Metcalfe's Law & Bubble Physics

*A network is worth something because people are on it. That is the whole idea, and it is old — the telephone, the fax, the web all obeyed it. In 2019 a team led by Sornette, the physicist who spent a career forecasting crashes, turned the same lens on Bitcoin in a peer-reviewed Royal Society journal. Their verdict was double-edged: a fundamental value does exist, and it grows with the network of users — but it had been heavily exceeded, on at least four occasions*, by bubbles that grew super-exponentially and burst.

The paper fuses two instruments. The first is a generalized Metcalfe's Law. Robert Metcalfe's textbook claim was that a network's value scales with the square of its users — the number of possible connections among n nodes is roughly n². Applied to Bitcoin, the 'users' are proxied by active addresses, and the market capitalization is regressed against that count. Crucially, the authors do not impose the exponent of 2; they fit it, and the data prefer a gentler slope: β ≈ 1.69. Value still grows faster than linearly with the network, but a touch slower than n². That fitted exponent is the fundamental-value anchor.

The second instrument is the Log-Periodic Power Law Singularity (LPPLS) model, Sornette's signature tool for detecting bubbles. Its premise is behavioural: in a bubble, price does not merely rise, it rises faster than exponentially. Traders imitate each other, herding and momentum feed on themselves, and the growth rate itself accelerates toward a mathematical singularity — a critical time at which the trajectory becomes unsustainable. Superimposed on that acceleration are log-periodic oscillations, the fingerprint of a market oscillating with growing urgency as it approaches its breaking point.

Put the two together and you get a diagnostic rather than a price target. Metcalfe tells you where fundamental value roughly sits given the size of the network. LPPLS tells you when price has detached from any fundamental and entered a self-reinforcing, faster-than-exponential regime that historically ends in a crash. The paper reports that across Bitcoin's history these bubble regimes appeared and burst on at least four occasions, and that LPPLS gave ex ante — before the fact — warning of elevated crash hazard, along with a probabilistic bracket for when the correction was likely.

The intellectual significance is where this module earns its place in the academic wing. This is arguably the most rigorous valuation framework published on Bitcoin, and it explicitly frames network-based value as an alternative to scarcity narratives — the digital-scarcity, gold-analogy story from which stock-to-flow later grew. Its claim is that fundamental value flows from adoption — from real users transacting on the network — not from the supply schedule engraved in the protocol. Where scarcity models say the halving makes coins rare and therefore valuable, Metcalfe says a coin nobody uses is worth nothing regardless of how few exist. The two worldviews are not compatible, and the peer-reviewed one sides with users.

Metcalfe (generalized): MarketCap ≈ 10^α · (active addresses)^β,  fitted β ≈ 1.69 (s.e. 0.0076), intercept α ≈ 1.51  (Metcalfe's textbook β = 2)
LPPLS (bubble regime): ln p(t) ≈ A + B(t_c − t)^m + C(t_c − t)^m · cos[ω ln(t_c − t) − φ]
  where t_c = critical time (crash hazard peaks), 0 < m < 1 (super-exponential), ω = log-periodic angular frequency