House of Views Challenge Process
Stress-test expert forecasts through structured disagreement
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 94%
The House of Views Challenge Process preserves individual expertise while creating a disciplined route to a shared forecast. The person responsible for a market, institution, or domain presents a clear view and its supporting evidence. Other specialists then test the assumptions, identify counterarguments, and explain why the conclusion could be wrong. The goal is not automatic consensus or obedience to a central house opinion; it is a healthy debate that exposes weak reasoning before publication. Once the team has examined both sides and turned every relevant stone, it adopts a working view while retaining awareness of dissent and uncertainty. The mechanism combines accountable ownership with adversarial review, producing forecasts that are more robust than either isolated opinions or imposed conformity.
Origin
Extracted from Masters in Business, where Torsten Slok contrasted a single institutional house view with Deutsche Bank’s more pluralistic house of views.
Core principles
- 01Let domain experts lead within their areas
- 02Challenge conclusions without suppressing independent judgment
- 03Test both supporting and opposing arguments
- 04Publish a view only after turning every stone
- 05Healthy disagreement improves collective judgment
How to run it
- 1
Assign ownership
Give the forecast to the person with the strongest relevant expertise and make that person responsible for presenting it.
Pro tip Separate ownership of the analysis from organizational seniority.
Watch out Shared ownership without a lead can produce vague conclusions.
- 2
State the view
Express the expected outcome, timeframe, assumptions, and important evidence in precise terms.
Pro tip Use a forecast that could later be evaluated as right or wrong.
Watch out Avoid unfalsifiable language that cannot be challenged.
- 3
Build both cases
List the strongest arguments for and against the proposed view before debating it.
Pro tip Ask the presenter to articulate the opposing case first.
Watch out Do not reduce the opposing case to a weak straw man.
- 4
Fight it out constructively
Invite specialists to challenge the mechanism, evidence, and assumptions while keeping the discussion focused on the decision.
Pro tip Distinguish factual disagreements from differences in judgment.
Watch out Personal or status-based attacks destroy the value of dissent.
- 5
Record the working conclusion
Publish the resulting view with its assumptions, unresolved objections, and conditions that would trigger revision.
Pro tip Name the evidence that would change the team’s mind.
Watch out Do not erase dissent merely to present artificial unanimity.
In the wild
The analyst responsible for the Federal Reserve presents a forecast that rates will remain on hold. Colleagues test the inflation assumptions, labor-market evidence, and plausible reasons for a cut or increase before the team publishes its conclusion.
→ The organization produces a challenged, evidence-based forecast rather than repeating one person’s assertion.
A product lead predicts that a new feature will increase retention. Marketing, finance, support, and engineering each challenge the evidence and identify failure conditions before leadership approves a measured launch.
→ The launch proceeds with clearer assumptions, monitoring thresholds, and rollback criteria.
Common mistakes
Imposing the leader’s answer
A nominal debate is useless if participants know that disagreement with the most senior person is unwelcome.
Debating without ownership
Someone must remain accountable for integrating challenges into a final working forecast.
Forcing false consensus
Document material dissent instead of pretending uncertainty disappeared during the meeting.
Is it for you?
Best for
It is best for research, investment, and policy teams making consequential forecasts under uncertainty.
Not ideal for
It is not ideal for urgent operational decisions that allow no time for structured debate.
From the transcript
“some banks have a house view, other banks have a house of use.”
“What are the arguments why this is right? One are the arguments why this is wrong. This gives a healthy debate.”
“we would sit around the table and fight it out and end up with a view”
From the episode
The US Economy and Trump Administration Policy with Apollo's Torsten Slok
Torsten Slok