The Analyst Paper Portfolio Learning Loop
Turn research recommendations into a tracked record of judgment and growth
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 97%
The paper portfolio learning loop gives every analyst a simulated portfolio within a defined area of expertise. Analysts first deconstruct businesses, estimate intrinsic value, and rank the securities they believe are mispriced. Those recommendations enter a paper portfolio whose performance and decision history are tracked over time. The record serves two purposes: it communicates actionable ideas to portfolio managers and creates a controlled learning environment for the analyst. Reviews can identify recurring analytical biases, reveal which kinds of situations suit the analyst, and separate isolated outcomes from persistent skill. At the firm level, the system channels the highest-conviction ideas from specialists closest to individual businesses into portfolio construction while allowing responsibility to grow alongside demonstrated judgment.
Origin
Extracted from Masters in Business during Graeme Forster's account of Orbis's paper-portfolio system for analysts.
Core principles
- 01Analysts learn faster when they must express research as decisions.
- 02A tracked portfolio exposes recurring strengths and biases.
- 03Deep specialist knowledge should feed portfolio-level capital allocation.
- 04Decision records help distinguish repeatable skill from isolated luck.
- 05The best ideas often come from specialists closest to the businesses.
How to run it
- 1
Define the analyst's niche
Give each analyst a coherent sector, geography, or business category to study deeply. The scope should be broad enough to compare alternatives but narrow enough to develop expertise.
Pro tip Choose boundaries that reflect real economic relationships rather than arbitrary index classifications.
Watch out A niche that is too broad encourages shallow coverage and weak comparisons.
- 2
Deconstruct the businesses
Analyze the operating drivers, assets, liabilities, cash flows, competitive position, and management decisions of each candidate. Rebuild the business model into an independent estimate of intrinsic value.
Pro tip Write down the assumptions that account for most of the valuation.
Watch out Do not let the current market price become the anchor for estimated value.
- 3
Rank mispriced opportunities
Select and order the securities with the greatest gap between price and estimated intrinsic value. Make the analyst express relative conviction rather than issuing disconnected opinions.
Pro tip Keep the highest-conviction list short enough that each inclusion has consequences.
Watch out Avoid filling the portfolio with low-conviction names merely to appear diversified.
- 4
Track the paper portfolio
Record selections, sizing, thesis, valuation, and subsequent performance. Preserve the original reasoning so later reviews are not rewritten with hindsight.
Pro tip Timestamp changes and require a reason whenever a position is added, resized, or removed.
Watch out Performance without an accompanying decision record provides weak evidence about skill.
- 5
Review process and biases
Compare the original thesis with what happened and identify recurring strengths, blind spots, and behavioral tendencies. Judge the process separately from short-term outcomes.
Pro tip Ask whether the same reasoning would remain sound across many similar cases.
Watch out Do not punish a good probabilistic decision solely because one outcome was unfavorable.
- 6
Feed the best ideas upward
Present the analyst's strongest recommendations and supporting evidence to portfolio managers. Use the tracked record to calibrate how much weight each recommendation receives.
Pro tip Focus discussion on the top three or four ideas where specialist insight is greatest.
Watch out A paper portfolio does not eliminate the portfolio manager's responsibility for sizing and aggregate risk.
In the wild
An analyst covering financial companies tears each candidate into its component economics, estimates intrinsic value, and ranks the most mispriced securities in a paper portfolio. Over time, the analyst's top three or four recommendations repeatedly contribute useful ideas to the live portfolio.
→ Deep specialist work becomes a repeatable source of portfolio performance and evidence about the analyst's strengths.
Years of paper decisions reveal that an analyst is especially effective when evaluating neglected cash-generative companies but weaker with speculative growth businesses. The firm directs that analyst toward the situations where the record demonstrates an edge.
→ Responsibilities are matched to observed skill rather than generic seniority.
Common mistakes
Tracking returns without reasoning
A performance number alone cannot show whether the result came from sound analysis, hidden risk, or luck.
Using paper portfolios as punishment
The system should support learning and calibration; a blame-oriented review encourages analysts to conceal uncertainty and avoid differentiated ideas.
Ignoring implementation differences
Live portfolios face liquidity, sizing, correlation, and client constraints that a paper portfolio may not capture.
Is it for you?
Best for
It is best for investment organizations with specialist analysts whose work must inform real capital-allocation decisions.
Not ideal for
It is not ideal when paper results are treated as mechanically equivalent to live portfolios with liquidity, sizing, and behavioral constraints.
From the transcript
“all of our analysts run paper portfolios.”
“it's a learning mechanism as a recommendation mechanism for portfolio managers and thinking about how to allocate capital.”
“the top three or four ideas coming from key analysts who are really deep in the weeds, generate a lot of our performance.”
From the episode
Graeme Forster on Global Longevity Investments
Graeme Forster