The market for the world’s largest asset class

We’re building the market for an asset class that already exists.

People already invest in other people. Income Share Agreements at Purdue and Lambda. Athlete-revenue advances at Big League Advance. Music-catalog trading at Hipgnosis, Royalty Exchange, SongVest. Litigation finance. Life settlements. Revenue-share deals on private cash flows. The asset exists.Humans and their future economic output. What’s missing is a liquid, standardized, clearable market for it. Preflop is that market.

Jump to pricing math →Read the spec →

Why this isn’t science fiction

Every row below is a market that clears today. None of them compose. None of them share a pricing rail. None of them let a backer allocate across categories. Preflop is the rail.

Analog market	Est. annual volume	What’s missing
Income Share Agreements (Purdue, Lambda, etc.)	~$300M+	Priced at a single handshake moment. No secondary. No repricing. Illiquid through maturity.
Athlete revenue advances (Big League Advance)	$300M+ deployed	Bilateral contracts. No exchange, no order book, no standard covenant. One-to-one.
Music-catalog trading (Hipgnosis, Royalty Exchange, SongVest)	Multi-$B	Mature instrument for one asset class. Doesn't compose with ISAs, athlete claims, or equity.
Royalty Exchange (auction platform)	~$100M+ cumulative	Auctions, not continuous markets. No secondary repricing. No cross-category allocation.
Life settlement market	~$4.5B	Illiquid, OTC, age- and health-gated. No standard pricing rail. Not accessible to non-specialists.
Litigation finance	~$16B AUM	Institutional-only. Case-by-case. No fungible instrument, no continuous pricing.

Every instrument in that table trades today. None of them compose, none of them share a pricing rail, and none of them let a backer allocate across categories. Preflop is the rail.

The three primitives

01TEB Definition

Total Economic Benefit, the comprehensive denominator. W-2 wages, 1099 self-employment, K-1 distributions, equity realizations, royalties, board fees, non-cash consideration at FMV. Closes every door an issuer could use to route around the covenant.

Read the full spec — TEB Definition →

02Obligation Ledger

Every listing is a (share_rate, time_window) tuple on an issuer's future TEB. The ledger tracks every active claim and enforces a single invariant: the sum at any moment can't exceed 25% of TEB. Two independent cap-check algorithms cross-validate. 75 passing tests guard the arithmetic.

Read the full spec — Cross-Listing Math →

03Conviction Engine

Eleven signals (six backward-looking, five forward-looking) blended 0.6/0.4. Today: curates who qualifies to list. Tomorrow: feeds the forecast that feeds the pricing model. The data accrues with every listing. The engine improves with every realized TEB outcome.

Read the full spec — Conviction v2 →

Pricing · the four stages

A liquid instrument requires three things: a fundamental anchor, a market-clearing price, and a continuous repricing loop. Preflop builds all three. Every token prices through four stages, each with a governing formula and a worked number that flows into the next.

1

Stage 1 — Forecast

Evidence → forward TEB path

We take every piece of evidence on the issuer (tax filings, trajectory, sector, conviction signals) and produce a forward path of their Total Economic Benefit with low / mid / high bands. The path is piecewise exponential: a blended near-term CAGR, then a terminal growth rate beyond a 10-year horizon.

\widehat{\text{TEB}}(t) \;=\; \begin{cases} \text{TEB}_0\,(1+g_1)^{t} & 0 \le t \le 10 \\ \widehat{\text{TEB}}(10)\,(1+g_2)^{t-10} & t > 10 \end{cases} \qquad g_2 = 2\%

Worked: mid-band TEB today $= \$500{,}000$ , blended $g_1 = 8\%$ for 10 years, then $g_2 = 2\%$ terminal. At year 10, $\widehat{\text{TEB}}(10) \approx \$1.08\text{M}$ ; the forecast extends to year 75.

Why it’s robust —The forecast isn’t one number. It’s a band. Pricing uses the mid and carries the confidence interval through to the reference price, so a reader can distinguish high-conviction narrow-band issuers from noisy wide-band issuers.

2

Stage 2 — Reference price

Gordon-growth anchor for the first trade

The forecast feeds a Gordon-growth present-value calculation. At listing, the reference price $P_{\text{ref}}$ is the PV of the issuer’s committed share of future TEB, divided by the fixed token supply $N = 10{,}000$ .

P_{\text{ref}} \;=\; \frac{1}{N} \cdot \frac{e \cdot \text{TEB}}{r - g}

Worked: $e = 2\%$ , $r = 12\%$ baseline discount rate, $g = 5\%$ baseline Gordon growth, $\text{TEB} = \$500\text{K}$ . Numerator: $0.02 \times \$500{,}000 = \$10{,}000$ . Denominator: $0.12 - 0.05 = 0.07$ . PV of perpetual claim: $\$10{,}000 \,/\, 0.07 \approx \$142{,}857$ . Divide by $N = 10{,}000$ tokens: $P_{\text{ref}} = \$14.29$ with confidence interval ±30%.

Why it’s robust —The reference isn’t the price. It’s the anchor the book is built around. Analyst-facing, transparent, and derived from the same public forecast every backer can inspect.

Roadmap —The discount rate $r$ is currently a platform constant (12%) in engine v2.1.0; cohort-conditional $r_c$ (with a survival-curve auxiliary on $V_{\text{HC}}$ ) is specified in design v1 at /spec/discount-rate Design v1 — In progress and ships in engine v2.2.0. The cohort priors that feed V_HC are also being re-anchored to entry-pool data sources (Census ABS, BLS, AAMC, NCAA, IRS) with a survival-mixture $P_{\text{stay}}(t)$ added to the integration — design v2 at /spec/cohort-priors Design v2 — In progress, ships in engine v2.3.0 alongside the discount-rate change.

3

Stage 3 — First trade

Anchor meets order book

Lockup ends. Market makers post two-sided quotes seeded from $P_{\text{ref}}$ . Circuit breakers halt trading on moves of more than ±20% from the reference until conviction-signal review clears the halt. The order book opens and the market clears. The reference is an anchor, not a price. It’s the fundamental against which order flow deviates.

Why it’s robust —Every venue for human-capital claims today either anchors on a single handshake-negotiated price (ISAs, BLA) or has no repricing mechanism (private cap-table secondaries). Continuous repricing against a public Forecast Engine is the structural break from ISAs as a category. The contract stops being a one-time loan and becomes a continuously priced asset. A pre-trade anchor plus an open order book is how equities, bonds, and commodities have cleared for a century.

4

Stage 4 — Continuous repricing

Every event updates the fundamental

Every quarterly TEB report, every material event (single-source income ≥ $100K, notifiable within 7 days), and every conviction-signal update reshapes the forecast. The forecast re-anchors the fundamental. The market clears the delta. Between events, secondary prices follow a geometric Brownian motion with drift set by conviction and volatility set by phase:

\frac{dS}{S} \;=\; \mu\,dt \;+\; \sigma\,dW_t, \qquad \mu = f(\text{conviction}), \quad \sigma_{\text{Phase 2}} < \sigma_{\text{Phase 1}}

Information events and their repricing direction

Event	Affects	Direction
Quarterly TEB report	Forecast mid	Re-anchors reference price
Material event ≥ $100K (notifiable within 7 days)	Forecast mid + CI	Anchor update + band narrowing
Conviction-signal update	Drift μ	Continuous, integrates over t
Sector regime shift	Terminal g	Long-tail repricing
New obligation on same issuer (cap-check passed)	Cross-claim α	Class-specific, not systemic
Missed report → deficiency notice	Liquidity premium π	Discount widens until cure
Verified extreme-positive income event	Band high + mid	Upside revaluation

Why it’s robust —The risk-adjusted rate $r$ is itself decomposed. Math-foundation Part 3.6 framing: $r = r_f + \pi_{\text{income}} + \pi_{\text{liquidity}} + \pi_{\text{moral hazard}}$ . Live engine v2.1.0 collapses this to $r = 12\%$ flat; Discount Rate v1 Design v1 — In progress replaces it with three cohort-conditional components $r_c = r_f + \beta_c \cdot \text{ERP} + \pi_{\text{illiq}}$ and ships in engine v2.2.0.

This is the mechanism nobody else has built. Existing venues for human-capital claims either anchor on a single handshake-negotiated price (ISAs, BLA) or have no repricing mechanism at all (private cap-table secondaries). Continuous repricing is the structural break from ISAs as a category. Preflop has a fundamental anchor, a market-clearing price, and a continuous repricing loop, the three things a liquid instrument requires.

Read the full spec — Direct Listing pricing→

The Obligation Ledger · cross-listing math

Every token (Covenant Phase 1, Covenant Phase 2, Direct Listing, or a second Direct Listing on the same person years later) is a slice of the issuer’s future TEB, defined by a share rate and a time window. The same pricing machinery handles all of them, and a single invariant keeps the stack coherent.

The unifying abstraction

\text{Token claim} \;=\; \left(\tfrac{\alpha}{N}\right)\,s_{\text{rate}}\,\text{TEB}(t),\qquad t \in [t_{\text{start}},\,t_{\text{end}})

$\alpha$ = fractional share of TEB the class claims. $s_{\text{rate}}$ = payout rate within the window. $N = 10{,}000$ tokens per class. Windows are half-open so back-to-back listings can’t double-count the boundary instant.

The 25% invariant

\sum_{i}\; s_{\text{rate},\,i}\,\mathbf{1}\!\left[\,t_{\text{start},\,i} \le t < t_{\text{end},\,i}\,\right] \;\le\; C_{\max} = 25\%

A human can list multiple claims on their future TEB simultaneously — a Covenant Phase 1, the linked Phase 2, a Direct Listing, a second Direct Listing years later. Without this ledger there is no way to guarantee the sum of those claims stays coherent and the issuer retains at least 75% of their TEB to live, pay taxes, save, and invest. We enforce it algorithmically, cross-validated by two independent cap-check algorithms: an analytic transition-point scan and a monthly-bucket scan. Delisted grace windows and content hashes preserve an audit trail.

Every class is a tuple

Class	share_rate	t_start	t_end
Covenant Phase 1 dividend stream	s	covenant start	year T
Covenant Phase 2 perpetual claim	e_cov	year T	∞
Direct Listing perpetual claim	e_dir	listing day	∞
Second DL on same person	e_dir,2	listing day 2	∞

Pricing integral + Gordon collapse

P(t) \;=\; \left(\tfrac{\alpha}{N}\right)\int_{\max(t,\,t_{\text{start}})}^{t_{\text{end}}} s_{\text{rate}}\,\mathbb{E}[\text{TEB}(\tau)]\,e^{-r(\tau-t)}\,d\tau

In the pure-perpetuity case (Direct Listing or post-T covenant), this collapses to the Gordon-style anchor from Stage 2: $P = \tfrac{\alpha}{N}\,\tfrac{s_{\text{rate}}\,\overline{\text{TEB}}}{r - g}$ .

Worked example · covenant adds a Direct Listing

Issuer Z is year 3 of a s = 5%, T = 7, e_cov = 2%covenant (already IPO’d, tokens trading). Z applies for a Direct Listing at e_dir = 3%.

Step 1 · cap check (analytic scan + monthly bucket, both must pass)

Years 3–7: 5% + 3% = 8% ✓ under 25%

Years 7+: 2% + 3% = 5% ✓ under 25%

Step 2 · mint 10,000 Direct Listing tokens at $P_{\text{ref,dir}} \approx \$30/\text{token}$

Step 3 · existing covenant tokens unaffected — their claim is unchanged

Step 4 · two markets on Z — covenant tokens (dual-character, phase-decaying) and Direct Listing tokens (pure perpetual)

Step 5 · at year 7, covenant tokens transition to pure Phase 2 — economically equivalent to the Direct Listing class

Tokens on the same person from different paths trade separately but consistently. Think Class A / Class B shares.

75 passing tests guard every number in this section. Every price you see on the site is produced by the same library that will price real listings.

Read the full spec — Cross-Listing Math→

The Conviction Engine

Phase 1 covenant conviction is forward-looking by necessity. there’s no past trajectory on the platform yet. Direct Listings ship 5+ years of actual data. So v2 is hybrid: 60% weight on backward-looking signals (what happened), 40% on forward-looking signals (what’s next).

\text{Conviction}_{v2} \;=\; 0.6\,\text{Backward} \;+\; 0.4\,\text{Forward}

Backward · 60% · 6 signals

—TEB CAGR · trailing 5-year, recent 3 years weighted 2×
—TEB volatility · std-dev / mean of annual TEB
—Income diversification · Herfindahl–Hirschman concentration of income streams
—Event-driven income capture · monetization of IPOs, acquisitions, book deals, contract renewals
—Tax compliance record · on-time filing, no audits, no penalties
—Professional continuity · gaps / pivots / sustained engagement

Forward · 40% · 5 signals

—Industry trajectory · sector growing / flat / declining
—Age-adjusted runway · earning years remaining given age + role
—Pipeline indicators · signed contracts, vesting schedules, known future events
—Brand / network momentum · audience growth, citation velocity, reputation
—Stated intent · what the issuer says they’ll do next (weighted low)

The flywheel

The conviction engine does two jobs. Today: curation. Gating who qualifies to list. Tomorrow: anchoring.refining the forecast that feeds the pricing model. Every listing that pays out against its forecast is a training point. Every listing that underperforms is a training point.

In 18 months, Preflop owns the only dataset in the world that correlates early-career signals with realized human economic output. That dataset is the moat. It is not the app; it is the corpus. We cannot be overtaken by a better UI.

What’s rigorous, what’s judgmental

The ledger math and pricing formulas are rigorous and tested. The 11 conviction weights are seeded by economic judgment and will be recalibrated quarterly against realized TEB outcomes. The calibration harness is in the codebase today, logging empty until the first outcome lands. This is how every valuation engine in financial history has bootstrapped. The difference is Preflop architects the logging from day one.

Read the full spec — Conviction v2→

The market · three-tier architecture

Each tier unlocks the next. Primary without secondary is a handshake deal. Secondary without derivatives is a scratch pad. The three together are a market.

01Tier · Primary

Issuance

Covenants, Direct Listings. Issuer commits a (share-rate, window) slice of TEB. Mints 10,000 tokens at the reference price. Primary market participants get first allocation at the published reference.

02Tier · Secondary

Trading

Continuous limit order book. Tokens trade against the evolving fundamental. GBM drift set by conviction; volatility set by phase. Every Preflop token is fungible on the same terminal, regardless of origination path.

03Tier · Derivatives

Composite exposure

Cap-weighted indices (e.g. IDX — a flagship issuer index). European-style options on individual tokens or sectors. Tranched products for diversified exposure. All composed from the same primitives — no re-architecture.

Robinhood routed retail access to markets that already existed. Preflop is the market. For an asset class that existed but never traded on a common rail. The prize is not a better broker. The prize is owning the rail.

How a Preflop raise actually clears

The pricing engine produces a number. The auction is what translates that number into a clearing price the market actually agrees to. Two concepts make every raise readable: the κ ratio (how the issuer’s ask compares to the engine’s forecast) and the κ-tier regime (which auction posture the issuer is allowed to take given that ratio).

The κ ratio in one line

\kappa \;=\; \frac{\hat V_{\text{HC}}}{V_{\text{HC}}^{\text{engine, mid}}} \;=\; \frac{P_{\text{target}} / e_{\text{eff}}}{V_{\text{HC}}^{\text{engine, mid}}}

$\kappa = 1$ means the issuer’s implied self-valuation matches the engine’s mid forecast. $\kappa > 1$ means they’re asking the market to price them above mid, asserting their forward trajectory is better than the backward signal currently shows. The platform requires that the κ value sit in a tier consistent with the issuer’s conviction.

Tier	κ range	Conviction floor	Posture
Anchored	≤ 1.2	Any (curation gate)	Standard auction; engine-mid is the credible anchor.
Modest premium	1.2 – 2.0	≥ 65	Standard; bidder disclosure flags κ.
Elevated	2.0 – 3.0	≥ 75	Forecast-divergence rationale required; reserve floor moves to 0.85 of low-band.
Speculative	3.0 – 5.0	≥ 85	Flagged listing; smaller per-bidder cap; tighter clearing threshold.
Market-discovery	> 5.0	Not eligible	Listing rejected; raise forecast or conviction first.

The auction in one paragraph

Sealed-bid, uniform-price Dutch auction with per-bidder allocation cap (20% of supply), engine-anchored reserve (80% of low-band MCap divided by N), AON partial-clearing with retained allocation if undersubscribed but above the 60% minimum-clear threshold, and threshold-encrypted on-chain settlement so no front-runner can race the clearing block. Every winning bidder pays the same clearing price $P^*$ , regardless of their original bid. Full mechanism in the raise-mechanism spec.

Maya’s clearing in numbers

Pre-indication disclosure (T-4 weeks)

$V_{\text{HC}}^{\text{mid}} = \$3.01\text{M}$ , $V_{\text{HC}}^{\text{low}} = \$1.38\text{M}$ , $e_{\text{eff}} = 1.66\%$ , conviction 62, CI ±54%. Reserve published: $P_{\text{reserve}} = (0.80 \times \$1.38\text{M} \times 0.0166)/10{,}000 = \$1.84/\text{tok}$ .

Tier check

Maya targets $\$75\text{K}$ : $\hat V_{\text{HC}} = \$75\text{K}/0.0166 = \$4.5\text{M}$ , $\kappa = 1.50$ — modest-premium tier requires conviction ≥ 65. Maya is 62. Platform forces target reduction to $\$60\text{K}$ — anchored-tier boundary, $\kappa = 1.20$ . Maya accepts.

Clearing

Bid book runs from $\$8.50$ down to $\$1.84$ reserve. Cap-adjusted demand at $\$7$ meets supply. $P^* = \$7.00$ , full 10,000 tokens cleared. Total raise: $\$70{,}000$ , 17% above adjusted target. $\kappa_{\text{realized}} = \$70\text{K}/0.0166/\$3.01\text{M} = 1.40$ .

Engine cross-reference (v2.1.0)

The walkthrough above runs on the math-foundation forecast (TEB(10) = $600K). Run Maya’s exact inputs through the v2.1 Forecast Engine — cohort priors + bounded adjustments, shrinkage $w_{\text{personal}} = 0$ at $N = 0$ quarters — and the engine produces $V_{\text{HC}}^{\text{mid}} \approx \$860\text{K}$ , $e_{\text{eff}} \approx 1.78\%$ . At a $\$60\text{K}$ target raise: $\hat V_{\text{HC}} = \$60\text{K}/0.0178 = \$3.37\text{M}$ , $\kappa = 3.37/0.86 \approx 3.92$ → speculative tier (math-foundation says anchored).

The cohort prior says the median pre-seed issuer in this cohort doesn’t reach the math doc’s aggressive trajectory. Engine output is the operational ground truth; math-foundation reconciliation queued for v2.2. Read the full Maya walkthrough at /spec/forecast-engine §7.

Why it’s robust —Every Preflop raise produces a clearing price that is reproducible from the bid book and the engine’s published methodology. Nothing is whatever- cleared.The issuer’s ask is checked against the κ regime, the reserve is derived from public inputs, the algorithm is transparent, and the settlement is on-chain.

The V_HC used in every κ calculation comes from the v2.1 Forecast Engine — cohort baselines + Bayesian shrinkage + bounded adjustments. Engine output is the ground truth for every reserve, every reference price, every clearing.

Read — Forecast Engine v2.1→

Read — Human capital, e_eff, κ→

Read — full auction mechanism→

Where the returns actually come from

Every Preflop token represents two bundled claims on the same issuer: a perpetual claim on lifetime TEB (Phase 2, the dominant return component), and a time-bounded dividend stream during the covenant (Phase 1, the floor that pays while the perpetuity converges). The token’s fair value at any time is the present value of both claims together. Whether a backer’s return comes mostly from capital appreciation on the perpetuity or from accumulated dividends depends entirely on where the issuer sits on their career arc at listing.

The decomposition

V(t) \;=\; \underbrace{\int_{T}^{\infty} \frac{s_2 \cdot \mathbb{E}[\text{TEB}(\tau)]}{N}\,e^{-r(\tau-t)}\,d\tau}_{\text{Phase 2 — perpetual claim (primary)}} \;+\; \underbrace{\int_{t}^{T} \frac{s_1 \cdot \mathbb{E}[\text{TEB}(\tau)]}{N}\,e^{-r(\tau-t)}\,d\tau}_{\text{Phase 1 — bounded dividend (floor)}}

Phase 2 is the perpetuity, the load-bearing return component. Phase 1 is a finite 10-year dividend that pays a floor while the perpetuity converges. Same issuer, same TEB path, two very different sensitivities to changes in that path.

Why Phase 2 leverages trajectory re-forecasts

In the Gordon-collapse form, the Phase 2 present value at time $t$ with the covenant ending at year $T$ reduces to:

PV_2(t) \;=\; \frac{s_2 \cdot \text{TEB}(T) \cdot e^{-r(T-t)}}{(r - g_{\text{terminal}}) \cdot N}

The perpetuity value scales as $1 / (r - g_{\text{terminal}})$ . When the market re-forecasts $g_{\text{terminal}}$ from 2% to 4% (modest re-rating after an early breakout), the denominator halves and $PV_2$ roughly doubles. Phase 1 has no such leverage. It’s a bounded cash flow whose reprice is proportional to the TEB path, not amplified by perpetuity math. This asymmetry is structural, not speculative. It falls straight out of the closed-form integral.

Two personas, same instrument

Persona	Phase 1 PV (listing)	Phase 2 PV (listing)	Dominant component	Return profile
Maya · 20, pre-revenue founder, cold-start cohort prior	~$10K total / ~$1.00/tok	~$5.3K total / ~$0.53/tok	Optionality on Phase 2 re-forecast	Venture
Dr. Amara Okafor · 45, orthopedic surgeon, Direct Listing (~$2M/yr TEB)	~$340K total / ~$34/tok	~$160K total / ~$16/tok	Dividend cash flow	Dividend-weighted · modest optionality

One instrument, two investor personas, both coherent.Pick your trajectory exposure.

The Maya worked example: three snapshots

Maya lists at age 20, pre-revenue. Covenant $s_1 = 3\%$ , $T = 10$ years, $s_2 = 1\%$ perpetual. $r = 12\%$ , $N = 10{,}000$ tokens. Snapshots show two coherent regimes — engine-derived (the live operational number from v2.1 Forecast Engine) and math-foundation pedagogical (the canonical piecewise forecast TEB(0) → TEB(10) = $20K → $600K, used to illustrate the year-4 breakout reprice mechanics).

Both narratives reference the same lib/obligation-ledger pricing library. The engine numbers are operational ground truth. The math-foundation numbers will reconcile to engine output once Part 9 is regenerated against the v2.1 cohort priors.

T = 0Listing · v2.1 engine output

$V_{\text{HC}}^{\text{mid}} \approx \$860\text{K}$ , $V_{\text{HC}}^{\text{low}} \approx \$185\text{K}$ , $V_{\text{HC}}^{\text{high}} \approx \$2.53\text{M}$ . CI band ±273% — log-normal cohort dispersion at $N = 0$ quarters of personal data (cohort dominates 100%). $e_{\text{eff}} \approx 1.78\%$ .

PV_1(0) \approx \$1.00/\text{tok}\;\;(66\%)

PV_2(0) \approx \$0.53/\text{tok}\;\;(34\%)

V(0) = P_{\text{ref}} = \$1.53/\text{tok}\;\;(\text{engine-mid})

At $\$60\text{K}$ target raise: $\hat V_{\text{HC}} = \$3.37\text{M}$ , $\kappa \approx 3.92$ → speculative tier. Engine more conservative than the math-foundation pedagogical walkthrough — cohort prior says the median pre-seed issuer in this cohort doesn’t reach $\text{TEB}(10) = \$600\text{K}$ .

Math-foundation reference — $V_{\text{HC}}^{\text{mid}} = \$3.01\text{M}$ , $P_{\text{ref}} = \$5.00/\text{tok}$ , $\kappa = 1.20$ , anchored tier. Used in the auction walkthrough to show clearing mechanics; reconciliation queued for v2.2.

T = 4Breakout · Series A closes

Forecast re-weights: TEB(4) = $400K, reaching $3M by year 10. Terminal $g$ re-rates to 4%. $V_{\text{HC}}(4) = \$23.32\text{M}$ (7.75× from listing).

PV_1^{\text{remaining}}(4) \approx \$15.19/\text{tok}

PV_2(4) \approx \$18.25/\text{tok}

V(4) = \$33.45/\text{tok}\;\;(\sim 6.7\times)

Of the $\$28.45/\text{tok}$ dollar reprice from listing, 57% came from Phase 2 ( $\$16.24$ of the $\$28.45$ ), 43% from Phase 1. Phase 2 leveraged the terminal-growth re-rate; Phase 1 leveraged the near-term TEB lift.

T = 10Phase-2 transition

Covenant Phase 1 ends. Token converts to pure Phase 2 perpetuity. Maya at 30, TEB(10) = $2.5M, terminal $g = 3.5\%$ .

PV_2(10) = \frac{\$2.5\text{M}\cdot 0.01}{0.12 - 0.035} \approx \$294\text{K}

V(10) \approx \$29.41/\text{tok}

Original auction-clearing buy at $\$7$ is worth $\$29.41$ at year-10 spot (~4.2×) plus accumulated Phase 1 dividends collected over the 10-year covenant. Capital gain bounded by $(r - g_{\text{term}})$ compression — the upside envelope is whether $g_{\text{term}}$ itself re-rates further.

57%

Phase 2 · perpetuity

43%

Phase 1 · dividend

Phase 2, the perpetual claim, is the dominant return component. The dividend is the floor that pays you while the perpetuity converges.

From Maya’s breakout snapshot, year 4 reprice

The truly asymmetric outcomes are cases where $g_{\text{terminal}}$ itself re-rates upward — Maya becomes a category-defining founder and her career growth looks like 6–8% indefinitely. The upside envelope is set by how high the market is willing to price $g_{\text{terminal}}$ .

For breakout humans, that’s where the capital-markets re-rating lives.

The covenant is the income. The token is the trajectory. Dividends give you a floor; the Phase 2 perpetual gives you the call option on who this person becomes.

Read the full spec — Pricing decomposition→

Liquidity when the trajectory arrives, not when the clock runs out

Here’s the hidden elegance of the instrument: the covenant (the time-bounded obligation to pay a share of TEB) and the lock(whether tokens can transfer on the secondary market) are two different things. Most ISAs glue them together. You wait ten years for both the dividend stream and the liquidity. Preflop doesn’t. A token can unlock early once the issuer’s trajectory warrants it, while the underlying covenant continues paying dividends for the full original term.

Prior ISAs vs Preflop

Prior ISAs: single handshake price, capped at 2–3× principal, sold upstream to a hedge fund, illiquid until the contract expires. Preflop: continuous market-clearing price, uncapped Phase 1 dividend that converts into a perpetuity, no secondary kickback to the platform, token tradable as soon as the trajectory warrants it.

Read the full structural comparison →

The three unlock paths

01Term expiration

Automatic at the end of the covenant term $T$ . Always available. Tokens convert from the dual-character instrument into a pure Phase 2 perpetuity and secondary markets open without any further review.

02Issuer-initiated early unlock

The issuer requests early unlock; Preflop reviews against a formal data + process checklist (below). If thresholds are met and disclosures are clean, the market opens ahead of term expiration. The covenant continues running in the background.

03Event-triggered consideration

Major observable liquidity events (an exit, an acquisition, a major funding milestone with independent verification) can trigger unlock consideration. Discretionary — Preflop’s review panel still gates. Never automatic.

The issuer-initiated process · seven gates

1Minimum data threshold. 8 quarters of auditable TEB reports for covenant issuers (two years on-platform). 4 quartersfor Direct Listings, which arrived with 5+ years of pre-listing history already. Without this, the reference price doesn’t have enough ground-truth to anchor defensibly.
2Reference price modeling. Preflop’s pricing team runs the current forecast through the Obligation Ledger, produces a refreshed $V(t)$ with updated CI band. Same pipeline, same invariants as a Direct Listing day-one anchor.
3Disclosure pack. Refreshed issuer disclosures, updated conviction score breakdown, and a current cap-ledger summary showing every active obligation under the 25% invariant.
4Market-maker engagement. A market maker contracts in for the first 90 days post-unlock and provides two-sided quotes on the book so early liquidity isn’t a thin-market guess.
5Tight circuit breakers on day one. A ±20% deviation from reference halts trading pending conviction-signal review. Bands loosen over the first 30 days as realized volatility establishes itself.
6Insider lockup — 18 months post-unlock on the issuer’s retained allocation. The issuer cannot sell their own tokens for 18 months after the market opens. This removes the obvious gaming incentive (open the market right before a good quarter to dump your own bag) while keeping the lockup from feeling permanent — standard post-IPO insider-lockup convention, applied to the unlock event.
7Covenant rate is inviolate. The issuer signed for $s_1 = 3\%$ over 10 years. Early unlock does not renegotiate the share rate, term, equity conversion, or TEB definition. The covenant is the covenant — this is the credibility hinge of the whole instrument.

Maya early-unlock walkthrough

Year 2.Maya’s Series A closes and her fair value reprices substantially; tokens are still locked, so the reprice shows up on /market as an updated $V(t)$ but nobody can actually trade it yet — the data threshold hasn’t been met. Year 4. Maya requests early unlock. Her data threshold — 16 quarters of reported TEB, well above the 8-quarter minimum — is clean. Preflop validates the reference price (now $\approx \$32/\text{tok}$ vs. $\$50$ at listing), commits to it with a ±20% CI band, contracts a market maker in for 90 days, and opens the book. Some backers sell — early believers realizing the gain. Others hold, preserving their claim on the remaining 6 years of Phase 1 dividends plusthe perpetual Phase 2 token. Maya continues paying 3% of TEB for years 4–10 exactly as covenanted. Her own retained issuer-allocation tokens are locked until month 22 (year 4 + 18 months). Everybody’s incentives remain aligned.

Why this matters

For issuers

No more ten-year wait to validate the instrument. The issuers who break out early get visibility, and their successful early-unlock events become the proof points that recruit the next cohort.

For backers

Risk capital isn’t illiquid for a decade. Liquidity appears when the thesis plays out — without forcing every backer to become a ten-year lockup holder at the point of investment.

For the market

Breakout names become public benchmarks within 2–4 years, not 10. Every early-unlock print generates comps that re-anchor the pricing of every subsequent listing. The price-discovery flywheel compounds faster.

Read the full spec — Early unlock mechanics→

What’s proven, what’s heuristic

Honest labelling is the credibility move. Everything on the left is backed by tests and math. Everything on the right is seeded by economic judgment and will harden as data accrues.

Rigorous today

✓Obligation Ledger arithmetic — 75 passing tests, two cross-validated cap algorithms
✓Pricing integral and Gordon collapse — closed-form, auditable, tested
✓Cap invariant (≤ 25% of TEB) enforced at every listing application
✓TEB definition — comprehensive, unambiguous, closes routing-around attacks
✓Risk-adjusted rate decomposition — every premium observable per issuer
✓Cap theorem (flow ⟹ stock) — Fubini-proven; the issuer always retains ≥ 75% of their own trajectory
✓ $e_{\text{eff}} = e$ exact for Direct Listings; closed-form for canonical covenant; piecewise via numerical integration in production
✓Auction clearing algorithm — sealed-bid uniform-price with allocation cap; reproducible from bid book

Judgmental today · rigorous tomorrow

◯Conviction weights — seeded by economic judgment, recalibrated quarterly against realized TEB
◯Forecast bands — mid/low/high derived from signals; CI tightens as issuer-history accrues
◯Terminal growth assumption — 2% placeholder; will become sector-conditional
◯Circuit-breaker thresholds — ±20% seed; will calibrate against realized volatility
◯Phase 2 drift/volatility model — GBM baseline; upgrades to regime-switching as data lands
◯κ-tier thresholds (1.2 / 2.0 / 3.0 / 5.0) — governance-set; will recalibrate against realized cohort performance
◯Reserve coefficient (0.80 of low-band V_HC), minimum-clear (60% of N), allocation cap (20%) — disclosed governance choices subject to revision

Read the spec

Every claim on this page has a rigorous document backing it, one click away.

Instrument

TEB Definition

The comprehensive denominator. Inclusion rules, exclusion rules, attestation channels. Every instrument prices against this quantity.

Open spec →

Instrument

Two-Phase Uncapped Covenant

Phase 2 perpetuity (the dominant return) + Phase 1 income-linked dividend (the floor that pays you while it converges). Four parameters, five holder rights, power-law reasoning for no-cap.

Open spec →

Instrument

Direct Listing

Perpetual claim from day one. Eligibility, review, lockup, cross-listing math with the Covenant stack.

Open spec →

Math

Direct Listing Pricing

Forecast → Reference price → First trade → Continuous repricing. Gordon anchor, GBM drift, information-event repricing. End-to-end worked.

Open spec →

Math

Cross-Listing Math

Two cap-check algorithms, 75 passing tests, five worked examples. The math that makes multi-listing composable and safe.

Open spec →

Math

Conviction Engine v2

Eleven signals, 0.6 / 0.4 hybrid. Today: curation. Tomorrow: anchoring. The flywheel, the calibration harness, and the data moat.

Open spec →

Lifecycle

Pre-IPO Lock & IPO Unlock

Three thresholds that open the book — 4+ income reports, score ≥ 60, funding ≥ 50%. Why each exists, what holders can and can't do during lock.

Open spec →

On-chain

Smart Contract Hardening

Contract surface, six on-chain invariants, access control, immutability posture, audit roadmap. How the Solidity holds up the math.

Open spec →