Why Industries Produce Different CEOs — Bank vs. Manufacturing

Spend a week with bank CEOs and a week with industrial CEOs and you will notice they talk differently, worry about different things and are proud of different things. The lazy explanation is that industries have personalities. They do not. What they have is machinery — five mechanisms — that produces recognizable CEO styles from the ordinary variation in human temperament.

This module names those five forces and takes two deep dives: into the mind of the bank CEO, whose institution makes money by taking risk and can die from taking it badly; and into the mind of the industrial CEO, whose feedback arrives in tons, yields, downtime and scrap, and whose job at scale is to operate several abstraction levels above the plant floor without losing contact with it. Both are described, not diagnosed.

Then the most important move. Industry is one term in the Fit Equation, and often not the largest. A troubled bank and a troubled tire company share a problem — survival, credibility, cash, hard choices — that may demand more similar CEOs than two healthy banks in different strategic positions.

In the Effectiveness Equation — Traits × Behaviors × Organizational Context × Current Moment — this module works on Organizational Context, and shows that Context is itself a product: Industry × Scale × Lifecycle × Strategy × Governance × Problem.

Industries don't have personalities; they have selection, conditioning and constraint — and the problem in front of the company usually matters more than the industry it is in.

Four forces act on who becomes CEO and how they were shaped; the fifth on how much their traits can show. Together they produce a characteristic bank CEO and a characteristic industrial CEO. But applied to a company in trouble, the same forces produce a characteristic turnaround CEO regardless of sector. Fit is multiplicative, and Problem is a term.

RESEARCH FINDING The bank evidence begins with Ho, Huang, Lin & Yen (2016). In a panel of listed U.S. banks from 1994 to 2009, banks run by CEOs classified as overconfident by an option-holding proxy loosened lending standards and raised leverage more than peers ahead of the 1998 Russian crisis and the 2007–09 crisis — faster loan growth, especially in real estate, and market leverage roughly 5 percentage points higher. During the crises these banks suffered larger loan-default increases, bigger performance drops, higher CEO turnover and higher failure rates (about 10% versus 4%). The proxy is indirect, the design observational, and overconfident CEOs may sort into riskier banks. What it supports: overconfidence, in a business whose product is risk, is associated with taking more of it and surviving it less often.

RESEARCH FINDING Fahlenbrach & Stulz (2011) complicate the story that bank CEOs gambled with other people's money. Among 95 U.S. bank holding companies and investment banks with 2006 compensation data, banks whose CEOs had incentives better aligned with shareholders — larger dollar ownership — performed worse in the crisis, by roughly 10 percentage points in stock return and ROE. Option pay and bonuses did not predict worse performance. CEOs did not sell down beforehand and lost on average about $31.5 million (median $5.1 million). Small sample, one crisis, correlational. INTERPRETATION These CEOs believed the risk was value-creating. The bank CEO's characteristic failure is not greed but a misjudged tail.

RESEARCH FINDING Buyl, Boone & Wade (2019) studied 92 U.S. commercial-bank CEOs (2006–2014) using archival narcissism markers. More narcissistic CEOs adopted riskier policies before the 2008 shock, especially when pay was option-heavy; strong board monitoring dampened this; their banks recovered more slowly afterwards. Single industry, single crisis, proxy measure. In banking, governance is part of the risk-control system.

RESEARCH FINDING Berger, Kick & Schaeck (2014) provide rare quasi-experimental bank evidence. Using changes in German bank executive boards driven by mandatory retirements — a difference-in-differences design — boards with younger executives took more portfolio risk and boards with more PhD holders took less; a higher female share was associated with higher risk, a weaker and contested result. Many banks were savings and cooperative institutions. Because the design exploits an exogenous shock, a causal reading is supported: who sits on the executive board changes how much risk the bank takes.

RESEARCH FINDING The manufacturing evidence is about practices rather than traits. Bloom & Van Reenen (2007), in a double-blind interview survey of 732 medium-sized manufacturing firms in the U.S., UK, France and Germany, found management-practice scores strongly associated with productivity, profitability, Tobin's Q and survival, with weak competition and family succession by primogeniture the two main correlates of poor management. Bloom et al. (2019), using Census data on about 35,000 U.S. manufacturing plants, found management practices explain more than 20% of productivity variation, with about 40% of the variation across plants within the same firm. INTERPRETATION In manufacturing the operating system is measurable, varies enormously and predicts results — which is why the industrial CEO's judgment is disproportionately about systems.

RESEARCH FINDING Bandiera, Prat, Hansen & Sadun (2020) collected time-use diaries for 1,114 CEOs of manufacturing firms in six countries and placed each on a "manager" (one-on-ones with production staff, plant visits) to "leader" (multi-function executive meetings) index. A one-standard-deviation move toward "leader" behavior was associated with about 7% higher sales, emerging only about three years after appointment. They estimate that 17% of firms have a CEO whose behavioral type does not fit the firm — 36% in lower-income countries, 5% in high-income ones — and caution that this is a matching story, not proof that leaders are better. One week of diaries, manufacturing only.

RESEARCH FINDING Environmental conditioning has its clearest evidence outside industry. Benmelech & Frydman (2015), among 4,013 executives at 2,402 U.S. firms (1980–2006), found CEOs with military service ran more conservative policies — capital investment about 8% lower and R&D about 11% lower relative to the mean — were roughly 70% less likely to be involved in corporate fraud, and performed better in industry downturns; identification uses birth-cohort variation in draft exposure. Malmendier, Tate & Yan (2011) found Depression-era CEOs debt-averse and WWII combat veterans more aggressive with leverage. INTERPRETATION Formative environments leave measurable marks. It is a HYPOTHESIS, not yet tested directly, that thirty years inside a bank or a factory conditions a CEO the way a war or a depression does.

RESEARCH FINDING Custódio, Ferreira & Matos (2013), from S&P 1500 CEO résumés (1993–2007), found generalist CEOs earn a pay premium of about 19% — roughly $1 million per year — over specialists, largest when firms hire externally, switch from a specialist to a generalist, and assign complex mandates such as restructurings or acquisitions. Pay is a market price for skills, not a performance measure. INTERPRETATION The market already believes this module's point: when the problem is complex, it pays for transferable judgment over industry depth.

Where the evidence is weak

Nothing here directly measures the personality of bank versus industrial CEOs. The bank studies use option-holding and archival proxies; the manufacturing studies measure practices and time use, not traits. Conditioning is demonstrated for wars and depressions, not industries. The five forces are a FRAMEWORK assembled from discretion theory (Hambrick & Finkelstein, 1987; Hambrick, 2007), selection findings (Graham, Harvey & Puri, 2013) and the conditioning studies — a coherent reading, not a tested model. No study compares a troubled bank's needs with a troubled manufacturer's; that is INTERPRETATION.

The five forces

FRAMEWORK Industries produce characteristic CEOs through five mechanisms.

Self-selection. People choose industries that suit their temperament. Someone who enjoys probabilities goes to a bank; someone who likes seeing a thing made goes to a plant. By CEO-candidate age, thirty years of self-sorting have narrowed the population.

Promotion selection. A bank promotes people who did not blow up; a manufacturer promotes people who hit volume, cost and safety targets. Whoever survives fifteen years of that filter has been chosen for the trait.

Board selection. Bank boards, watched by regulators, favor candidates regulators will accept. Industrial boards favor people who have run plants.

Environmental conditioning. Decades of feedback shape judgment. The bank executive learns that a loan book that looks fine can be rotten; the industrial executive learns that a line either ran or it did not. Benmelech & Frydman (2015) and Malmendier, Tate & Yan (2011) show conditioning from formative experiences; the industry version is the hypothesis.

Managerial discretion. Module 11's force. Regulation and capital intensity narrow the bank CEO's latitude; competition and physical assets narrow the industrial CEO's. Whatever the first four forces produced shows up only in proportion to the room the fifth allows.

Three select, one shapes, one governs expression. Industries do not give people personalities. They filter, train and constrain them.

Deep dive 1: the bank CEO

A bank makes money by taking risk. It borrows short (deposits, wholesale funding) and lends long (mortgages, business loans), earning the spread — and every dollar of spread is compensation for a risk the bank chose to carry. So the elite bank CEO is not risk-averse. Their challenge is to take enormous amounts of correctly priced risk without ever threatening the survival of the institution — a different psychology from the gambler's or the accountant's. The vocabulary, in plain language:

Credit risk is the chance a borrower does not pay you back — the bank's core product and its main way of dying. Credit losses arrive late: a loan made in a good year defaults in a bad one, so a CEO can look brilliant for five years on decisions that were already wrong.

Liquidity is whether the bank can meet its obligations today. A bank can be solvent (assets worth more than liabilities) and still die in a week because it cannot turn assets into cash fast enough. Liquidity is the risk that kills quickly.

Capital adequacy is the cushion of the owners' own money that absorbs losses before depositors and creditors are hit. Regulators set minimum ratios. Capital lets the bank survive being wrong.

Regulation is the external layer — capital and liquidity rules, lending limits, stress tests, examinations, consent orders. A permanent second board with a veto and no interest in your growth story.

Market risk is loss from movements in interest rates, currencies or securities prices.

Counterparty risk is the chance the other side of a trade or hedge fails to perform — your protection is only as good as whoever sold it.

Reputational risk is loss of trust. Because a bank's funding is confidence, a scandal can trigger a liquidity event; reputation is not soft in banking.

Systemic risk is the risk that one institution's failure spreads through the system — why regulators care about large banks as they do not about large tire companies.

Balance-sheet management is the discipline of matching assets to liabilities, risk to capital and growth to funding so that all of the above stay within limits at once.

INTERPRETATION From that vocabulary you can describe the elite bank CEO's psychology. Numerate and comfortable in probabilities, because every decision is a distribution. Disciplined and institutionally minded, because the bank is older than they are and should outlive them. Skeptical, because the loan that looks fine is the dangerous one. Confident, because they must lend, and lending is a bet. Politically and regulatory-aware, because the regulator is a co-owner in all but name. Emotionally steady, because a rattled CEO is a liquidity risk. And attentive to tail risk in a way that would look neurotic elsewhere: not "what happens on average?" but "what happens in the year that kills us?"

The course's phrase for this is ambition combined with institutional paranoia — a description of a job requirement, not a diagnosis of a person. Ambition, because the bank must take risk to earn its spread. Institutional paranoia, because survival is the constraint under which all growth happens. Ho et al. (2016) show what happens when the ambition is present and the paranoia is not: faster lending, higher leverage, a 10% failure rate versus 4%. Fahlenbrach & Stulz (2011) show the failure was sincere. Buyl et al. (2019) show that board monitoring is how the paranoia gets institutionalized when the CEO lacks it.

Deep dive 2: the manufacturing CEO

A tire manufacturer, the course's standing example, buys natural and synthetic rubber, carbon black, steel cord and energy; runs capital-intensive factories whose economics depend on utilization; manages a large hourly workforce with safety and quality obligations; measures throughput, yield, scrap and downtime daily; depends on suppliers and logistics; ties up cash in raw-material and finished-goods inventory; sells to OEMs (vehicle makers, on multi-year contracts with brutal pricing) and to the replacement market (through dealers, with more pricing power); funds engineering for the next compound and the next plant; and decides capex years ahead of demand.

The distinctive feature is physical feedback. Volume was hit or it was not. Scrap was 2.1% or 3.4%. The line ran or it did not. A safety incident happened or it did not. This is a very different information environment from banking, where feedback is delayed and probabilistic. It rewards operational-economic judgment (what does a point of utilization cost, and what is it worth?), process and systems thinking (fix the constraint, not the symptom), discipline, capital allocation, continuous improvement, and comfort with tangible metrics.

The critical distinction is between the plant manager and the CEO of a global manufacturer. The plant manager runs the factory. The CEO operates several abstraction levels higher: who runs the factory and by what standard they are judged; what operating system governs all the factories; where the factories are, given labor, energy, tariffs and customers; which technologies get capital and which are allowed to age; which products and markets matter — OEM versus replacement, premium versus budget, which regions; and which capacity is added or removed, a decision that commits hundreds of millions of dollars and thousands of jobs years in advance. Bandiera et al. (2020) is the empirical warning: the CEO whose week looked like a plant manager's was associated, three years on, with lower sales. That is not "stay out of the plant." It is "the CEO's job is the system of plants, not a plant."

INTERPRETATION The good industrial CEO is a Player who became an Architect without forgetting what the Player knew. They still understand a scrap rate. They no longer manage one.

The sophisticated point: industry matters less than the problem

FRAMEWORK— the Fit Equation. Industry × Scale × Lifecycle × Strategy × Governance × Problem = CEO Fit. Multiplicative, so fit fails at the weakest term.

Put a troubled bank and a troubled tire company side by side. The bank has a bad loan book, a consent order, nervous funding and a board that has lost credibility. The tire company has two loss-making plants, a covenant it will breach next quarter, an OEM threatening to resource, and a board that has lost credibility. Both need a CEO who can face facts fast, size the hole honestly, cut without flinching, restore the confidence of the external party who can kill the company, and make irreversible decisions on incomplete information. Both need dials toward decisiveness, skepticism, urgency and centralization, at least for a year. The industry knowledge required is real but bounded — a credit officer, a plant operations chief — and can sit one level down.

Now put two healthy banks side by side. A deposit-rich regional bank with more funding than good lending opportunities, deciding whether to build a fee business, needs a patient, consensus-building CEO who can build a new business without breaking the culture. A fast-growing specialty lender whose growth has outrun its risk infrastructure needs an institution-builder with a paranoid streak who can slow a machine without stalling it. Same industry, close to opposite profiles.

INTERPRETATION This is what Custódio et al. (2013) price: the generalist premium is largest for complex mandates because the mandate, not the industry, sets the skill. Industry is a term in the equation, often not the largest, and hiring for it alone is how a board gets a bank CEO who is very good at banking and very bad at the problem.

Fictional composite.

Ines Valdry ran the industrial and financial-services practices at a mid-sized executive search firm, which is how she found herself, in the same quarter, running two searches her partners assumed had nothing in common.

The first was Ostrander Tire & Rubber: $1.4B revenue, 6,200 employees, four plants, 54% owned by the founding family with a public minority. Two plants were losing money; utilization at the oldest had fallen to 61%. A covenant on a $380M term loan would be breached at the next test. The largest OEM customer, 22% of revenue, had put its contract out to bid. The family CEO had stepped down after the board voted against him for the first time in its history.

The second was Kestrel Bancorp: $11B in assets, 1,900 employees, a listed regional bank. A commercial real estate book built over six good years had begun to sour; non-performing loans had tripled in eighteen months. The regulator had issued a consent order requiring more capital, a new credit-review function and board-level risk reporting. Two large depositors had quietly moved balances. The CEO had resigned.

Ostrander's board asked for "a tire person." Kestrel's asked for "a banker the regulator will accept." Valdry wrote down the problem before the profile. Ostrander: survive the covenant, keep the OEM, fix or close two plants, restore the lender's and the family's confidence — in twelve months. Kestrel: satisfy the consent order, stop the deposit leak, work out the CRE book, restore the regulator's and the market's confidence — in twelve months.

The profiles she wrote were nearly identical. Numerate. Skeptical of the numbers they would be handed. Calm under external scrutiny. Able to make irreversible decisions on incomplete data. Willing to centralize for a year and then let go. Credible to the party that could kill the company. Industry depth required at the CEO's right hand, not necessarily in the CEO.

Ostrander hired Tomas Reinholt, who had never made a tire. He had run the turnaround of a $900M packaging manufacturer, closing three plants and renegotiating with two anchor customers, and he brought a former tire-plant operations chief as COO. In year one he closed the 61% plant, moved its volume, held the OEM contract at a 6% price cut, and refinanced. Kestrel hired Priya Mahal, a chief credit officer from a bank three times its size who had run a CRE workout in the previous cycle. In year one she rebuilt credit review, sold $600M of loans at a discount the board hated, raised capital at a price the board hated more, and got the consent order lifted in nineteen months.

By year three the problems had diverged. Ostrander needed capital for a new compound technology and a decision about a plant in a lower-cost country — an Architect's problem, requiring patience and technical judgment Reinholt had less of. Kestrel needed to grow again in a market that remembered its trouble — and Mahal's institutional paranoia, exactly right in year one, was now slowing lending the bank could safely make. Both boards had a new fit question. The industry had barely changed. The problem had.

Put four archetypes into Ostrander's chair in the month the family CEO stepped down.

The Turnaround CEO does what Reinholt did, faster. Dials at decisiveness, urgency, centralization, skepticism. Gain: the covenant, the OEM and the plant decision get handled in sequence by someone who has done it before. Cost: the turnaround CEO's characteristic blindness is the day after the turnaround — cutting and centralizing when the company needs building and delegating (Module 10). Ostrander's year-three problem is where this archetype starts to misfit.

The Operator CEO — a lifelong tire executive, say a former COO of a competitor — knows exactly why plant four is at 61% and what it would take to reach 80%. Gain: credibility with the OEM and the plants on day one, and a fair chance the weak plant is fixed rather than closed. Cost: operators tend toward hands-on and patience, and the covenant does not care about the improvement plan. This archetype may fix the plant before the balance sheet and lose the company while improving it.

The Visionary CEO sees, correctly, that tires are becoming a technology business — sensors, sustainable compounds, fleet services — and that Ostrander's real problem is that it makes a commodity. Gain: a story the family and eventually the OEM may find compelling. Cost: optimism, innovation and aggression are the wrong dials for a company that must pass a covenant test in ninety days. The vision is the year-three answer offered in year one.

The Bank CEO — a former regional bank chief executive, transplanted — knows nothing about rubber but a great deal about surviving a period in which an external party has lost confidence, and about restoring credibility with a lender by showing them the bad news first. Gain: the covenant and capital problems are handled with unusual sophistication; institutional paranoia is precisely calibrated to survival mode. Cost: the plant decision and the OEM negotiation need operational judgment this CEO lacks, so the COO hire becomes the most important decision of the tenure — and if the bank CEO does not know that, the transplant fails.

In a survival year, the Turnaround and Bank archetypes — from opposite industries — are closer to what Ostrander needs than the Operator, the industry native. In a building year, the order reverses.

FRAMEWORK— the Overuse Ladder by industry. Each industry's conditioning produces a characteristic climb.

The bank CEO's ladder runs skepticism → cynicism and caution → timidity. Institutional paranoia, overused, becomes an institution that cannot lend: it stops taking correctly priced risk, loses its best lenders to competitors who will, and slowly earns less than its cost of capital — a quiet failure boards rarely fire for. The opposite climb — confidence → arrogance, risk tolerance → recklessness — is the loud failure Ho et al. (2016) document, and it usually arrives after a run of good years in which the delayed feedback of credit risk has made the CEO look better than they were.

The manufacturing CEO's ladder runs rigor → bureaucracy and detail → micromanagement. The operating system that made the company good becomes a religion; the CEO who once knew every scrap rate still wants to; the Player never becomes an Architect. The complementary failure is the Architect who has forgotten the plant — who approves a capacity decision on a spreadsheet and learns two years later what the line could never do.

The cross-industry failure is the one this module most wants you to see: hiring for the industry when the problem needed something else. The board that wants "a tire person" in a survival year is solving the Industry term and ignoring the Problem term. Because the equation multiplies, an excellent score on Industry does not rescue a zero on Problem.

Early warning signs

For a bank CEO: loan growth consistently above the market's with no change in stated standards; a credit committee that has become a formality; a regulator described as an obstacle rather than a co-owner. Or the reverse: no new lending product in five years, and the CEO is proud of it.

For an industrial CEO: a calendar dominated by plant visits in a company with a dozen plants; capacity decisions made without a technology view; or a CEO who could not name the constraint at the largest plant.

For a board: a search brief that names an industry before it names a problem; a shortlist of industry natives when the problem is survival or transformation; a CEO who was perfect for the last problem being kept for one they have never faced.

1. Which of the five forces most shaped you: did you choose your industry, were you promoted for its favored trait, or were you conditioned by its feedback? What trait did it strengthen that you now overuse?
2. Is your industry's feedback fast and physical, or slow and probabilistic? How has that shaped what you believe about your own judgment?
3. Where do you sit on "ambition combined with institutional paranoia"? Which half is naturally stronger, and what would it cost your company if the other half went missing for a year?
4. What abstraction level do you actually operate at — the factory, the system of factories, or the decision about which factories should exist? Is that the level your company needs?
5. Write your company's current problem in one sentence without naming the industry. Are you the CEO you would hire for it?
6. Think of a leader from a different industry you admire. Which of their traits would transfer to your problem, and which would not?

• Industries produce characteristic CEOs through five forces: self-selection, promotion selection, board selection, environmental conditioning and managerial discretion. Three select, one shapes, one governs expression.
• The bank CEO's job is to take large amounts of correctly priced risk without threatening survival — "ambition combined with institutional paranoia," a description of the role, not a diagnosis. Overconfidence in banks is associated with faster lending, higher leverage and more failure; governance is part of the risk system.
• The industrial CEO's judgment is operational-economic and systemic, exercised several abstraction levels above the plant: who runs it, what operating system governs it, where it is, which technologies get capital, which capacity exists.
• Industry is one term in the Fit Equation and often not the largest. A troubled bank and a troubled tire company may need more similar CEOs than two banks in different strategic positions. Write the problem before the profile.