Yew. I refrain from comparing with any well known narcissistic leaders today. Narcissism is a personality disorder, a psychopathology. Strong authoritarian leadership can be benevolent, whereas narcissistic leadership seldom is. Certainly in the corporate world many people, especially employees, suffer under narcissistic leaders / managers. This is usually subsequently reflected in employee morale and often ultimately in corporate profitability, so i think your proposed project is well justified as part of a search for investment alpha.

Thanks for the link to the article by Eric Jackson, but unfortunately i think the article over-simplifies too many things. One in particular is that companies are not just run by their CEO but also by their senior management. A wise but strongly authoritarian CEO may be open to having other members of senior management who disagree with her/him, as long as they can play well as a team. A narcissistic CEO is more likely to surround him-/herself with toadyish yes-men & yes-women, so the whole culture of a company with a narcissistic CEO is likely to be poisoned from the top down. Unfortunately however it is quite possible (and often happens) that a company with a good CEO who is not narcissistic may still have one or more narcissistic managers at lower levels. So, even if the top of the company is "clean", there may be toxic work environments in other parts of the company further down. This means that things like the size of the CEOs picture or salary compared to subordinates may NOT be very good indicators at all of the overall health of the company.

I would suggest using other metrics of corporate psychological health that are more broad-based. If the work environment is too toxic (for whatever reason, but often because of narcissistic managers although not necessarily the CEO), then employees will be unhappy. Many will not stay around for long. I would suggest using employee turnover data as a better proxy for corporate psychological health than the size of the CEO's photo.

Here's a general factor using the employee turnover idea. I used Factset Fundamental dataset and as you can see the turnover is very low ~3% with annual spikes as expected.

Loading notebook preview...

Hi @Daniel, @Bernino,

The "level of CEO Narcissism" is likely to be only one (probably fairly small) contributing factor to corporate profitability and share price performance, so i do expect we are looking for a very small signal in a lot of noise, but nevertheless any bit of incremental alpha is worth finding if we can :-)

In order to try to strengthen the signal, my suggestions are to:

  • Look not only at the individual CEO, but also the entire "C level" management and other senior managers as well.

  • Look at broad-based proxies for corporate morale in general. The first and most obvious one is employee turnover. Thanks @Daniel.

  • Employee turnover in some industries is likely to be highly Seasonal. Therefore, if the aim here is to study the impact of CEO or broader corporate management narcissism, then (unless the CEO is Seasonally manic-depressive :-)) i think it makes sense for the purpose of this study to look at long-term rather than short-term changes in employee turnover. So that means either making some sort of "Seasonal adjustments" or else, probably more easily, just using an averaging of employee turnover over a reasonable window length (e.g. at least quarterly = 63 trading days or longer).

  • My guess is that employee turnover is probably related to Industry, with some industries tending to have higher rates of employee turnover than others. Therefore i think it makes sense to consider not only the actual employee turnover number itself, but also the turnover relative to the Industry or Sector to which the company belongs. So i would suggest considering: a) Employee turnover%, and b) Employee turnover minus Sector average turnover%, and also c) Employee turnover minus Industry average turnover%.

  • I wonder what other factors we could find as proxies for employee satisfaction or lack of it due to poor working conditions or poor management (with "CEO narcissism" being one potential contributor) ? Of course employee sick leave is often due to sickness, but most people's "sick" days off (especially when just one single day at a time) are often also related to employee unhappiness. I don't know if "sick leave days" data are available, but if so then my guess is that this might also be potentially useful info.

I'm completely humbled that anyone reacted to this post, thank you so much for your thoughts and efforts @Tony and @Daniel..!

I'm 100% with your lines of thought and did some additional digging for more proxy data. It turns out https://glassdoor.com has employee review data for fortune 1000s and I've reached out to them to ask if we can have access to their data.

Its interesting that there is a slice of alpha in the churn of employees - but I still wonder if unhappy employees really do leave their jobs as an assumption? My hearsay experience is that people would rather stay in a known and bad situation than to go into an unknown situation but remediating the bad part of the old situation. I would suggest this is related to homosapiens being the risk averse and loss averse species -- there is a paper somewhere showing that we are x2 or x3 more likely to take a decision to avoid a loss than to take a decision get a gain (in a normal distribution cross sectional population - investors will be very different from this).

Here is more on loss aversion and links to the papers: https://en.wikipedia.org/wiki/Loss_aversion

Also, the more I think about this idea the more I'm puzzled why this data doesn't exist publicly already - a public scoring of the leadership of companies; as everybody talks about how brilliant CEO x,y,z is - and punishes a company if they loose a perceived good CEO - however - "good" and "bad" isn't quantified... to compare: PMs and Presidents are being polled monthly for approval rates.

Anyway, I will let you know if we get access to that dataset.

Meanwhile the inverse of the employee churn might be interesting as well - companies that expand their workforce must be very bullish about their future, have a bit of free cash flow and know a bit about their market and upcoming deals in order to invest. Do you know if such data exists?

@Bernino, yes of course you are correct that risk aversion tends to bias people's actions in ways that are not always in their best interests (e.g. staying in abusive relationships, staying in bad jobs, etc) rather than taking a dangerous step into unknown or uncomfortable situations (no relationship, no job, etc). From my own personal experience of about 40 years in industry, i have certainly seen a lot of people gritting their teeth and hanging in there in bad or worsening work situations that they dislike, complaining more, taking more sick leave, buying more lottery tickets, but still not leaving, at least not until it becomes totally unbearable, or they see an advertisement for a better job, which may take a very long time.

Really its not actually employee attrition per-se that we want, but rather some readily available measures of employee satisfaction or lack thereof. The problem is that i don't know what else there is.

You write: "I'm puzzled why this data doesn't exist publicly already - a public scoring of the leadership of companies"
Personally i'm not surprised at all. Firstly i doubt that accurate scoring of companies in this way can be done by "the public". Generally it is only the people actually inside a company who know what the company (and its leadership) are really like. Many employees are frightened to speak up about bad or abusive bosses or work environments because they are scared of being fired. Many companies discourage internal criticism or comment, even if it might actually help them. During my own career, i tried on several occasions to introduce the idea of extending annual employee reviews in a way that would give "360 degree feedback", so that employees could rate and comment on their bosses, rather than just having superiors do performance reviews of their subordinates. In fact i actually did request my own staff to anonymously do "performance appraisals" of me as their boss, which was very helpful even if a bit uncomfortable for me. However any attempts to introduce this more widely were immediately blocked. Most managers just do not want to be given honest feedback about their shortcomings, and most employees are too scared to do it, even when given the opportunity. Clearly however, more and more people are starting to think about possible ways to implement honest feedback systems ... in industry as well as in politics. Best wishes for your research, TonyM.

Thank you Tony!

Regarding the dataset "measures of employee satisfaction or lack thereof" this is exactly what Glassdoor.com has on offer but sadly not for external customers as of now. I have reached out to them and await their answer.

Regarding public scoring that was a wrong semantics - what I meant is that such data should be publicly available, but of course sourced by the employees themselves who knows best how things are going.

I'm starting to imagine the alpha hunting with the datasets of:

  • CEO gender
  • CEO degree of narcissism vs. servant leader
  • Employee satisfaction

Thanks for the feedback and thoughts once again!

Sounds good Bernino. I suggest doing all of the 3 things that you suggest on a Company-vs-Sector and a Company-vs-Industry basis, as well as just on the basis of the individual company on its own. Cheers, all the best.

humble CEOs who are Servant Leaders outperform arrogant ones

My intuition is that this is not the case. Just look at AAPL, TSLA, T-Mobile, etc etc. I think big personalities and big vision go hand-in-hand.

On a related note, are you familiar with Ray Dalio's CEO personality tests?
https://www.cnbc.com/2019/05/02/bridgewaters-ray-dalio-had-elon-musk-bill-gates-do-personality-test.html

Yes, 'm familiar with the type of psychological / personality tests that Ray Dalio used, and I would certainly agree with your statement that: "... big personalities and big vision [sometimes] go hand-in-hand" .... but that is not exactly the point being made as per the title of this thread.

I repeat my previous comment:
Strong authoritarian leadership and narcissistic leadership might perhaps appear to be similar, but they are not. An excellent example of a well-known past leader who was highly authoritarian but definitely NOT narcissistic was the former leader of Singapore, Mr. Lee Kwan Yew. Narcissism is a personality disorder, a psychopathology. Strong authoritarian leadership can be benevolent, whereas narcissistic leadership seldom is.

It is a mistake to confuse strong authoritarian leadership which is definitely compatible with "big personalities and big vision", with narcissistic leadership, under which everyone and everything suffers, as do many people working in the corporate world with narcissistic managers who are nowhere near as good as they imagine themselves to be. Nevertheless, people can change and improve their interpersonal style. Steve Jobs is probably a good example.

Thank you both! Interesting thoughts. On my end I would be mostly interested in the data and what the back testing on the data says with a hypothesis that narcissistic and sociopathic CEOs underperform. But who knows what the data will say ...

Thanks also for the test - I guess that could be a starting point and maybe Dalio's hedgefund have some data...

"... who knows what the data will say ..." .... indeed.

@ Bernino, I understand fully that your hypothesis is that "narcissistic and sociopathic CEOs underperform" and, given that such managers are reasonably common (even if not necessarily in CEO position), i think your study is worthwhile, but not likely to be easy. Two difficulties are 1) deciding what data to get, and then 2) getting it.

Obviously CEOs themselves are not going to tell you if they are narcissistic / sociopathic. Their employees could, but they are probably not going to .... unless they are just about to quit. So the difficulty is in finding appropriate proxy data. Employee dissatisfaction and turnover are probably useful proxies, as we have already said.

As @Viridian implies "Big Personality" might perhaps sometimes be an indicator of "Big Vision" (although i venture to say that there are plenty of big noisy personalities with little or no vision at all), but "Big Personality" is certainly neither synonymous with nor a good proxy for "Narcissistic / Sociopathic". The industry examples given above would in fact tend to support this. TSLA = Elon Musk is undoubtedly a "big personality", maybe loud & abrasive, but not narcissistic / sociopathic, which is clearly demonstrated by the fact that Musk demands his employees have the courage & willingness to to speak up openly and disagree with him. I don't know what data might be available that indicates whether or not a corporate culture permits disagreement with the CEO without incurring retribution, but it sure would be useful.

1) I think it would be very difficult to do unbiased personality assessments. Sociopaths are often charismatic and hard to detect until after some terribly egregious act has led to their undoing. Elon Musk is a great example of an ambiguous figure. Depending on who you ask he is either a hero or a fraud. Ray Dalio's personality assessment attributes the same qualities to him as to other world-changing CEOs. However -- High executive turnover, SEC settlement, FBI investivation, "pedo-guy" lawsuit, accounting irregularities, "beta-testing" potentially dangerous new features on human customers, etc. are just among a handful of the evidence TSLA short-sellers point to. I'm not making an argument one way or the other, only pointing out how difficult it is to make an unbiased assessment as to whether somebody is a narcissist or sociopath.

2) I think it would be very difficult to produce a point-in-time dataset unbiased by look-ahead bias. Today we may view the Enron executives as sociopaths, but can we say for certain we would have been able to detect it before Enron collapsed? How would we even go about assessing personalities back in time, since as Tony suggests, people change. Contemporary public accounts? Very difficult to reconstruct.

So while I think personality assessments are a hopelessly difficult project for DIY quants such as us, as suggested earlier there are perhaps other, less subjective ways we can pursue the thesis. Accounting irregularities is one. Or even looking at things such as executive compensation, insider trading. Or is the company engaged in anti-social behavior -- such as Philip Morris or Herbalife. If the companies' business practices are sociopathic, then likely it would require a sociopath to run them.

Company executives skew more towards sociopathy than the general population.
https://www.cnbc.com/2019/04/08/the-science-behind-why-so-many-successful-millionaires-are-psychopaths-and-why-it-doesnt-have-to-be-a-bad-thing.html
https://www.forbes.com/sites/victorlipman/2013/04/25/the-disturbing-link-between-psychopathy-and-leadership/#7723af5e4104

Question is -- while sociopathy is bad for society, is it bad for business? Sure, it can lead to an occasional Enron. Unfortunately though, I think generally capitalism favors those who will do anything to make a buck over those who do what is compassionate and good.

Another thing to consider is company culture. How pervasive is the anti-social behavior? Is it just somebody at the top or is it what guides the entire company? The financial crisis wasn't just caused by people at the top. The whole industry was drinking the kool-aid. Wells Fargo had a scandal a couple years back along the same lines -- people not even high up in the company were signing customers up for accounts they hadn't enrolled in, in order to earn bonuses.

1) "difficult to do unbiased personality assessments. Sociopaths are often charismatic and hard to detect until after some terribly egregious act" Yes, true.

"Elon Musk is a great example of an ambiguous figure". Yes, "ambiguous" is probably a good description. Some of the things that Viridian mentions were definitely not nice at all, although other things like "accounting irregularities" are by no means indicative of sociopathic / narcissistic behaviour, just "normal, run-of-the-mill" corporate bad.

2) Yes, indeed very difficult task.
Enron is a good case study in many ways. Before it all went wrong for the criminals who were at the top, Enron was the absolute darling of the market and was widely written up as "one of the best companies in America to work for". But were the top people actually narcissistic psychopaths, or simply criminals with rat-like cunning? I would tend to argue probably the latter, but it is debatable.

"... can we say for certain we would have been able to detect it before Enron collapsed?" Actually Enron's impending collapse was quite detectable (especially with the benefit of hindsight) before it happened, not by psychological measures, which no-one probably even thought about doing at the time, but certainly detectable by using some of the measures of accounting forensics.

"personality assessments .... difficult project for DIY quants such as us".
Yes, difficult indeed, but not impossibly so.
However, as Viridian wisely notes: "perhaps other, less subjective ways we can pursue the thesis".

"Accounting irregularities is one". -yes, definitely agree.

"executive compensation" - IMHO this is actually a rather unreliable measure to use on its own. However what about executive compensation vs performance of the company, after normalizing for differences between industries ?

"insider trading" - is legal if disclosed.

"company engaged in anti-social behavior" - this is a very interesting values-based one. Personally i agree about cigarette companies because i am anti-smoking. I'm also anti-war. So then, are companies that promote war sociopathic? OK then, what about so-called "defence" companies? It gets to be a very grey area....

"Company executives skew more towards sociopathy than the general population"
yes, probably so, and my own personal observations from working in industry for many years would tend to support the conjecture.

The question at the end of Viridian's post is a very good one for every decent and ethical person to ponder if they are interested in business / investment.

If i may take the liberty of extending this just a little bit further, i notice that recently a number of wealthy senior people in industry are talking about "improving Capitalism", presumably (or at least so they claim) that everyone would benefit more. The words might sound nice, but words are cheap. To me it sounds more like a fashionable piece of corporate PR blah-blah. Personally i tend to doubt the sincerity of some of the people in industry who are talking but who give very little evidence that they are actually compassionate & good.

Perhaps characteristics that we can find in the numbers is the way to look for things...: https://www.forbes.com/sites/ericjackson/2013/09/16/the-top-25-most-narcissistic-ceos-in-tech/#12789b9368a7

"He - along with several colleagues - found that a more narcissistic CEO was more likely to:

  • make more acquisitions,

  • pay a bigger premium for those acquisitions,

  • spend more on advertising,

  • spend more on R&D,

  • continue to make big M&A deals even if their recent performance had been poor, and

  • lead an organization to more volatile performance (i.e., it would work out really well or be a total disaster)."

Other metrics from the article:

The prominence (size) of the CEO’s photo in the annual report
CEO prominence (number of mentions) in company press releases
CEO’s use of first person singular pronouns in transcripts of public comments to shareholders
The gap between the CEO pay (salary, bonus, deferred income, stock grants, and stock options) and the pay of the 2nd highest paid executive

The author of the article doesnt think he finds much but he didnt do it with as much data I guess.

EDIT: Just look at Wayfair, not much in the way of Acquisitions but huge cash burn on "efficiencies" and the like and you only have to have tried to have traded this stock to know how volatile its been!

Thanks for these great replies - very nice of you!

As such I think the conclusion is that though an interesting idea - the data reality is that it's not feasible.

I will instead play with female led companies as the data is almost there :-)

Thanks all for contributing to this - I had a similar hypothesis and appreciated coming across this thread. One note: If I'm reading the documentation correctly, the FactSet Fundamentals have a number of different "turnover" metrics, but none of them refer to employee turnover. Rather, trading turnover, inventory turnover, receivables turnover, etc. In general, most of those would all also have a "lower is better" dynamic.

Hey

You may want to look at Directors and Officers (D&O) insurance as a signal. https://www.investopedia.com/terms/d/directors-and-officers-liability-insurance.asp Companies need to buy insurance for their leaders and the insurance industry has a set of underwriting guidelines to check for factors. Admittedly, it may not be directly what you are looking for but potentially correlated.