1 Curated Finance Prompts

Finance Prompts

Production recipes and workflows categorized under #finance.

Pernas Research
@pernasresearch

Finance research

✹GPTFinance research

Two-layer portfolio risk factor analysis

Analyzes a portfolio's observable and hidden risk factors, clusters them and keeps the analysis updated.

Original X post
Analyze the attached portfolio's risk factors in two layers. First, identify conventional and relatively observable factors, including: Interest rates and discount rates Economic growth and recession Consumer spending Credit availability and refinancing Inflation and labor costs Commodity and energy prices Foreign exchange rates Housing and commercial-real-estate activity Government spending and regulation Industry capital-expenditure cycles Equity-market liquidity and investor risk appetite Geography, customers, suppliers, and end markets Technological disruption Competitive intensity Second, search creatively for less obvious common exposures. Look for companies from apparently unrelated industries that depend on: The same customer group or source of customer funding The same industry’s capital-expenditure budget Shared suppliers, infrastructure, distribution channels, or technology platforms Electricity, cloud computing, telecommunications, payment rails, cybersecurity, or electronic connectivity Continuously available and inexpensive financing Government licenses, reimbursement, subsidies, contracts, or emergency-policy support Similar investor ownership, market liquidity, index inclusion, or susceptibility to forced selling The same assumptions about consumer behavior, demographics, globalization, or institutional stability Second- and third-order economic relationships A single external event that could simultaneously affect otherwise unrelated businesses Government intervention that might protect some assets or companies while excluding others Be imaginative but disciplined. A creative factor must have a coherent causal mechanism. Do not manufacture connections merely because two securities have historically moved together. For every material factor or cluster, explain: Which portfolio positions are exposed? What specifically connects them? What event or change would activate the factor? How would the factor travel through each business? Would it affect fundamentals, valuation, financing, market liquidity, or some combination? Would the damage likely be temporary or potentially permanent? How confident are you that the connection is economically real? Could a reasonable investor have identified and acted upon the factor beforehand? Why might the proposed grouping be misleading or overstated? Pay particular attention to companies classified in different industries that ultimately depend on the same economic engine. Recognize the limitations of factor analysis. After a crisis, it is always possible to redraw boundaries around the affected companies and give their shared exposure a new name. Distinguish among: Established factors that were identifiable and actionable beforehand Hidden factors that could have been identified through deeper analysis Background dependencies that are real but nearly impossible to avoid Factors whose importance could not reasonably have been understood before an unprecedented event Retrospective narratives that explain the past but offer little practical foresight Do not assume every shared characteristic constitutes a useful risk factor. A useful factor requires a plausible mechanism through which multiple holdings could be affected together. Coverage and exclusions: Do not stop the analysis because one or more securities cannot be adequately researched. Complete the analysis using every position for which sufficient information is available. Classify every portfolio position as: Fully analyzed Partially analyzed Excluded Mark a position partially analyzed or excluded when you cannot identify the security confidently, cannot access reliable information, do not adequately understand its business, lack necessary portfolio or thesis information, encounter conflicting information, or cannot establish its risk exposures without speculation. For every partially analyzed or excluded position, disclose: Company name and ticker Portfolio weight, if available Whether it was partially analyzed or excluded The precise reason for the limitation Which portions could and could not be analyzed What information would be needed to complete the analysis How the omission could distort the portfolio-level conclusions Calculate the total portfolio weight that was fully analyzed, partially analyzed, and excluded. Never describe the work as covering the entire portfolio when material positions were omitted. Do not invent a company’s business model or exposures to achieve complete coverage. Include tentative conclusions only when clearly labeled as low confidence. Continue the analysis without waiting for clarification. Request missing information at the end so omitted positions can be incorporated into a later update. Output requirements: Present the results in two layers. Begin with an intuitive executive summary written in clear, conversational language. It must be understandable without reading the supporting tables. Avoid excessive jargon and long lists of immaterial observations. The executive summary should explain: The three to five most important risk clusters Which companies belong to each cluster and what connects them Where the portfolio is more concentrated than it initially appears Which apparent concentrations are probably less serious than they look The most important hidden or surprising connection Which risks are conventional and readily identifiable Which risks are more creative, uncertain, or dependent on unusual circumstances The percentage of the portfolio included in the analysis Any material securities partially analyzed or excluded The bottom-line conclusion about how the portfolio is diversified Explain each cluster causally and intuitively. Do not merely name a factor such as “interest-rate risk.” Explain how it travels through the businesses and why it could cause the positions to struggle simultaneously. Prioritize factors according to the portfolio weight exposed, potential severity, and confidence in the connection. After the executive summary, provide: A. Portfolio coverage and exclusions B. Major identifiable risk factors C. Cross-industry risk clusters D. Hidden or unconventional shared exposures E. Imaginative but plausible stress scenarios F. Detailed position-by-factor analysis G. Blind spots and information needed For each important cluster, show: Shared factor or dependency Positions exposed Approximate portfolio weight exposed Transmission mechanism Potential severity Confidence in the connection Ex-ante identifiability Reasons the grouping could be overstated Use current primary sources such as regulatory filings, earnings materials, and company investor presentations. Cite material factual claims. Do not provide position-sizing, trading, or portfolio-reconstruction recommendations unless specifically requested. Maintain this as an ongoing analysis. When the portfolio changes, update the clusters, exposed portfolio weights, coverage assessment, and overall conclusions rather than starting from scratch.
·21.9k uses·View original