Investment Signal
Applying structural lock-in analysis to equities and crypto assets
Analytical framework only. Not financial advice. BHI has not been backtested against historical returns. All investment decisions involve risk.
Core Thesis
The market systematically misprices structural lock-in. Markets evaluate companies through earnings, revenue growth, market share, NPS, and user counts. None of these directly measure structural switching costs. The BHI fills this gap by asking a question markets do not ask: How expensive is it for customers to leave?
B high, sentiment low. Market sees bad NPS and low adoption. BHI sees a deep gravitational well. The 'failure' is the opportunity.
B low, sentiment high. Market sees viral growth and high NPS. BHI sees zero switching costs. One better competitor captures the base.
B_struct >> B_state. Infrastructure deployed, adoption lagging. Lambda is working. Convergence inevitable. Market will be surprised by stickiness.
Five Investment Signals
Measure: Re-score quarterly. Sustained dB/dt > 0 for 3+ quarters is the strongest bullish signal.
Highest h-decay sectors: AI coding assistants, financial terminals, enterprise ERP, specialized design tools.
Equity Protocol
- Identify the revenue-critical platform (M365 for Microsoft, CRM for Salesforce, Creative Cloud for Adobe)
- Score 11 parameters from customer perspective (CIO for enterprise, median paying user for consumer)
- Compute B_struct via Observatory
- Estimate B_state from public metrics: NRR > 120% suggests B_state > 1.5; Logo churn < 5% suggests B_state > 2; DAU/MAU > 0.5 suggests high closeness; Multi-year contracts suggest o >= 7
- Compute gap. Generate combined signal.
- Re-score quarterly after earnings.
Signal Matrix
| Condition | Signal | Thesis |
|---|---|---|
| High B + Gap>0.3 + dB>0 | STRONG BUY | Deep moat, underexploited, deepening. |
| High B + Gap~0 + dB>0 | BUY | Strong moat, fully exploited, deepening. |
| High B + Gap~0 + dB~0 | HOLD | Stable fortress. Income strategy. |
| High B + dB<0 | WATCH | Moat eroding. Monitor quarterly. |
| Low B + dB>0 + t high | SPECULATIVE | Moat forming. Venture-style risk/reward. |
| Low B + dB<=0 | AVOID | No moat, no trajectory. Commodity economics. |
Crypto Protocol
Four-layer model:
Token signal: Exchange tokens (BNB, CRO) derive value from exchange lock-in.
Binance — d=8, m=6, a=7, p=7, n=9, c=8, x=4, s=4, h=6, o=7, t=6. B estimate: 3.5-4.5.
Governance tokens derive value from protocol lock-in. High B = sticky TVL. The fork myth: open source code is forkable. Liquidity is not. On-chain history is not.
Aave — d=5, m=5, a=7, p=6, n=8, c=6, x=7, s=5, h=5, o=7, t=7. B estimate: 1.8-2.5.
Native tokens derive value from developer and capital lock-in. Developer lock-in is the strongest signal — deployed contracts, toolchain specificity, ecosystem grants.
Ethereum: B = 2.5-3.5 for developers. Solana: B = 2.0-3.0 (Rust/Anchor non-portable). Rising fast.
Wallet lock-in affects chain/protocol usage indirectly.
MetaMask: c=9, m=6, x=3. B estimate: 1.5-2.0.
B_total = w_exchange x B_exchange + w_protocol x B_protocol
+ w_ecosystem x B_ecosystem + w_wallet x B_wallet
Weights sum to 1, reflect capital allocation.Portfolio Construction
Long 8-12 positions with B > 2.0 and dB/dt > 0. Hold until dB/dt negative for 2 consecutive quarters. Exit if B drops below 1.3.
Target positions where B_struct > 2.5 but B_state < 1.5. Alpha from convergence as B_state approaches B_struct. Position sizing: 50% of normal conviction (timing uncertainty).
High valuation + B < 0.7 + dB/dt <= 0. Market pricing a moat that does not exist structurally.
Limitations
- No backtesting. The thesis is logically sound but empirically unvalidated.
- B is necessary but not sufficient. High B does not save a company in a collapsing market.
- Timing gap. B predicts eventual outcome, not when. 2-5 year horizon.
- Crypto-specific: smart contract exploits can destroy lock-in overnight regardless of B.
- Scoring subjectivity. Use 3+ analyst panel with ICC validation for investment-grade analysis.
- Goodhart's Law risk if BHI becomes widely adopted.