postdoc.jp

Valuation Framework

The Methodology

GX assets are not spreadsheets. They are options on a non-stationary world. We price them accordingly.

Why DCF Fails

Discounted Cash Flow assumes a deterministic future: a fixed price curve, a fixed discount rate, a fixed technology. GX assets violate all three. Carbon prices jump on policy shocks. Hydrogen economics flip on electrolyzer learning curves. Ammonia arbitrage bends under CBAM. Static models produce static mistakes.

Four Pillars

01  V = DCF + Σ option_i

Real Option Valuation

Every GX asset embeds optionality: defer, expand, contract, switch fuel, abandon. We price that flexibility using Black-Scholes extensions and binomial lattices calibrated to observable markets.

02  dP = κ(θ−P)dt + σdW + J·dN

Stochastic Carbon Price

EUA and J-Credit prices follow mean-reverting jump diffusions. We fit the process to historical EEX auctions and simulate forward scenarios under policy regimes (tightening, loosening, CBAM extension).

03  E[NPV] ± σ, VaR₉₅, ES₉₅

Monte Carlo Aggregation

10,000-path simulation over the joint distribution of JCC, EUA, exchange rates, and technology cost curves. Output is a probability-weighted NPV distribution, not a single number.

04  HCI = Σ wᵢ · score(team)

Human Capital Index

PhD-level expertise is a measurable factor in project viability. We quantify domain coverage, publication depth, and prior deal experience into a scalar that multiplies execution probability.

Model Registry

Every valuation is a Git commit. Every parameter change is auditable. Every version can be reproduced against pinned data snapshots. No model leaves the registry without a passing test suite.

  • Git-versioned Python / Julia / R notebooks
  • Pinned data snapshots with SHA-256 manifests
  • Automated sensitivity test suite per commit
  • Public audit log of rating changes

Read the Protocol

The Ammonia Protocol is our first applied hackathon: build a model under these constraints, publish it in the registry.

The Ammonia Protocol →