The Most Dangerous GM in Your Portfolio Owns the Room. His ROE Is Bleeding.
He walks into the quarterly review with a tight deck, a confident story, and a command of the room that makes the board lean in. His peers defer to him. The CEO points to him as the model. His business unit’s ROE has dropped 420 basis points over eight quarters. Nobody has said it out loud because his confidence makes the question feel like an insult. He is the single most expensive leader in your portfolio, and you are about to promote him.
Dunning-Kruger effect describes the inverse relationship between confidence and competence in specific domains. When plotted against actual performance data, leaders with the highest self-reported confidence frequently show the steepest performance decline, because confidence suppresses the feedback loops that would otherwise correct the trajectory.
The Fusion: Pop Psychology Meets Hard Finance
Dunning-Kruger is a cocktail-party concept. People reference it to describe their annoying coworker, then go back to running their business on gut instinct. DuPont decomposition is finance-textbook machinery. CFOs reference it in footnotes, then go back to presenting P&Ls that obscure the actual trajectory of return on equity. Neither, alone, is a leadership weapon.
Welded together, they become the single most honest audit of your executive bench you will ever run. The Dunning-Kruger axis measures a leader’s self-assessment. The DuPont axis measures what their business unit actually produced. Plot them against each other across 12 quarters and the Dangerous GMs appear as unmistakable outliers. They are the ones who consistently rate themselves above the 80th percentile in capability while their net margin, asset turnover, or both, trend down. The math is on the wall. The math doesn’t care about the deck.
The comfortable delusion is that great leaders are confident leaders. That’s a selection bias masquerading as wisdom. You remember the confident leaders who succeeded because they became CEOs and wrote books. You forget the confident leaders who failed because they were quietly reassigned, paid off, or allowed to retire with their reputations intact. The graveyard of value destruction is full of the articulate.
The Review Meeting Where the Deck Died
Quarterly portfolio review. Eight business unit GMs presenting 90 minutes each. The third GM on the agenda was the room favorite. Three-year tenure, clean narrative, relentless optimism. Revenue was up 12% cumulative. His deck opened with a slide titled “Building Momentum” and a bar chart that looked like a staircase.
I asked one question: “Can you walk me through your ROE decomposition over the last 12 quarters?”
He couldn’t. His CFO tried to rescue the moment and pulled up the numbers. Revenue up 12% cumulative. Net margin down from 9.1% to 5.8%. Asset turnover down from 2.1x to 1.6x. Leverage flat. ROE had collapsed from 31% to 14% while the GM ran his “building momentum” story. The 12% revenue growth had come from a 34% increase in inventory and a 22% stretch in receivables. He was buying growth with the balance sheet and presenting it as operational excellence.
The room went quiet in the way rooms go quiet when everyone realizes they have been watching a magic trick for three years. I asked the GM to rate his own leadership on a 1-10 scale. He said 8. I asked his team, privately, to rate him later. The team came in at a 6.2. His CFO came in at a 4. The delta between his self-assessment and his team’s assessment was 1.8 points, the widest gap in the portfolio by a factor of two.
He was transitioned out within 90 days. His successor inherited a business with $23 million in trapped working capital and a leadership culture that had been trained for three years to nod along with a story that the math did not support. Unwinding that culture took 14 months. The enterprise value cost of keeping him 18 months too long was approximately $40 million.
The Playbook
Move 1: The Confidence-Competence Matrix
Build a 2×2 for every executive in your organization. Horizontal axis: self-reported confidence, captured through a simple quarterly self-assessment on a 1-10 scale across five operating dimensions. Vertical axis: 12-quarter ROE trajectory of their business unit, decomposed into margin, turnover, and leverage. High confidence with improving ROE is your Core Team. High confidence with declining ROE is your Dangerous List. Low confidence with improving ROE is your Developmental Bench, the quiet operators who are under-promoted. Low confidence with declining ROE is your Exit Queue.
Most leadership teams are staffed with more confidence than competence because promotion systems reward articulation over outcome measurement. The Confidence-Competence Matrix forces that imbalance into visible, measurable form, which is the first step toward fixing it.
Move 2: The Quarterly DK:DuPont Review
Every quarter, every GM presents two slides before the rest of their deck. Slide one: their self-assessment on five operating dimensions, with a delta against last quarter. Slide two: their ROE decomposition across 12 rolling quarters, with a delta against the same period prior year. If the self-assessment trends up while the DuPont trends down, the meeting stops until the gap is explained. Not rationalized. Explained, with mechanism and math.
This is uncomfortable. That is the point. Comfort is what let your Dangerous GM run a stagnating unit for three years while everyone nodded along.
Move 3: Separating Dangerous from Developmental
Not every leader with a confidence-competence gap needs to be exited. Some are coachable. The test is whether, when confronted with the math, they collapse into defensive posture or move into diagnostic posture. A Developmental GM responds to a declining DuPont with “walk me through what you are seeing, because my read was different.” A Dangerous GM responds with market conditions, temporary headwinds, and a pivot to the slide where revenue is up. The first is coachable within 90 days. The second is a 30-day decision per the 30-Day Rule.
Move 4: The 90-Day Question
If I pulled your last eight quarters of DuPont data, what would it say about your self-assessment? Ask every direct report. Not in a one-on-one. In a room, with the numbers projected on the wall. The leaders who can answer honestly are the ones you keep. The leaders who explain why the numbers are misleading are telling you exactly who they are.
The delta between a leader’s self-assessment and the DuPont trajectory of the business unit they run is the most predictive single metric for future value destruction in an executive portfolio. A gap wider than 2 points sustained across three quarters is a leading indicator of a business unit that will require a turnaround within 24 months.
Monday Morning
Pull the ROE decomposition for every business unit in your portfolio over the last 12 quarters. Cross-reference against the last three 360-reviews for each GM. The names where self-assessment runs high while DuPont runs low are the names you need to confront this quarter, not next year. Every month you wait, the enterprise value cost compounds.
For the 80/20 Matrix and the full Replacement Matrix that pairs with this diagnostic, visit toddhagopian.com/freetools. The complete framework is in The Stagnation Assassin at toddhagopian.com/book. Weekly operator conversations on leadership decisions that actually move the needle are at The Stagnation Assassin Show: toddhagopian.com/podcast.
The most dangerous GM in your portfolio is in your calendar this week. His deck will be tight. His story will be confident. The math will not match. The only question is whether you will ask.