# UNI ownership valuation: September 19, 2026

## Scope

This edition values outright ownership of UNI. Its process is to identify the scope, trace the economic benefits, infer a value, validate the inference and report separate personal and institutional ownership decisions. The market-based valuation concerns UNI. Separate clearly labeled worked examples explain LP, swap, execution, company equity, governance and builder decisions. The builder calculator values an assumed business, not a named operating company. None of those hypothetical returns is added to UNI value. Economic benefits from every product are eligible if a connection to UNI is established. The connected market pages supply separate dated technical, futures and activity analysis. Their signals and prices do not automatically change this frozen valuation.

The fee cutoff is September 19, 2026 at 00:00 UTC, so September 18 is the final complete observation day. The market snapshots are from September 19, at the exact timestamps in evidence.json. The ownership horizon is three years, ending September 19, 2029. The explicit forecast has five annual periods, ending September 19, 2031. The estimate is a present value, not a 2029 price target. Planned review dates are analytical checkpoints, not scheduled jobs.

## Trace

CoinGecko supplies USD price, unburned total supply, circulating supply and update time. The total-supply denominator includes treasury holdings. It does not use the original one-billion launch supply or a permanently fixed maximum: UNI governance can authorize annual minting.

DefiLlama supplies daily total fees and daily holders revenue through its fee-summary API. Complete 7-, 30-, 90- and 365-day windows are recomputed from dated daily rows. All rows must be present exactly once. Chain breakdown sums are reconciled to the daily-total sum. The JSON field chain_reconciliation_gap is daily total minus chain total; the signed gaps described below use the opposite convention, chain total minus daily total. For UNI holder revenue only, the 365-day chain sum minus daily sum is −$3.23; 7-, 30- and 90-day gaps are zero. CAKE holder revenue has a +$7.79 gap over 90 days; total UNI fees have a −$11.60 gap. These small rounding/revision differences do not materially affect this model. The normalized headline uses the 90-day sum multiplied by 365/90. This annualization is a starting assumption, not an audited measure of sustainable earnings.

The UNI holders-revenue adapter counts Firepit Released events, multiplies by its threshold and values the inferred UNI token amount in USD. It is a release-based burn-value proxy. It does not directly sum same-day raw fee-asset receipts. Historical thresholds, release timing and UNI pricing affect measurement. This price dependence introduces circularity: forecasting USD burn value cannot independently establish operating revenue growth. All “fee base” and “fee growth” controls refer to this proxy, pending receipt reconciliation. We neither subtract the proxy from total swap fees nor calculate a fee-allocation ratio. Dollar burns are not dividends or guaranteed increases in market capitalization.

The adapter combines v2 and v4 holder revenues within the v3 label. An additional primary-source check found official Proposal 100 EXECUTED on July 27, 2026. It covers eligible static, CCA and aggregator-hook families on Ethereum, Arbitrum, Base, BNB, Polygon, Optimism and Robinhood. The proposal leaves five other chains for a later vote; this edition does not assert their status. The broad activation contradiction is resolved in favor of the executed record; chain-specific receipts remain unaudited. The Unichain Firepit contributes $14,438 over 90 days, without distinguishing swap and sequencer sources. Net sequencer fees are eligible, but no separate amount is added because inclusion cannot be separated. Hooks and third-party application income require specific holder attribution. The page maps coverage and does not assert that every eligible product has an audited contribution.

The UNI documentation establishes governance and the fee-to-burn mechanism, while stating that holders have no individual or pro-rata claim on revenue. The source permits up to 2% annual minting and reports no active inflation. The starting issuance assumption is therefore zero. It is a forward assumption, not a promise of permanent zero issuance. The model does not add treasury assets as distributable NAV. Selling or spending treasury tokens already included in total supply does not create new tokens; it may affect market float, spending and future fee generation, which this simple version does not forecast separately.

## Infer

The intrinsic calculation is an explicitly conditional, discounted economic-benefit proxy. It applies cash-flow valuation mathematics to a conditional recurring economic-benefit stream inferred from the release-based USD burn proxy. It is not a statutory equity valuation, a dividend-discount claim or a forecast of achievable sale proceeds. A burn is not automatically a proportional price increase.

Let B be the selected annualized core fee base, S the snapshot total unburned UNI, g the five-year annual core-fee growth rate, r the annual required return, d annual new issuance, c the realization fraction, and q mature annual fee growth. All rates and cash flows are nominal USD. The default g=25%, r=25%, d=0%, c=100%, q=3% are transparent analyst assumptions. The discount rate is not represented as a statistically estimated cost of equity. Sustainable reinvestment, incentives, collection losses and future redirection of fees can reduce the benefits; c is a stress control, not a measured conversion ratio. The model assumes the remaining platform can sustain the forecast under the retained fee allocation, an assumption requiring subsequent financial reconciliation.

An optional event adds annual fees A with probability p starting in forecast year k. A must be incremental to every fee already included in B. It remains flat through year five. A defaults to zero. These are hypothetical event inputs, not forecasts of a particular proposal. E_t = A*p when t>=k, otherwise E_t=0.

For t from one to five:

    B_t = B * (1+g)^t
    F_t = c * (B_t + E_t)
    S_t = S * (1+d)^t
    f_t = F_t / S_t
    PV_t = f_t / (1+r)^t

Terminal per-token growth is h = (1+q)/(1+d)-1. The perpetuity is finite only if r>h:

    TV_5 = f_5 * (1+h) / (r-h)
    Value_today = sum(PV_t, t=1..5) + TV_5/(1+r)^5

Future burn benefits are represented in F_t. Supply is not also reduced by those future burns, which would count the same effect twice in this proxy. New issuance is handled in S_t. This convention is a limitation of a distribution-equivalent valuation, not a claim that actual token supply will follow this path. No explicit probability of protocol failure is separately estimated. A fee decline, realization haircut, higher discount rate and adverse scenarios test failure risks, but do not measure their probabilities. Terminal survival remains an important assumption.

Price/value = snapshot UNI price / Value_today. If value is zero the ratio is undefined. The percentage change to estimated value is Value_today/price-1. It is not an annualized return or a prediction of convergence. The reverse model uses bisection to find g in [-95%,300%] that reproduces the snapshot price, holding all other assumptions fixed. If there is no solution in that range the result is unavailable, rather than an invented growth estimate.

The bear sensitivity reduces g by 20 percentage points and raises r by 5 points. Growth has a -95% lower bound. The bull sensitivity increases g by 25 points and reduces r by 5 points, subject to a 5% floor and r at least one point above terminal per-token growth. These scenarios have no assigned probabilities and are not confidence intervals. Sensitivity cells show g of 0%,15%,30%,45%,60% against r of 15%,20%,25%,30%,35%.

## Relative pricing

CAKE is an initial fee-and-burn peer. For both assets, aggregate pricing is current price multiplied by total unburned supply. The denominator is holders revenue over the same selected observation period, annualized with 365/days. UNI's denominator includes any user-selected source-retention adjustment, which is explicitly identified in the table. CAKE's denominator remains its unadjusted observation. The peer comparison is before reconciling differences in emissions, incentives, product mix, growth and risk. It is a diagnostic rather than a fair-value target.

    Peer_multiple = CAKE_price * CAKE_total_supply / CAKE_annualized_holder_fees
    Mechanical_UNI_price = Peer_multiple * UNI_selected_annual_fee_base / UNI_total_supply

There is one peer, not a sector median. The quote timestamps are close but not identical. The page does not label this multiple P/E, EV/EBITDA or a distributable cash yield. It does not blend this result with intrinsic estimates.

## Catalysts and option eligibility

The July 7 v4 proposal alone is not proof of execution. The subsequent official Proposal 100 record establishes the first activation on July 27 for its specified chains and pool families, not every v4 deployment. The September 16 StablePair Hook explanation describes product mechanics, including LP benefits. Neither is automatically incremental UNI revenue. The model requires executed governance, deployment state and attributable receipts before incorporating a specific additional fee stream into the published baseline. Existing receipts from a new product are already included in B and cannot also appear in A.

A genuine real option requires an economic underlying, an exercise decision and cost, a time window and a defensible connection to the investor's rights. Those inputs have not been established here, so no separate premium is modeled. This is an applicability judgment rather than proof that all future opportunities have zero value. A Black–Scholes option on UNI would value a derivative, not UNI itself.

## Validate

The page exposes time-window sensitivities, individual chain contributions, removal of Robinhood Chain, new issuance at 2%, a 50% reduction in the selected realization rate, reverse growth and terminal-value dependence. Concentration alone is not a defect. Removing a source is a stress scenario rather than an allegation of fabricated fees. Missing data never becomes zero. We do not use external market narratives as quantitative proof of durable growth.

There is no claim to have independently validated five years of growth, every executed burn, future costs or an optimal peer set. The independent review must rerun calculations from the saved raw API evidence and challenge the methodology. It receives no author conversation or reasoning transcript. Review does not transform analyst assumptions into observed facts.

## Report

Both audiences receive the same token value. Costs, desired valuation cushion, position size and investment constraints affect the decision route. The personal defaults are a $100,000 portfolio, $2,000 proposed exposure, 1% portfolio stress-loss budget, 70% adverse price move, a 1% all-in three-year cost allowance and a 30% margin of safety. Fund defaults are a $100 million portfolio, $1 million exposure, the same loss budget and stress, a 2% cost allowance and a 35% margin. These are illustrative and editable, not an institution's actual policies or observed costs.

The all-in comparison is P*(1+cost). The allowance expresses estimated aggregate costs relative to the original position value, conservatively charged at acquisition for the screen; it is not a detailed cash-flow schedule or an after-tax forecast.

    Potential_add_price = Value_today*(1-margin)/(1+cost)
    Potential_hold_upper_price = Value_today/(1+cost)

Below the add threshold the output is 'potential add'; between it and the hold upper bound it is 'potential hold'; above it the output is 'not supported at this price.' A formal buy/hold/sell recommendation remains unavailable until the material evidence gaps are resolved. These bands are screening conventions, not mathematically unique optimal decisions. An already-owned token is assessed on its prospective economics, not its purchase history. Cash-flow uncertainty can warrant rejection even below an apparent buy threshold.

The stress-loss check is:

    Stress_loss = Position_USD * Assumed_drawdown
    Portfolio_loss_fraction = Stress_loss / Portfolio_USD
    Position_ceiling = Portfolio_USD * Loss_budget_fraction / Assumed_drawdown

The ceiling is not a recommended allocation. A full-loss case is disclosed as position/portfolio. No probability is assigned to either drawdown. Fund execution capacity is unknown until independently supported venue depth, custody and mandate evidence is available. A pass on valuation or stress arithmetic cannot approve those missing requirements.

The first planned observation review is October 19, 2026, followed by December 19. A 30-day annualized holder-fee pace more than 25% below the selected fee baseline is an analyst review trigger. Governance redirection, issuance changes and security events trigger earlier reconsideration. No automatic alerts or trades are created.

## Publication and reproducibility

This is a frozen research edition. Data fetching happens outside the page. The page makes no market API calls and no periodic refresh is installed. Changing assumptions recalculates locally. The downloadable analysis includes the evidence, selected assumptions, valuation, reverse-growth result and ownership screen. A new dated edition requires a fresh collection, validation and independent review.

The API collectors used are CoinGecko's coin endpoint and DefiLlama's fees-summary endpoint. Public protocol documentation supplies the rights and mechanism evidence. Raw input SHA-256 hashes are stored in evidence.json. The collection script, valuation model and tests are included in the source package. Chart assets are served locally.


## External comparisons and questions

The initial independent calculation was completed before the author’s outside-estimate comparison. SEBA Bank’s March 2021 report, page 10, gives $55.06 equally weighted and $18.78 base cash-flow value before its 2× governance multiplier. The Mispriced Research’s September 2, 2026 article gives a $5.00–$6.75 base scenario. Different dates, supply definitions and undisclosed inputs preclude treating proximity as accuracy. These reports did not set Charlie’s assumptions.

All three method families are assessed. They are not mechanically averaged. The intrinsic proxy is conditional; the one-peer pricing result is diagnostic; a separate option premium is not estimable from the documented rights. Neither the retail nor institutional route converts those limitations into an approved recommendation.

The question service uses only the frozen public page, methodology and a generated default-results digest. It uses the existing VPS OpenAI integration (gpt-4.1-mini), with the Gemini 2.5 Flash Lite through OpenRouter as a backup, without browsing or other tools. Questions and up to six preceding messages supply conversational context, not factual authority. It refers to published default assumptions, not altered browser controls. Credentials stay on the server. Answers are rendered as plain text with server-approved section links. Request length, concurrency and rate limits bound resource use; model-call metadata, but not question text, may enter the existing usage log.


## Remaining evidence, error direction and next procurement

Official Proposal 100 (https://vote.uniswapfoundation.org/proposals/100) resolves the first v4 activation. Official Unichain implementation (https://raw.githubusercontent.com/Uniswap/unichain-contracts/main/README.md) confirms fee splitting and forwarding to TokenJar. Optimism receives max(15% net fees, 2.5% gross fees); use net attributable amounts rather than gross chain fees. The recorded route and same-collector address establish overlap, not a separately measured increment.

To replace the burn-value proxy with an independent receipt measure, obtain per-chain fee-asset collections, historical USD valuations, starting and ending unspent balances, historical Firepit thresholds, release events and UNI transfers/burns. Reconcile opening balance plus receipts minus releases and other outflows to closing balance. Match product, chain and observation time before annualizing. In Unichain, the documented earnedFees(forwarder) read can establish an accrued balance at a historical block if an archival endpoint is available; that single balance alone cannot establish a 90-day revenue flow.

A cutoff supply ledger must reconcile minting, burns, bridge representation and treasury balances without counting bridged tokens twice. A full economic forecast must also establish incentive funding, necessary costs and reinvestment, governance allocation risk and possible failure. Neither a favorable mathematical sensitivity nor an executed proposal replaces these records.

Error direction is two-sided. Backlog release, duplicate product streams, too little denominator supply and omitted necessary spending can overstate value (false attractiveness). Unreleased attributable receipts, omitted live streams and too much denominator supply can understate value (false rejection). These are mechanisms, not measured probabilities or a confidence interval. At the original rates, +$1m of annual starting benefit adds approximately $0.0109 per UNI. Keeping every other assumption fixed, a $9.35 result requires approximately $857.8m annual starting benefit. No evidence here establishes that missing amount.

Publication status is a conditional research screen with incomplete economic reconciliation, not an affirmative investment rating or proof of forecasting accuracy.

The follow-up VPS request to DefiLlama’s Unichain dailyRevenue endpoint succeeded, but its methodology describes gas fees and burned coins; it does not establish a separate UNI-attributable fee flow. The block lookup for the cutoff returned height 59027641. Neither response replaces historical contract reads and receipt reconciliation, so neither adds value to the model.


## Revised decision presentation

The completed calculation and the incomplete evidence must be stated separately. Price gap uses estimate / price − 1; the price premium uses price / estimate − 1. At the published values, the estimate is 87.9592% below the September 19 $9.35 quote, while price is 730.5072% above the estimate. The denominator differs; these are the same discrepancy. Neither is a return forecast. Price is a frozen reference, not a live quote or yesterday’s observation.

The investment inference is objective within the assumed case: the modeled benefits do not justify the acquisition price. Costs further widen the required economic support. An existing holding bought for these same modeled benefits needs a stronger forward case; past ownership does not change future value. This is not a personalized sell instruction or a complete estimate of every possible token benefit.

Completed: saved API observations and windows, model calculations, sensitivity comparisons and rule-based ownership decisions. Assumed: growth, discount rate, issuance and benefit realization. Incomplete: receipt-to-burn reconciliation, separated product contributions, independent supply ledger and necessary costs/reinvestment. The missing evidence can cause under- or overestimation and is not described as complete.

Institutional relevance extends beyond UNI ownership. Pool LPs, execution providers, integrators, company-equity investors and strategic governance buyers have different claims, revenue and cost functions. This edition maps those decision routes and their required inputs; it does not calculate unsupplied position-specific income, business valuation or strategic control benefits. The UNI price gap cannot be transferred to those activities as a conclusion that they are unattractive. UNI's core numerical scope and financial inputs remain unchanged.


## Reader decisions and business examples, revision 2

The main page uses reader-facing headings. Detailed incomplete-evidence lists remain in its Assumptions, calculations and data limits section and this method. An assumption label stays next to every estimate. The saved $9.35 quote is September 19, not yesterday's $8.31: 8.31× is the price/value ratio. Financial raw observations and UNI model equations have not changed.

Use-case examples are hypothetical and separated from the saved evidence. Swap example: 990−985=$5 more output, assuming the same asset, trade size, timestamp and all-in costs. LP example: assets excluding fees 10100+fees600−cost100=10600 versus10500 buy-and-hold; extra100 equals1% of starting10000. Scaling requires proportional returns/costs and is not assumed achievable. Execution-service example: 1000000×0.001−700=$300 contribution before tax. Equity example: (100m+20m−10m)/10m=$11; offered12 is9.0909%above estimate. This is not a Uniswap Labs valuation. Governance hurdle:100000×(9.35−1.1258180299486806)=$822418.19700513 before extra costs. This credits the model's underlying token value and demands incremental strategic benefits; it does not estimate control, the outcome probability or votes needed.

The three-year business example is separate from the five-year UNI model. Assumed default build cost100000, annual retained revenue100000, annual running cost40000, annual required return20%. With t=1,2,3, factor=sum((1+r)^−t), net income=revenue−running cost, PV_income=net×factor, surplus=PV_income−build cost. Break-even annual retained revenue=running cost+build cost/factor. Default factor2.10648148148, PV126388.888889, surplus26388.888889, break-even87472.527473. Cash is received at year-end, the build cost today, no terminal resale value. Running costs should include taxes, losses, maintenance and required reinvestment; all values are illustrative. A positive result meets the specified discount return, not every operational or commercial condition. Zero discount uses factor3. Negative income is retained as a loss. Blank/nonfinite/out-of-range inputs suppress the result.

Ownership controls describe a proposed purchase and aggregate acquisition/ownership costs. An existing owner should compare future costs only, excluding sunk entry costs, and consider taxes and alternative investments. A price between the cushion threshold and model value is not an automatic hold recommendation. Audience controls now preserve separately entered personal and fund inputs.

Institutional sources distinguish actual fund exposure, equity and execution. Grayscale's July31 2026 fact sheet reports39.49% UNI composition; it is not a purchase price. Bitwise's fund directly invests in UNI. BlackRock's February11 joint release says a strategic ecosystem investment without identifying its instrument or price. Talos's August21 2025 report confirms v2/v3/v4 integration. Labs's October2022 $165m round is company financing, not UNI demand. DEF's July2021 disclosed $20.40 Genesis sale was facilitated execution, not a long-term buyer cost basis. These observations add no separate numerical UNI premium.

Connected research pages have independent timestamps. Crosslinks are navigation, not a combined model that automatically recomputes valuation from market signals. A positive technical signal alone does not close a valuation gap. The completed recommendation under the published token assumptions is to wait on a new purchase for those benefits; this does not forecast short-term returns or ban trading under a separately specified strategy.

Connected UNI sections added with this revision: Binance completed daily prices through September18 give7d+47.4551%,30d+143.879%,SMA50=4.96028USDT,latestclose8.865USDT and7d/prior21d average quotevolume1.4388079×. Futures coin-unit OI September12→19 00UTC grows13.6525%; notional grows67.8016%, partly a price effect, not newlongdemand. DefiLlama v3+v4 completedweeks total19.989238047bn versus20.941555709bn (−4.5475%). These are different assets/venues and are not an equivalent-volume comparison. The price rally does not establish greater UNI-holder benefits. Raw snapshots and the build script for all three connected sections are included in the source archive.