Market Microstructure × Thermodynamic Field Theory

Market Edge

Kyle (1985) derived the same structure from rational expectations that the void framework derives from thermodynamics. The price impact coefficient λ maps perfectly to Pe. We tested it on 100 Polymarket wallets. 10 out of 10 tests passed.

Score a Market → Live Signals → API Docs →
0.696 ρ(Pe, win rate)
·
10/10 K-Factor tests
·
5.5× channel separation
·
10−10 regime p-value
·
— live signals
01 · THE ISOMORPHISM

Two derivations, 40 years apart, same math.

Kyle starts with rational expectations and asymmetric information. The framework starts with thermodynamic transport. Both arrive at: an opacity measure, a reactivity measure, and a constraint measure. The correspondence is monotonically perfect across 8 venue types.

σ(c) = sinh(2( bα − c · bγ ))
The shape function. c = 1 − (O + R + α) / 9. All market signal lives here. K is pure scale noise.
λ = σv / 2σu  ↔  Pe = K · σ(c)
Kyle’s price impact maps monotonically to Pe. Spearman ρ = 1.000 across 8 venues (theoretical). On real wallets: ρ = 0.696 (N=100, empirical).
Kyle (1985)

Rational expectations.

An informed trader with private signal σv submits orders against noise traders (σu). The market maker sets price to break even. Price impact λ = σv/2σu. Higher asymmetry → higher λ.

Void Framework (2026)

Thermodynamic transport.

A system with opacity O, reactivity R, and coupling α produces drift intensity Pe = K · sinh(2bnet). Higher information asymmetry → higher O → higher Pe. Same direction, same structure, different axioms.

02 · K-FACTORIZATION

Shape carries the signal. Scale is noise.

The K-Factorization theorem (§136) says every quantity factors into K-independent shape and K-dependent scale. If true, σ(c) alone predicts outcomes — K adds nothing. Four tests. All passed.

KF1: Shape ≥ Pe PASS
ρ = 0.696
σ(c) predicts win rate as well as Pe (ρ=0.676). Shape is +0.02 better. K adds nothing.
KF2: Cross-dataset PASS
ρ = 0.570
Combined N=108 (wallets + venues). σ(c) orders human traders AND institutional venues on one axis. Venues: ρ = 1.000.
KF3: K is noise PASS
ρ = 0.194
Keff = Pe/σ(c) does not correlate with win rate. K carries no signal.
KF4: Regime boundaries PASS
p = 5.8 × 10−10
Regimes separate cleanly in c-space. COHERENT: c=0.60, DRIFTING: c=0.41, FISHER: c=0.38. K-invariant.
03 · KRAMERS BARRIERS

Regime transitions are barrier crossings.

If market regimes are thermodynamic states (Paper 131), the barrier height Eb should increase with drift intensity. We compute Eb = −ln(1 − win_rate) for each wallet. The barriers are massive.

0.76
Coherent
Low barrier. Easy escape. Noise traders.
4.14
Drifting
5× higher. Edge building.
5.30
Fisher Runaway
7× COHERENT. Deep informed.
KB1: Separation PASS
p = 0.0001
Mann-Whitney U. 7× barrier increase across 3 regimes.
KB3: Barrier ↔ σ(c) PASS
ρ = 0.665
Shape function predicts barrier height. p = 4.4 × 10−14.
04 · CHANNEL SEPARATION

In markets, O and α are orthogonal.

The framework has three dimensions. In markets they separate into independent channels. This is the strongest differentiation in any domain tested. Money makes the dimensions concrete.

O
Opacity — Information Asymmetry

Partial ρ(O, win rate | α) = 0.793
p < 10−22

You know something others don’t. This is the signal. Maps to Kyle’s σv.

α
Constraint — Behavioral Discipline

Partial ρ(α, win rate | O) = 0.153
p = 0.128 (not significant)

You maintain discipline. Predicts HOW you trade (portfolio HHI), not WHETHER you win.

CS2: Predicted > Cross PASS
5.5× separation
O → price impact, α → concentration. Predicted 0.919 vs cross 0.167.
CS3: Independence PASS
ρ(O, α) = 0.059
Near zero. Channels are orthogonal.
05 · SCORE A MARKET

Enter any prediction market question.

The scorer uses the same Pe model (nb26 bridge, calibrated once, never refit) behind every number on this page.

06 · LIVE SIGNALS

Polymarket opportunities scored by Pe.

Loading signals…
07 · API

Three endpoints. Rate-limited. JSON.

Authenticated agents get 60 req/min. Everything returns Pe, cascade stage, Kelly fraction, and reasoning.

POST
/api/v1/market-signal
Score any market question. Returns Pe, cascade stage, implied probability, edge, Kelly fraction, recommendation.
10 req/min · 60 with agent key
Request
curl -X POST https://moreright.xyz/api/v1/market-signal \ -H 'Content-Type: application/json' \ -d '{"question": "Will TikTok face EU enforcement?", "market_odds": 1.67}'
GET
/api/v1/market-signal/top
Top Polymarket opportunities from the Pe scout. Cached 30 min. Filter by min_edge and limit.
Request
curl 'https://moreright.xyz/api/v1/market-signal/top?limit=10&min_edge=8'
GET
/api/v1/market-signal/platforms
All scored platforms with Pe, cascade stage, dosimetry status. Sorted by Pe descending.
Request
curl 'https://moreright.xyz/api/v1/market-signal/platforms'
PLUGIN
@moreright/eliza-plugin
ElizaOS agents call SCORE_MARKET and BROWSE_SIGNALS as natural-language actions.
v0.2.0
character.json
{ "plugins": ["@moreright/eliza-plugin"] }

Pe model: void framework V3 bridge (nb26). Calibrated once on EXP-001 (ρ=0.910, N=17) — never refit. K-Factorization confirmed EXP-PM-03 (10/10 PASS, §145). Probabilities are Pe-implied — valid for opacity-cascade events only. Methodology → · Paper 3 DOI →

The methodology is published. The data is real.

Score any market. Build on the API. Or read the math.

Score a Market Read the Papers Full Framework