Doomsday Outcome Litmus Test
Trace a scary forecast through the conditions required to make it real
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 97%
Begin with the alarming outcome and reason backward through the chain of events necessary to produce it. Replace a vague question such as whether a recession will happen with a causal examination of the drivers, probabilities, interactions, and responses that would bring the economy to that point. Determine whether the warning describes a credible central scenario or merely a real but remote tail risk. The method also checks whether governments, markets, companies, or households would react in ways that interrupt the projected path. Its output is not a precise point forecast but a structured judgment about plausibility, missing links, and decision relevance. This reduces susceptibility to media alarmism while preserving attention to risks that have a credible mechanism.
Origin
Extracted from Masters in Business, where Philipp Carlsson-Szlezak described the questions he uses to challenge economic doomsday narratives.
Core principles
- 01Treat alarming outcomes as scenarios, not central forecasts
- 02Demand a causal pathway rather than a model output
- 03Assess drivers, probabilities, and interactions together
- 04Include policy, market, and social responses
- 05Distinguish genuine risks from likely outcomes
How to run it
- 1
Define the feared outcome
Translate the warning into a specific, observable outcome with a plausible time horizon. Avoid ambiguous labels such as crisis or collapse.
Pro tip Ask what measurable conditions would confirm that the outcome had occurred.
Watch out Do not begin by accepting the forecaster's probability estimate.
- 2
Build the causal pathway
Work backward from the outcome and identify the drivers and intermediate events needed to reach it. Make every assumed link explicit.
Pro tip Use a simple causal chain rather than a complex forecasting model.
Watch out A noteworthy data point is not itself a complete pathway.
- 3
Assess probabilities and concurrence
Evaluate how likely each condition is and whether the necessary conditions can occur together. Downgrade scenarios that require several unlikely events to align.
Pro tip Separate conditional probability from the probability of the entire chain.
Watch out Do not move a tail risk to the center of the distribution merely because its consequences are severe.
- 4
Add adaptive responses
Consider how policymakers, markets, businesses, and households would respond as the danger became visible. Determine whether those reactions would weaken or break the causal chain.
Pro tip Review comparable crises to identify the institutions most likely to intervene.
Watch out Static extrapolations often assume no one changes behavior.
- 5
Compare competing scenarios
Contrast the doomsday pathway with a base case and at least one milder downside case. Use judgment to determine which scenario deserves the most decision weight.
Pro tip Track indicators that discriminate among the scenarios instead of committing to one point forecast.
Watch out Do not interpret uncertainty as proof that every scenario is equally likely.
In the wild
An investment committee hears that higher rates guarantee a recession. It maps the required pathway: rates must materially reduce household disposable income, corporate investment must contract, employment must weaken, and policy responses must fail. It then observes that many US households hold long-term fixed-rate mortgages and that firms continue investing where returns remain attractive.
→ The committee retains recession as a downside scenario but does not position the portfolio as though it were inevitable.
A company hears that rising global rates will trigger widespread emerging-market defaults. Its risk team identifies which countries face refinancing needs, whether bond markets remain available, what reserves and multilateral support exist, and which failures could transmit to its markets. The team distinguishes exposed countries from the entire asset class.
→ Management creates targeted contingencies instead of making a costly global retreat.
Common mistakes
Treating a model output as a mechanism
A numerical forecast does not explain the causal conditions that would make the outcome occur. Require the forecaster to expose the pathway behind the number.
Ignoring intervention and adaptation
Markets, policymakers, firms, and households change behavior when risks become visible. Leaving those responses out makes linear disaster forecasts unreliable.
Confusing severity with probability
A scenario can be genuinely dangerous while remaining at the edge of the probability distribution. Consequence and likelihood must be assessed separately.
Is it for you?
Best for
It is best for executives, investors, and policymakers evaluating recession calls, crisis warnings, and other high-impact forecasts.
Not ideal for
It is not ideal for immediate emergencies where observable harm already requires action regardless of forecast probability.
From the transcript
“the litmus test for me is often what would it take for a certain outcome of a certain doomsday outcome.”
“walk me through the conditions that actually lead us to the precipice and then make us fall off that microeconomic cliff.”
“We need to talk about drivers, causes, We need to talk about their probabilities and their constellations.”
From the episode
Economic Risks Versus Media Alarmism with Philipp Carlsson-Szlezak