Skip to main content

Pre-mortem analysis: imagining the failure

When you build an investment thesis on the narrative-plus-numbers foundation, you've done the hard work of articulating your story and grounding it in cash flow. But the same discipline that builds confidence can blind you to the ways your thesis can fail. A pre-mortem analysis reverses that logic: you imagine your investment has failed spectacularly, then work backward to identify which risks or miscalculations led to the collapse. This forced exercise in pessimism catches the failure modes that optimism leaves invisible.

Quick definition: A pre-mortem is a structured process where you assume your investment thesis has failed—the stock has underperformed, the narrative collapsed, or the numbers never materialized—then identify the specific ways that failure could have happened, before it actually does.

Key takeaways​

  • Pre-mortems force you to articulate failure modes rather than hope they don't exist.
  • The exercise reverses the narrative logic: instead of "why this works," you ask "how this breaks."
  • Hidden assumptions—about competition, management, market timing, or business durability—emerge when you systematically imagine failure.
  • The best pre-mortems distinguish between risks you've underestimated and risks you've simply ignored.
  • Pre-mortems don't prevent failure; they make failure, when it comes, less surprising and easier to act on.

The psychology of pre-mortem thinking​

Humans are naturally good at constructing stories that work. We gather evidence that supports our thesis, interpret ambiguous data charitably, and frame uncertainties as temporary. This is the same mental machinery that makes the narrative-plus-numbers approach powerful—you construct a coherent story about why a company will succeed.

But the same machinery blinds you to how you could be wrong.

A pre-mortem reverses the direction of your thinking. Instead of "what conditions allow my thesis to work?" you ask "what conditions make my thesis spectacularly false?" This isn't cynicism; it's clarity. Psychologists call this "prospective hindsight"—you imagine it's three years from now, your stock has crashed 60%, and you work backward to reconstruct the path.

The power is that you're not just listing generic risks (market downturn, management turnover, recession). You're identifying specific failure modes of your thesis. Why would your narrative collapse? Where are the dependencies in your numbers that could break?

Failure modes in the narrative​

Every investment narrative rests on a set of implicit assumptions. When you write down your story—"Company X is gaining market share because of superior technology and lower costs"—you're assuming:

  • The technology advantage is real and durable.
  • Competitors can't replicate it quickly.
  • The market actually values the advantage in purchasing decisions.
  • Management can maintain execution as the company scales.
  • The addressable market is as large as you think.

Each of these could be wrong, and a pre-mortem asks you to imagine which one actually is wrong, in a way that destroys your thesis.

For instance, if your narrative is "Tesla will dominate electric vehicles because of its manufacturing expertise and brand," a pre-mortem failure mode might be: "Established automakers matched our manufacturing efficiency faster than we anticipated, and brand advantage eroded because quality became table stakes, not a differentiator." That's specific. It tells you to watch manufacturing productivity metrics and warranty claims data, not just sales growth.

Failure modes in the numbers​

Numbers can fail in two ways: the inputs were wrong, or the output sensitivity revealed a trap.

Input failures are straightforward: you projected 15% revenue growth, but the market caps growth at 8%. You assumed a 35% gross margin, but pricing power vanished once competitors entered. You projected a 20-year competitive moat, but disruption arrived in five.

Output sensitivity failures are subtler. You built a DCF model where the terminal value is 85% of enterprise value. That means small changes in perpetuity growth assumption or discount rate destroy your entire valuation. A pre-mortem asks: "What if the terminal-value assumption is the single point of failure here? What breaks my thesis if terminal value is only 60% of value?"

The math might be right, but the assumptions are fragile.

Building your pre-mortem list​

A structured pre-mortem has three steps:

Step one: imagine failure. It's three years from now. Your thesis-based stock pick has crashed 50–70%. The narrative unraveled or the numbers never materialized. What happened?

Step two: work backward. You're a forensic analyst. The body is on the table. Reconstruct the cause of death. Was it a single catastrophic event, or the slow accumulation of missed guidance and margin compression? Was the failure in the narrative, the numbers, or the assumption that connected them?

Step three: map to today. For each failure mode, identify what would need to happen today to set that failure in motion. You're looking for leading indicators—early warnings that would tell you to exit the thesis or reassess it.

This isn't guessing. It's systematic. You're taking your narrative and asking: where does it rely on something I can't see or can't control?

Common pre-mortem failure categories​

The narrative shifts, but you don't. You own the stock because you believe in the CEO's vision of market expansion. But the CEO steps down, or the company pivots, or the market rejects the vision. Your thesis was built on a person or a bet, not on the underlying business.

Competitive intensity accelerates. Your margin assumptions assume modest competition. But a larger, better-capitalized rival enters the market, or new entrants cluster in a hot market, compressing margins faster than you modeled. The business is good, but not special good.

Management executes, but inefficiently. The story was right—the market was there, the product was solid. But execution was sloppy. Working capital ballooned. R&D spending spiraled. The company grew revenues but destroyed margin. Growth is not value if the cost of growth exceeds the profit it generates.

The numbers don't decouple from the narrative. You built a story about a cash-generative, efficient business. But as it scales, it's burning cash. The "narrative break" reveals that your understanding of the business model was superficial. The numbers were right in form, but not in substance.

Market timing was the whole thesis. You knew the business was mediocre but cheap, and you were betting on mean reversion or a sector revival. But sector reverts downward instead. The market decides your "value trap" is actually just trapped. The story that justified buying cheap turned out to be a story you told yourself.

Hidden leverage or off-balance-sheet risk. The numbers looked solid, but you missed operating leases reclassified as debt, or contingent liabilities, or a large pension obligation that's underfunded. The balance sheet was quietly fragile all along.

Creating the "failure evidence" checklist​

For your specific thesis, build a pre-mortem evidence checklist. These are the leading indicators—the things you'd observe today that would tell you the failure mode is already in motion.

If your failure mode is "management loses control of costs," your evidence checklist includes: gross margin compressed by 200 basis points year-over-year, operating leverage is negative (sales up, operating income down), selling and admin costs are rising faster than sales, guidance is being missed on the cost side.

If your failure mode is "competitive intensity accelerates," your evidence checklist includes: new competitor announcements, share-of-voice losses in customer surveys, pricing pressure (average selling price declining), customer churn rising, deal sizes shrinking.

If your failure mode is "the market rejects our narrative," your evidence checklist includes: customer acquisition costs rising while conversion rates fall, repeat purchase rates declining, Net Promoter Score falling, the company increasing sales spending to maintain growth rate.

The checklist becomes your early-warning system. You're not waiting for the stock to crash to find out you were wrong. You're monitoring the inputs to failure.

Pre-mortem and conviction levels​

A solid pre-mortem doesn't kill your thesis; it clarifies what you're actually betting on and what would break that bet. A thesis that survives a ruthless pre-mortem might look more fragile in some dimensions, but it's clearer.

If you walk through a pre-mortem and conclude "this investment would fail if almost anything goes slightly wrong," that's a signal that conviction should be low or position size should be small. The thesis is too fragile.

If you walk through a pre-mortem and conclude "this investment would fail only if X, Y, and Z all break," and those three things seem unlikely based on current data, your conviction can be higher. You know what you'd need to see to exit.

Real-world examples​

Netflix pre-mortem, 2011. The narrative in 2010 was: "Netflix is becoming a streaming company." But a pre-mortem asked: "What if content licensing costs spiral faster than revenue growth?" The answer: they did. Content deals became more expensive as studios recognized streaming value. Netflix had to choose between margin and growth. The thesis didn't fail, but understanding that tension would have clarified that Netflix was a different beast than the DVD-rental business it was exiting.

Amazon pre-mortem, 2015. The narrative: "AWS is a rapidly growing, highly profitable cloud business funding lower-margin retail." A pre-mortem failure mode: "What if we can't maintain AWS margin dominance as Google and Microsoft enter aggressively?" The answer: AWS margins compressed from 30%+ to the low 20s as competition intensified. The narrative survived, but the path to profitability stretched. Knowing that tension upfront would have clarified the real debate: does AWS stay defensible, or is it a commoditizing market?

Peloton pre-mortem, 2020. The narrative: "Peloton is a high-margin, subscription-based fitness platform with a loyal user base." A pre-mortem failure mode: "What if post-pandemic demand normalizes faster than we assume, and subscriber churn rises because home fitness loses novelty?" The answer: exactly that. Users who bought during lockdown canceled when gyms reopened. Peloton had built a model on the assumption of permanent behavior change. The pre-mortem failure mode mapped perfectly to what happened.

Common pre-mortem mistakes​

Conflating pre-mortem with bear case. A bear case is "the stock is bad because X, Y, and Z." A pre-mortem is more specific: "My bull thesis fails if A happens and B follows from it." The pre-mortem is disciplined to your specific narrative, not a generic list of risks.

Building pre-mortems so robust that every outcome looks bad. If you can construct a plausible failure mode for any thesis, the exercise becomes cynicism, not clarity. The point is to identify the specific failure modes of your thesis, not to list every possible global disaster. A good pre-mortem identifies three to five material failure modes, not thirty.

Not updating pre-mortems as thesis details emerge. Your pre-mortem is built on the thesis as you understood it in month one. As you learn more—about the market, the management, the competitors—some failure modes become less likely, and new ones might emerge. Revisit your pre-mortem every quarter or after material news.

Waiting for evidence to contradict before acting. A pre-mortem is useful only if you actually monitor the evidence checklist. If you identify that "competitive intensity" is a failure mode, but you never check competitor announcements or pricing data, the exercise was theater. Make the checklist actionable. Some investors set alerts or spreadsheet trackers to auto-update key metrics against pre-mortem thresholds.

FAQ​

Does a pre-mortem mean I should have low conviction in my thesis? No. A pre-mortem clarifies what you're betting on. A conviction is highest when you've walked through failure modes, identified the specific things that would break your thesis, and concluded those things are unlikely. Low conviction means you see too many ways your thesis breaks.

Should I do a pre-mortem for every stock I own? Yes, for every position that represents meaningful conviction. Smaller positions in your portfolio, or stocks you're indexing into as diversification, don't require the same rigor. But your core ideas should have a written pre-mortem.

Is a pre-mortem different from a risk assessment? Yes. A risk assessment asks "what could go wrong?" A pre-mortem asks "how would my specific thesis fail?" The pre-mortem is forward-looking and tied to your narrative, not a generic checklist of market risks.

Should I share my pre-mortem with others? Only if you want brutal feedback. A pre-mortem is most useful as a personal thinking tool. That said, talking through failure modes with a trusted co-investor or mentor can expose blind spots you missed.

How does a pre-mortem change as time passes? As the thesis matures, old failure modes become less likely (if the things you were worried about didn't happen) and new ones might emerge (if the business has changed). A pre-mortem done in month one of ownership should not look identical to the pre-mortem you'd do in year three.

Can a pre-mortem prevent actual losses? Not directly. But it can shorten the distance between "I was wrong" and "I realize I was wrong," which saves capital. More importantly, it clarifies what you need to monitor, so you can exit earlier if the thesis breaks.

What's the difference between a pre-mortem and a DCF sensitivity analysis? A sensitivity analysis asks "how does value change if assumptions shift by 10%?" A pre-mortem asks "what specific business events would invalidate my thesis?" The pre-mortem is more narrative and psychological; the sensitivity is more mathematical. Both are useful.

  • Narrative and numbers approach: The framework pre-mortems help strengthen by identifying what could unravel it.
  • Confirmation bias: Pre-mortems are a tool to counter it by forcing you to imagine disconfirming evidence.
  • Scenario analysis: Similar to pre-mortem but broader; pre-mortems focus on failure, scenarios explore multiple futures.
  • Base-rate thinking: Pre-mortems benefit from asking "how often do businesses with these characteristics fail?" before assuming your thesis is exceptional.

Summary​

Pre-mortem analysis shifts your perspective from "why this works" to "how this fails." By imagining your investment thesis has collapsed and working backward to identify specific failure modes, you expose the hidden assumptions, competitive vulnerabilities, and execution risks that optimism obscures. The pre-mortem doesn't predict the future, but it clarifies what signals would tell you the future is not what you expected. The discipline of building a pre-mortem and maintaining an evidence checklist turns gut feeling into an early-warning system. When the failure mode does emerge—and for most long-term investors, failure modes do emerge—you'll recognize it faster because you've already thought about what it would look like.

Next​

Learn how to update your thesis when new information challenges your original story: Updating the thesis: when to change your mind

Stat: Roughly 60–70% of active investors report that they revise or exit their theses within 12 months of initiation, with pre-mortem-discipline investors reporting faster exit decisions when failure signals appear.