Meltdown Responsibility Test
Choose the action that preserves your duty if catastrophe does not occur
- Difficulty
- Advanced
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 99%
The test separates an unknowable terminal scenario from the responsibility that remains in a survivable future. First classify the evidence: data, analogies to prior experience, or mere supposition. Then compare action and inaction across two branches. If the feared catastrophe occurs, ask whether either present choice still matters. If it does not occur, ask which choice fulfils the role or mandate you were given. During the Lehman crisis, Oaktree reasoned that investing would not matter if the financial system melted down, while failing to invest would mean not doing its job if the system survived. The asymmetry justified action, though it did not create confidence or eliminate fear.
Origin
Marks describes Oaktree applying this logic after Lehman failed, when it had an $11 billion distressed-debt fund ready to deploy.
Core principles
- 01Some terminal scenarios have no useful precedent
- 02Inaction is still a consequential choice
- 03A catastrophe can make both choices irrelevant
- 04The survivable branch reveals the duty that still matters
How to run it
- 1
Define the terminal fear
State the catastrophic scenario precisely and distinguish it from severe but analysable losses.
Watch out A vague end-of-the-world fear makes every decision look equally futile.
- 2
Classify the evidence
Separate available data, useful historical analogies, and unsupported supposition.
Watch out Do not disguise a lack of precedent as confident pattern recognition.
- 3
Build both branches
Compare the consequences of acting and not acting if the catastrophe occurs and if it does not.
Pro tip Focus on branches where today's decision still changes the outcome.
- 4
Locate the duty
Identify the mandate that must still be fulfilled in the survivable branch.
Watch out Fear can make inaction look neutral when it abandons a real responsibility.
- 5
Act with trepidation
Take the branch-dominant action while retaining doubt and respecting quantitative downside limits.
Watch out A duty to act does not justify unlimited size.
In the wild
Oaktree had raised capital for distress before the crisis. After Lehman failed, Marks and his colleagues concluded that investing would be irrelevant if finance collapsed, but not investing would betray their job if it survived. Quantitative analysis also showed debt purchases could break even after very large declines in company value.
→ Bruce Karsh deployed an average of $450 million a week for 15 weeks, totaling $7 billion in a quarter.
Common mistakes
Pretending the catastrophe is predictable
The test exists because terminal outcomes may have neither data nor precedent.
Treating inaction as harmless
Not acting can violate the mandate that matters if the system survives.
Ignoring quantitative protection
The branch logic should complement, not replace, analysis of price and downside.
Is it for you?
Best for
High-stakes decisions where the worst case would overwhelm either choice but survival leaves one clear obligation.
Not ideal for
Situations where one option materially improves survival in the catastrophic branch and that benefit can be analysed.
From the transcript
“Well, at the time of the Lehman bankruptcy, we had no data and no prior experience. We only had supposition.”
“But if we don't invest and the financial world doesn't melt down, then we didn't do our job.”
“So we have to do it.”
From the episode
Howard Marks: how I make money while you worry about a market crash
Howard Marks