开展证据驱动的经济形势分析
围绕经济问题梳理数据来源、传导机制与不确定性,区分事实、推断和情景并形成分析报告。
完整提示词
I want you to perform a maximally deep, systematic, evidence-first and context-aware economic analysis of a country, region, industry, market, company or business decision. Main objective: > Determine which economic factors actually affect the problem at hand, through which channels they act, how strong the evidence is and where alternatives, tradeoffs or uncertainty exist. This is not: - a collection of macro indicators unrelated to the specific question - a political analysis, unless it directly affects the economic topic - automatic conclusions from a single indicator - an assumption that correlation means causation - a forecast without assumptions and uncertainty - mixing nominal and real data - comparing different countries/periods without normalization - relying on a single source when relevant primary data exist If you use current data: - check the date - period - the revision - the jurisdiction - the methodology - source authority ## 1. DEFINE QUESTION Before the analysis, write down: ```text Economic question: Geography: Period: Population/market: Decision/use case: Key unknowns: ``` ## 2. SOURCE HIERARCHY Prefer: ```text official statistical office / central bank / regulator > international institution > audited/company data > reputable research > secondary analysis > media > anecdote ``` ## 3. DATA VINTAGE Which version of the data? ## 4. REVISIONS GDP, employment and similar data get revised. ## 5. NOMINAL VS REAL Mandatory distinction. ## 6. PER CAPITA ## 7. PPP Use only where relevant. ## 8. LEVEL VS GROWTH ## 9. RATE VS CHANGE IN RATE ## 10. STOCK VS FLOW ## 11. SEASONALLY ADJUSTED ## 12. ANNUALIZED ## 13. BASE EFFECT ## 14. INDEX BASE YEAR ## 15. POPULATION DENOMINATOR ## 16. LABOR FORCE ## 17. INFORMAL ECONOMY If relevant and source-supported. ## 18. GDP ## 19. GDP COMPONENTS - consumption - investment - government - net exports ## 20. PRODUCTIVITY ## 21. EMPLOYMENT ## 22. UNEMPLOYMENT ## 23. PARTICIPATION ## 24. WAGES ## 25. REAL WAGES ## 26. INFLATION Distinguish headline, core and sector inflation, the price level from the rate of change, and nominal from real values. ## 27. INTEREST RATES Distinguish the policy rate, market yields, the credit spread and the actual cost of borrowing; transmission works with a lag. ## 28. FX Distinguish the nominal and real exchange rate and transaction, translation and economic exposure; the exchange rate does not follow the interest differential deterministically. ## 29. CREDIT ## 30. MONEY ## 31. HOUSEHOLD BALANCE SHEET ## 32. CORPORATE BALANCE SHEET ## 33. GOVERNMENT BALANCE ## 34. FISCAL POLICY ## 35. DEFICIT ## 36. PUBLIC DEBT ## 37. EXTERNAL BALANCE ## 38. CURRENT ACCOUNT ## 39. CAPITAL FLOWS ## 40. SAVINGS ## 41. INVESTMENT ## 42. BUSINESS CYCLE ## 43. LEADING INDICATOR ## 44. LAGGING INDICATOR ## 45. STRUCTURAL VS CYCLICAL Critical distinction. ## 46. SUPPLY SHOCK ## 47. DEMAND SHOCK ## 48. POLICY SHOCK ## 49. TECHNOLOGY SHOCK ## 50. DEMOGRAPHICS ## 51. MIGRATION ## 52. URBANIZATION ## 53. ENERGY ## 54. COMMODITIES ## 55. TRADE ## 56. TARIFF ## 57. SUPPLY CHAIN ## 58. REGULATION Only economic effect, unless broader scope requested. ## 59. MARKET STRUCTURE ## 60. COMPETITION ## 61. CAPACITY ## 62. EXPECTATIONS ## 63. CONFIDENCE ## 64. CONSUMER SENTIMENT ## 65. BUSINESS SENTIMENT ## 66. CAUSAL CHANNEL For each major driver, write: ```text Driver ↓ Transmission channel ↓ Affected variable ↓ Observed evidence ↓ Economic impact ``` ## 67. CORRELATION Do not infer causation without mechanism/evidence. ## 68. COUNTERFACTUAL What would likely happen without driver? ## 69. SECOND-ORDER EFFECT ## 70. DISTRIBUTIONAL EFFECT Different groups may experience different effects. ## 71. SHORT RUN ## 72. LONG RUN ## 73. ELASTICITY ## 74. SUBSTITUTION ## 75. COMPLEMENTARITY ## 76. MARKET CLEARING Only where applicable. ## 77. CONSTRAINT ## 78. BOTTLENECK ## 79. SCENARIO ## 80. BASE CASE ## 81. DOWNSIDE ## 82. UPSIDE ## 83. PROBABILITY Do not invent probabilities. ## 84. SENSITIVITY ## 85. FORECAST Clearly separate observed from forecast. ## 86. FORECAST HORIZON ## 87. FORECAST ERROR ## 88. CONSENSUS Consensus is not truth, but useful benchmark. ## 89. ALTERNATIVE VIEW Represent credible alternatives. ## 90. FALSE POSITIVE RULES Do not automatically conclude: - GDP growth = broad prosperity - unemployment down = labor market universally strong - inflation down = prices falling - rate cut = easier conditions for everyone - currency appreciation = stronger economy - trade deficit = economic weakness without the relevant context. ## 91. EVIDENCE TIERS ```text A - primary official or audited quantitative evidence B - multiple high-quality consistent sources C - strong derived analysis with transparent calculation D - plausible interpretation/inference E - scenario/hypothesis ``` Materiality is the severity scale of this analysis: **HIGH** changes the overall conclusion or the decision; **MEDIUM** changes a key metric, driver or trend but not the conclusion; **LOW** has a limited effect and is noted for completeness. Judge it against an explicit base (the value of the economic question: GDP, the market, the company's revenue or the decision) and state that base. A finding is CONFIRMED only with evidence tier A or B; tier C is SUPPORTED; tier D stays NOT VERIFIED, and tier E is a SCENARIO. When credible sources disagree, the status is CONTESTED and both views are shown. Never present an estimate, a forecast or a scenario as a fact. ## 92. STATUS ```text CONFIRMED SUPPORTED NOT VERIFIED CONTESTED SCENARIO NOT APPLICABLE ``` ## 93. FINDING FORMAT ```text ID: Materiality: Topic: Status: Evidence tier: Geography: Period: Indicator/driver: Observed data: Comparison: Transmission mechanism: Economic impact: Affected groups: Alternative explanation: Evidence: Uncertainty: Implication: ``` ## 94. MACRO MATRIX | Driver | Current condition | Direction | Transmission | Evidence | |---|---|---|---|---| ## 95. SCENARIO MATRIX | Variable | Base | Downside | Upside | Main driver | |---|---:|---:|---:|---| ## 96. SECOND PASS Check again: - nominal vs real - base effects - revisions - seasonality - causality - structural vs cyclical - alternative explanations - distribution - lag - policy transmission - external shocks ## 97. FINAL QUALITY GATE Confirm: - question defined - geography/period - source quality - data definitions - real vs nominal - mechanisms - causality carefully handled - uncertainty - alternative explanations - scenarios separated from facts ## 98. OUTPUT `ULTIMATE_ECONOMIC_ANALYSIS.md` # FINAL RULE An economic analysis is not good when it lists many indicators. It is good when it shows: ```text what changed + why + through which channel + for whom + in which period + with what level of evidence and uncertainty ```
分类与标签
财务理财、经济分析、市场研究、宏观指标、情景分析