Lodestar Protocol v0.1·published 2026-05-06·sha256:7f83b1626...e3c2d9a4

Lodestar Protocol

Version 0.1 · Published May 2026 · SHA256: 7f83b1626...e3c2d9a4

The complete methodology documentation for Lodestar valuations. Every step from inputs to valuation range is documented here.

Read ProtocolDownload PDF

Overview

The Lodestar Protocol defines how we derive valuation ranges for private companies. It is designed to be auditable, reproducible, and defensible. Every Lodestar valuation follows this protocol.

Core Principles

  • Transparency: Every step is documented and auditable.
  • Reproducibility: Given the same inputs, any analyst should reach the same range.
  • Falsifiability: Our predictions can be tested against outcomes.

Three-Lens Framework

Every valuation uses three independent analytical lenses. These lenses triangulate toward a defensible range.

LENS 1

Transaction Comparables

M&A and secondary transactions with similar characteristics

LENS 2

Trading Multiples

Public company benchmarks with private company adjustments

LENS 3

Scenario Modeling

Bull/base/bear cases with explicit assumptions

Comparable Transactions

We identify relevant M&A and secondary transactions based on:

  • • Business model similarity (SaaS vs. marketplace vs. hardware)
  • • Scale proximity (revenue within 0.5x-2x range)
  • • Market conditions (recency-weighted, with higher weight to recent transactions)
  • • Geographic relevance (US, Europe, emerging markets)

Each comparable is source-graded (A/B/C) based on data reliability.

Trading Multiples

Public company multiples require adjustment for private company factors:

  • • Liquidity discount: 15-30% depending on company stage and anticipated exit timeline
  • • Size adjustment: Smaller companies typically trade at lower multiples
  • • Growth premium/discount: Adjusted based on relative growth rates
  • • Profitability adjustment: Rule of 40 considerations

Scenario Modeling

Three scenarios with explicit assumptions:

Bear Case

Growth decelerates, margins compress, market conditions deteriorate. Not a worst-case scenario, but a realistic downside.

Base Case

Current trajectory continues with reasonable assumptions about market and execution.

Bull Case

Execution exceeds expectations, market conditions improve. Not a blue-sky scenario, but a realistic upside.

Confidence Grades

Every valuation includes an explicit confidence grade:

AHigh confidence: Strong comparable data, clean financials, well-understood business model
B+Good confidence: Adequate comparables, some data gaps, standard adjustments
BModerate confidence: Limited comparables, material assumptions required
B-Lower confidence: Few comparables, novel business model, significant uncertainty
CLow confidence: Highly speculative, insufficient data for reliable valuation

Outcome Humility§ 3.4

Lodestar publishes accuracy claims only when the data supports them. We do not assert calibration performance until enough resolved outcomes exist to make the claim falsifiable.

The threshold

Cohort-level accuracy claims require 50+ resolved outcomes per cohort. Until that bar is met, we publish coverage and methodology — not accuracy.

This is why the Lodestar Index ships in two versions: v0 (Q1 2027) publishes cohort coverage, sector volatility, and methodology references with no accuracy claims. v1 (2027+) adds confidence-grade error rates, protocol-version performance, and cohort calibration histograms once the 50+ threshold is cleared per cohort.

“When does Lodestar publish accuracy?” The honest answer: coverage and methodology are publishable now; accuracy is publishable when the data supports it — not before.

Falsifiability Tests

Every valuation is subjected to three falsifiability tests:

1. Hype Injection Test

Would the valuation change if we removed all narrative assumptions and relied only on financial data?

2. Backtest Validation

When applied to historical data, does our methodology predict past outcomes?

3. Adversarial Review

Can an independent analyst find material flaws in our analysis?

Inputs Grading

Every input is graded for reliability:

  • Grade A: Audited financials, verified transaction data, regulatory filings
  • Grade B: Management-provided financials, reputable data providers, disclosed terms
  • Grade C: Estimated figures, undisclosed terms, secondary sources

Asymmetric Impact Rule§ 5.2

Evidence quality determines what an input is allowed to do. Strong evidence can move the midpoint up or down; weak evidence can only widen the confidence band.

The rule, stated

Tier 3 inputs (self-reported metrics) and Tier 4 inputs (narrative, press releases, news) cannot raise the midpoint and cannot narrow the confidence band. They can only trigger evidence requests or widen the band.

A press release saying “we crossed $100M ARR” cannot raise the valuation. It can only trigger a request for the Stripe export that would. An audit-grade verified ARR figure can. A Tier 1 (integration-verified) comp-set update from a peer transaction can. A Tier 4 narrative cannot.

This is the falsifiability anchor of the protocol. It also makes source-grading load-bearing rather than cosmetic: integration-verified data is the difference between an Issuer-Verified memo and a Public-Only one.

Pulse cross-reference: Lodestar Pulse enforces this rule on every cascade. Tier 4 news widens the band only; Tier 1-3 comp-set updates move the midpoint deterministically; Tier 3 self-reported metrics can never raise the midpoint without verified evidence. Every cascade is logged in the assumption ledger.

Output Format

Every Lodestar valuation includes:

  • Valuation range (low, base, high)
  • Confidence grade
  • Comparable transactions with source grades
  • Key assumptions
  • Sensitivity analysis
  • Inputs hash for verification
  • Protocol version citation

Cite This Protocol

Lodestar Protocol v0.1 (May 2026). Lodestar Technologies, Inc. SHA256: 7f83b1626...e3c2d9a4. https://lodestar.finance/protocol