The REWF Framework V4.5: Production Calibration of Buyer-Specific Resource Access Under 18-Year Historical Validation, 40-Nation Supplier Coverage, and Stock-Only Endurance

Abstract

This paper presents Version 4.5 of the REWF (Resource Energy War Front) Framework. V4.5 retains the entire mathematical structure of V4.3 — buyer-specific Depletitude D_opt(c) under orthogonal constraints (W, Π, λ, P), segment-decomposed σ across eight route segments, the extended τ composing stock with flow, and political decision constraints captured in the S_eff scaling — and adds six production-calibration extensions that elevate the framework from research-grade to production-grade for institutional consumption.

The first extension is Data Source Unification: V4.3 had a structural divergence between the live API and the historical replay. V4.5 routes both paths through a single build_access_for_date() function so that the API value at time t is mathematically identical to the historical CSV value at time t — a property V4.3 could not claim.

The second extension is the 18-Year Historical Backbone. V4.5 extends the operational window from 4 years (1,082 trading days) to 18 years (4,613 trading days, 2008-01-02 → 2026-04-22), with sanctions calibration expanded from 21 to 57 rows. The backbone reveals empirical regularities invisible at 4-year scale: D_opt's monotonic +10.6% rise, the 2015–2016 Trump-max-pressure peak, the post-Fukushima Japan PDD = 35 days on 2011-06-13, and the 93.93% prevalence of negative Narrative-Reality Gap.

The third extension is Supplier Universe Expansion. V4.3 calibrated W and Π for 18 Tier 1 suppliers. V4.5 adds 22 Tier 2 suppliers, bringing total coverage to 40 nations. The Tier 2 layer captures the diversification suppliers institutional buyers actually use to hedge against Tier 1 concentration.

The fourth extension is the Stock-Only Endurance Model v5. V4.3's countdown layer admitted crude tanker arrivals × yield as a contribution to refined-product days-of-supply, double-counting future refining. V4.5 enforces product_stock_only_v5 — refined product days are computed from refined-product inventory only, with crude-yield contribution permanently disallowed at the calculation boundary. METI primary-source PRODUCT_INVENTORY values replace previous estimates that overstated gasoline inventory by 1.84×.

The fifth extension is AIS-Based Departure Tracking. V4.5 introduces VesselFinder integration scraping 25 Middle East ports of origin for tanker departures, filtering to DWT ≥ 40,000 with confirmed crude or oil-products cargo, reading draft ratios for realized loading rates. The departure layer captures the upstream half of the supply chain that the arrival layer cannot see.

The sixth extension is the Narrative-Reality Gap. NR Gap is the daily differential σ_news − σ_brent. V4.5's 18-year backbone reveals that the gap is negative (physical prices moving first) on 93.93% of trading days. This statistical regularity is the empirical foundation of REWF's commercial differentiation: REWF measures physical reality before media narrative catches up. The remaining 16.4% of days where news leads physics are flagged as "narrative-leading" events worth analyst attention.

V4.5 does not deprecate V4.3. Every V4.3 equation is preserved verbatim. The V4.5 changes are at the calibration layer (data sources, supplier coverage, inventory model) and at the observability layer (18-year backbone, AIS departures, NR Gap), not at the mathematical layer.

For Japan in May 2026, V4.5 produces D_opt(JP) ≈ 8.62 against a global D_opt^V4.5 ≈ 4.57, with S_eff^V4.5(JP) ≈ 10.0 (P-saturated). Compound NCI(JP) ≈ 8.62. These values are tracked daily and reproducible to the third decimal between the live API and the historical CSV.

1. Core Architecture — V4.5

1.1 Preserving V4.3, Adding Production-Grade Calibration

The REWF Framework retains exactly three quantitative constructs — Depletitude, Severitude, Civilization Level — and one integration variable, NCI. V4.5 does not add a fourth construct, nor does it modify the equations of V4.3. The V4.5 changes operate outside the equation layer: at the calibration data layer (what numerical values are fed into V4.3's equations), at the inventory model layer (how refined-product days-of-supply are computed downstream of D_opt), and at the observability layer (what additional diagnostic streams are exposed alongside V4.3's outputs).

A V4.3 implementation can be promoted to V4.5 by replacing its A_i source with build_access_for_date(today), importing Tier 2 suppliers, replacing its naphtha calculation with product_stock_only_v5, and subscribing to the Mid East departures and NR Gap streams. Nothing in the V4.3 mathematics changes.

1.2 Design Principle: Orthogonality of Constraints (V4.3-preserved)

V4.5 inherits V4.3's strict orthogonality requirement. Three distinct obstructions to resource flow exist, and they must not be conflated:

• Political willingness (W) — does the seller want to sell to this specific buyer? Event-driven, shaped by sanctions regimes and alliance posture.
• Port infrastructure (Π) — can the seller physically load the claimed volume? Structural, slow-moving, measured in VLCC-capable throughput.
• Logistics passability (λ) — can the ship complete the voyage on the buyer's contracted route? Kinetic, fast-moving, driven by segment-level σ.

These three do not substitute for each other. V4.5 adds one further orthogonal layer: the data calibration source, independent of W/Π/λ/P, concerning only how the underlying market observables (TC, PD) are obtained at time t.

2. Civilization Level System

The Civilization Level architecture of V4.2 and V4.3 is preserved without modification in V4.5. The bloc-only Level 6 theorem, the Level 5 ceiling for nation-states, the intrinsic Severitude ranges for Levels 1–4, and the discrete transition rule S_eff ≥ 10.0 → L − 1 all stand.

For the May 2026 Level 5 hierarchy, V4.5 ranking is identical to V4.3: nations with high political-foreclosure scores (Japan, Germany for different reasons) sit at the upper end. Japan remains the weakest major Level 5 nation. The 18-year backbone provides historical evidence that this ranking is stable across 2011-06 (Fukushima + Libya compounding) through 2026-04 (Hormuz closure declaration).

16. Diagnostic Applications (Overview)

Under a Hormuz-escalation scenario for Japan (σ_hormuz → 0.90 with other segments at baseline), the V4.5 diagnostic chain operates as follows: λ_JP collapses from ~0.50 baseline to ~0.11; the extended τ (with crude reserves at 175 days and inflow ratio collapsing) yields ED^eff ≈ 194 days, so crude τ remains high; D_opt(JP) rises from ~8.62 toward ~9.4 as the denominator of accessible supply narrows; S_eff^V4.5(JP) reaches the 10.00 transition threshold under P-amplification; compound NCI(JP) approaches 9.4; and naphtha 27.4 days collapses to ~6 days within 60 days under sustained disruption. Each step is traceable to a named calibration value. Each constant is registered in the parameter registry. Each assumption is either an observable or a parameter on which the framework reports.

V4.5 dashboards expose the full decomposition via dedicated endpoints for segment σ, λ, D_opt(c), willingness, port constraints, Mid East departures, NR Gap, and refined-product days-of-supply. Every number displayed on a briefing binds to exactly one of these endpoints. Fixed values do not appear in REWF outputs; when an endpoint is unreachable, the surface degrades to "—" rather than masking the gap with imputed data.

17. Relationship to V4.3

V4.5 is strictly additive at the equation surface. Every V4.3 equation is computed in parallel and exposed on the V4.3 endpoints, which remain unchanged. Consumers of D_opt_global, the V4.3 D_opt(c), the V4.3 segment σ, V4.3 λ, V4.3 extended τ, and V4.3 S_eff (relabeled S_eff^V4.5 in V4.5 surfaces) do not need to migrate.

The V4.5 changes are at three layers:

1. Calibration layer. The TC, PD, W, Π values fed into the V4.3 equations are now sourced from build_access_for_date(), the 18-year backbone, and the 40-nation universe. A V4.3 implementation reading the V4.5 calibration files reproduces V4.5 results to floating-point precision.
2. Inventory layer. The refined-product days-of-supply calculation is replaced with product_stock_only_v5. This is a downstream consumer of D_opt and is independent of the V4.3 equation set proper.
3. Observability layer. New endpoints expose departure flows and NR Gap. These are pure additions; no V4.3 endpoint is removed or modified.

No V4.3 parameter has changed value in V4.5. The V4.5 calibration values for individual W_ic, Π_i, p_intensity_c are May-2026 calibrations against the V4.3 March-2026 baseline; differences reflect the passage of time and the post-Hormuz-closure events of April 2026, not framework redefinition.

19. Conclusion

V4.5 elevates the V4.3 framework from research-grade to production-grade. The mathematical structure of V4.3 — buyer-specific D_opt(c) under orthogonal W/Π/λ/P constraints, segment-decomposed σ, extended τ, P-scaled S_eff — is preserved verbatim. The V4.5 contribution is in the calibration data, the supplier coverage (40 nations), the inventory model (stock-only v5 with METI primary sources), and the observability streams (AIS departure tracking, NR Gap historical regularity).

The 18-year historical backbone is the empirical foundation. It validates that the framework reproduces well-known historical episodes — Fukushima compounding 2011-06, Trump max-pressure 2018-2019, 2020 COVID demand collapse, Ukraine 2022, Hormuz closure 2026 — under consistent calibration. It also reveals empirical regularities that are not visible at 4-year scale: NR Gap 93.93% physical-leading, monotonic D_opt secular trend, and individual events traceable to W/Π/λ/P decomposition.

The Stock-Only Endurance Model v5 closes the most embarrassing operational gap of V4.3: refined-product days-of-supply are now computed from refined-product inventory only, with crude × yield contribution permanently disallowed at the calculation boundary. Naphtha 27.4 days, gasoline 14.2 days, diesel 15.2 days, kerosene 62.4 days — calibrated to METI primary-source inventory and prior-year-same-month demand — are the V4.5 production headline numbers.

The Narrative-Reality Gap — physical prices leading news on 93.93% of 4,613 trading days — is the empirical claim that distinguishes REWF's product surface from credit-rating, financial-CPE, and news-sentiment competitors. This is not a marketing assertion. It is an 18-year statistical regularity.

The purpose of V4.5 remains what V4.2 articulated and V4.3 deepened: temporal awareness. Knowing how many days remain. Knowing that the number is smaller than yesterday. Knowing which suppliers, which routes, which decisions made the number smaller. V4.5 reports the number to the third decimal, reproducible between live API and historical CSV. V4.5 reports the source of every change. V4.5 maintains the audit trail across 4,613 trading days.

A model whose outputs only confirm intuitions has no analytical value. A model whose outputs diverge from intuition has analytical value if and only if the divergence is traceable to named assumptions. V4.5 is constructed so that every divergence is traceable, every assumption is either an observable or a registered parameter, and every value is reproducible against build_access_for_date(t).

MASTER AI — Pioneering the Future of Intelligence
REWF Working Paper Series, WP-2026-014

References

• BP (2024). Statistical Review of World Energy 2024.
• Caldara, D., & Iacoviello, M. (2022). Measuring geopolitical risk. American Economic Review, 112(4), 1194–1225.
• Dusseault, M. B. (2001). Comparing Venezuelan and Canadian heavy oil. CIPC 2001-061.
• Erisman, J. W. et al. (2008). How a century of ammonia synthesis changed the world. Nature Geoscience, 1, 636–639.
• Financial Stability Board (2023). Cross-border Payments Monitoring Report.
• Hubbert, M. K. (1956). Nuclear energy and the fossil fuels. American Petroleum Institute.
• Klare, M. T. (2012). The Race for What's Left. Metropolitan Books.
• Le Billon, P. (2012). Wars of Plunder. Columbia University Press.
• McGlade, C., & Ekins, P. (2015). Geographical distribution of fossil fuels unused when limiting warming to 2°C. Nature, 517, 187–190.
• Nincic, D. (2002). Sea lane security and U.S. maritime trade. Naval War College Review, 55(4).
• OPEC (2024). Annual Statistical Bulletin.
• Ross, M. L. (2012). The Oil Curse. Princeton University Press.
• Smil, V. (2001). Enriching the Earth. MIT Press.
• Sovacool, B. K., & Mukherjee, I. (2011). Conceptualizing and measuring energy security. Energy, 36(8), 5343–5355.
• SPE (2018). Petroleum Resources Management System.
• Tainter, J. A. (1988). The Collapse of Complex Societies. Cambridge University Press.
• UN Comtrade Database (2008–2026).
• U.S. Energy Information Administration (2008–2026). Country Energy Profiles.
• Yergin, D. (2011). The Quest. Penguin Press.
• METI 「資源・エネルギー統計」 (2008-01 〜 2026-04, transcribed via 石油連盟 PAJ-02 series).
• Petroleum Association of Japan (PAJ) (2025). 製油所の所在地と原油処理能力 (2025年3月末現在). PAJ-04.
• VesselFinder (2026). Recent Departures, Mid East ports.
• yfinance (2008–2026). Brent, WTI, NG, ZW, ZC, ZS daily closes.

MASTER AI — Pioneering the Future of Intelligence
REWF Working Paper Series, WP-2026-014