מסמך טכני מיומן המחקר · באנגלית, כפי שנכתב בשלב העבודה
Stage 129 — equity valuation and investment sensitivity
Test four equity quotation assumptions (constant real, fixed nominal, profit multiple, real drawdown), three productive-investment targets (residual, 5% and 7.5% of output), four capital-output elasticities, three recipient consumption floors and five retention rates.
For each identical scenario key, compare 25–100% retention with 0% retention. Hold owner spending response at 0.5 and collection ceiling at 0.1, to isolate the sensitivity grid from additional behavioral dimensions. The full grid contains 720 paths (144 matched cells with five rates). Report completion, qualified success and distinct failure reasons. Do not confuse zero-quote failures with corporate cash shortages. Scenario counts are not empirical probabilities.
The grid is declared and uncalibrated. The equity quotation rules are stylized assumptions; this is not a causal estimate. Results to be filled after successful CI.
Reproduce: python scripts/pricing_investment_sensitivity_v129.py --output data/processed/pricing_investment_sensitivity_v129.json
Verified results — GitHub Actions run 407
The 720-path grid completed with successful research, web and frontend checks. Each retention rate has 144 scenario paths.
| Retention | Completed | Qualified | Cash/capital/equity combined guard stops | Zero equity quote stops |
|---|---|---|---|---|
| 0% | 132 | 132 | 12 | 0 |
| 25% | 120 | 93 | 24 | 0 |
| 50% | 120 | 69 | 24 | 0 |
| 75% | 120 | 48 | 24 | 0 |
| 100% | 90 | 18 | 18 | 36 |
Under the profit-multiple quotation rule, all 36 scenarios at 100% retention stop because the dividend-based quote becomes zero, even though this is a valuation-rule artifact and not evidence of corporate insolvency. At 100% retention, the other three price rules each complete 30 of 36 paths and qualify 6 of 36. At 0% retention, each of the four quotation rules completes and qualifies 33 of 36.
Across all four price rules, the 25–75% retention paths have 120 completions per rate, compared with 132 at zero retention. This narrower sensitivity grid holds owner spending response and collection ceiling fixed, so its counts are not directly interchangeable with the larger v124 grid. The combined guard category has not been independently decomposed for this new grid. The experiment remains uncalibrated and noncausal.