NEAR Protocol NEAR

NEAR spent years as an easy pass.

Not because anything about it was broken. As just another general-purpose blockchain (an "L1," short for Layer 1, the base network everything else is built on, with its own native token, also called NEAR, that trades on exchanges), it simply wasn't exciting. For plain smart-contract exposure, Ethereum or Solana offered more liquidity, more developers, more real usage, more staying power. NEAR worked fine technically; it just didn't have much happening on it. A functional, empty city is still empty.

What changes that assessment is a product called near.com.

For the first time, NEAR doesn't read as a blockchain waiting for developers to build something worthwhile on top of it. It reads as a financial app that hides crypto's messy technical parts from the person using it. That's the case this piece makes.

The bet isn't that NEAR will process more transactions. It's that NEAR is turning into the thing that executes trades, protects user privacy, and eventually powers payments between AI programs, what's often called "agentic finance."

That's also why this is interesting, and why it isn't a large position yet. Based on how much money NEAR earns today, the token is genuinely expensive, not "expensive but justifiable." The case for holding any at all rests on today's earnings being the wrong number to judge it by: if near.com becomes the account people actually use, and the technology behind it, NEAR Intents (more on that shortly), becomes the plumbing behind wallets, private transactions, native crypto assets, tokenized assets, and AI agents making payments, the earnings base changes.

This is treated as a venture-style bet, high risk and high potential payoff, sized at around 3% of a tracked research position. The position was first entered at $1.55 with a 2% allocation, and has grown since as the price has risen.

Blockspace Is Turning Into a Commodity

Start with something most people in crypto avoid saying out loud: blockspace, the raw computing capacity blockchains sell, is a commodity, and its price keeps falling.

From roughly 2015 to 2021, the smart move was to buy the infrastructure itself: the base blockchains (Layer 1s) and the scaling networks on top of them (Layer 2s), rather than guess which app would win. That's why Ethereum went from a few dollars to thousands, and Solana did the same a cycle later. Own the base layer, and the trade won big.

Here's why that trade worked, and why it stopped. Back then, everyone knew useful apps were coming but nobody knew which would succeed, so the rational move was buying the "picks and shovels," the layer every future app would have to run on. It's the same logic as buying Cisco in the late 1990s: the internet was coming, but nobody knew if the winner would be Pets.com or Amazon. There was no need to pick the website. The routers and switches all of them would run through were the thing to own.

That era is over, for two reasons. First, blockspace supply exploded: dozens of credible chains plus a wave of Layer 2s pushed the cost of onchain computing toward zero. Second, the industry has matured enough that it's now fairly clear which applications actually matter, fewer than 20 genuinely do. Once the winning apps are visible, there's no need to gamble on the platform beneath them; the apps themselves become investable, because that's where real usage and cash flow show up.

In other words, value moves up the stack over time: it sits in infrastructure when nobody knows what will get built, then shifts to the application layer once it's clear what people actually use. That shift is underway now.

This matters even more for a trend already well underway: real-world assets moving onchain. Once those assets exist onchain, they need somewhere to actually be used: traded, lent against, used as collateral, settled. That "somewhere" is an application, not just a blockchain. So value concentrates next in the app a tokenized asset flows through, not the chain it happens to sit on.

Keep that in mind, because NEAR turns out to be an unusual, specific answer to exactly this problem.

Why NEAR Was Easy to Overlook

NEAR was easy to overlook for years, and at the time, that was the right call.

For most of its existence, NEAR was a chain where little happened: low total value locked (TVL, how much money is actually parked in a chain's apps), thin real activity, and a token price that mostly moved with general crypto sentiment rather than anything NEAR was doing. A well-built city nobody moved into: good roads, empty streets. As an infrastructure bet, it failed the one test that matters: was anyone using it? The honest answer was no.

Before this year, the fair assessment was that NEAR was a technically solid chain with no real reason to own it. That was accurate at the time.

What changed is near.com, and the technology it's built on. The product only makes sense once the problem it solves is clear.

The Product That Actually Changes Things

near.com is the reason the category shifts.

The pitch is simple: one account gives you swaps, yield, and private transactions across 40 blockchain networks, all from one place, with independent market makers competing to give you the best price. The technology behind it has reportedly already processed around $19 billion in volume across 103 assets and 40 chains.

But the marketing line isn't the important part. The user experience is.

Using crypto across different blockchains is still genuinely painful; even experienced users tolerate it rather than enjoy it. Say a user holds USDC, a stablecoin (a crypto token designed to hold a steady $1 value), on Solana and wants to buy something on Base, another blockchain. Funds need to "bridge" over, the right token has to be held to pay network fees, a route has to be chosen, delay accepted, slippage managed, all while hoping for a rate that isn't quietly worse than advertised. Starting from USDT on Tron is more annoying still; starting from actual Bitcoin is close to absurd. Testing each of these routes firsthand confirms it: the experience is unpleasant, and a mistake can cost real funds.

This complexity isn't all bad. It's exactly the friction that's let sophisticated users make good money, because moving value between chains is hard. But it also caps adoption: it basically takes becoming a technical expert to do this directly, and only so many people will bother. That ceiling looks close: most users who will ever go direct, the hard way, likely already have.

That ceiling is the opportunity. If someone makes the hard way disappear, the next wave of users doesn't need to be braver, they just need a better product. That's what near.com is trying to be.

near.com and "Intents"

The engine running underneath near.com is called NEAR Intents, and it's the part that does the actual work.

Normally, crypto puts the burden on the user to plan the entire path: which chain, which bridge, which exchange, which route, which token to pay fees with. Intents flips that. The user simply states the outcome wanted, and a network of "solvers," competing market makers, race to find the best route there. Whoever wins settles the trade through a verification system built into NEAR. In plain terms: instead of specifying every step, the system is told the result, and it figures out how.

A user selling one asset for another doesn't want to manage five separate transactions. They want the best price. Intents turns routing into a competitive market where solvers shrink their own margins just to win the trade. That's why the design is genuinely powerful: it's not just a nicer bridge, it functions like an onchain prime brokerage, the kind of service that handles execution and settlement for large institutional investors.

near.com is the consumer app sitting on top of all this: one account, swaps across 40 chains, transfers, and private transactions, without asking the user to understand the machinery underneath.

In testing, most transactions settled in about 5 seconds; one took roughly 3 minutes, the kind of thing that'll likely get fixed, and even that was tolerable. For anyone who's spent years bridging, approving, wrapping, and waiting, just stating what's wanted and having it happen is a genuinely new experience, not a small one.

Two caveats. First, the assets available in the near.com interface are deliberately limited right now, a testing-phase decision, not a technical ceiling; the underlying rails support far more, and that will expand. Second, a card is reportedly coming, the thing that turns this from "a crypto app" into "the account you actually spend from." That claim is worth treating as unconfirmed until it ships, but the direction is clear.

The Privacy Angle

There's a second product worth naming, because most people are sleeping on it: Confidential Payments.

This isn't a claim that NEAR's privacy tech cryptographically beats Zcash, the cryptocurrency built specifically around private transactions. Zcash has its own design and a much longer history as private money. The real point is different: Zcash is private money; NEAR is trying to become a private financial account. That's a bigger surface.

Most people don't want to move their entire financial life into one volatile privacy coin. They want to hold whatever they hold, send whatever they want, receive on whatever network makes sense, swap when needed, and not broadcast their full transaction history just because they paid someone $100. That expectation is normal in traditional finance and strangely controversial in crypto: nobody would use a bank account if the account number exposed their entire history, yet public blockchains normalized exactly that.

near.com says deposits, transfers, and swaps stay confidential, visible only to the account holder. If that works, the privacy pitch isn't crypto ideology. It's normal financial discretion.

This is why the Revolut comparison helps: Revolut isn't compelling because it invented money, it's compelling because it collapsed fragmented financial tasks into one interface. NEAR is trying to do that for crypto: onchain assets, confidential transfers, cross-chain execution, and eventually tokenized assets, bank accounts, and cards, all in one place. That's the real privacy thesis: not privacy-coin speculation, but a private financial life.

It's worth being careful not to oversell this point. NEAR's confidential design is presented as compliance-aware and selectively disclosable, provable to a regulator if required, rather than a black-box mixer. That distinction matters: compliance-aware, disclosable privacy is the kind that tends to survive regulators, while opaque, undisclosable mixing is the kind that gets shut down. Whether it holds up legally and technically at scale is unproven. But the intent, privacy across a user's whole financial life and not just one coin, is correct, and a better answer to the real problem than what Zcash offers.

The Business Model Is Inverted (In a Good Way)

This is the part that flips the argument this piece opened with.

The opening argument was that blockspace is a commodity and the base layer captures less value over time by default. NEAR escaped that default by building the dominant trading application on top of its own blockchain, then wiring that app's revenue straight back into its own token.

When a swap happens through near.com's own front end, 100% of the fee goes to NEAR as revenue. Through a third-party app plugged into Intents, the fee splits roughly 50/50. That revenue doesn't sit in a treasury. It funds buybacks of the NEAR token: the same mechanic as a company using profits to repurchase its own stock, permanently removing tokens from circulation and increasing the value of each one that remains.

Think about what that means. NEAR stops being a chain valued on typical Layer-1 metrics like total value locked or active addresses, metrics that never made it investable. It's now a chain that owns a high-volume exchange, that exchange generates real revenue, and that revenue buys back the token. It can be valued the way an exchange is valued: revenue, a multiple on that revenue, and a buyback returning value to holders. NEAR is the coordination layer taking a fee on every dollar that moves through it, and near.com is the front door capturing the largest cut.

Base layers almost never capture value this way. NEAR flipped it by building the killer app itself and routing its revenue back into the base layer's token.

This also gives the cleanest forward lever in the whole thesis: as more volume shifts to near.com's own front end instead of partners, NEAR's blended cut of every fee dollar rises, since it keeps 100% instead of 50%. Revenue can grow faster than volume.

AI and Privacy: The Two Themes NEAR Actually Sits On

If the flipped revenue model explains why NEAR is a real business today, this section covers why it may have a second act bigger than the first. NEAR sits right at the intersection of two of the most important forward-looking trends in the space: "AI x Crypto" and privacy.

Most "AI x Crypto" plays are lazy. A project slaps an AI-themed page onto its website, the market rewards it with a higher valuation for a while, and then the story fades because there's no actual usage or revenue behind it.

NEAR is more genuinely interesting than that, though it still deserves some healthy skepticism.

NEAR's co-founder, Illia Polosukhin, was one of eight authors of "Attention Is All You Need," the 2017 Google paper that introduced the Transformer, the architecture behind essentially every major AI model today, including GPT, Claude, and Gemini. The person who helped invent the technology modern AI runs on left Google to build NEAR. So when NEAR talks about an AI-driven future, that's not a marketing pivot chasing a trend. It's the founder returning to the problem he's arguably among the most qualified people alive to work on.

The underlying thesis: AI agents, software that can act and transact on their own, will outnumber humans this decade. They'll transact constantly, paying each other for services, buying compute, settling jobs, hiring other agents, and none of that can run through banking, because no bank is going to KYC a piece of software. The only financial system natively compatible with agents is one where identity is a public key, settlement is programmatic, and a payment is a function call. That's crypto.

The question is which piece of crypto becomes the rails, and here's why Intents beats a plain chain for that job. An AI agent doesn't want to hold gas tokens (the tokens used to pay network transaction fees) on 40 chains or know that USDT on Tron needs a different token for fees. It wants to express an outcome: best price, move this value, execute. That's exactly what an intent is. Intents are, almost by accident, the most agent-native transaction model that exists: the same product that makes near.com pleasant for a human is what makes it usable for a machine, since both just want to state a result and not manage how it happens.

NEAR has built the lower-level primitives too. "Chain signatures" let a single NEAR account control keys on other chains via multi-party computation (splitting control of a private key across multiple parties so no single one ever holds the whole key), letting an agent act natively across Bitcoin, Ethereum, and beyond from one identity. NEAR is also building agent frameworks, including agents running inside trusted execution environments (secure hardware that hides data even from the machine running it) so they can hold keys and act autonomously without ever exposing them to a human.

To be clear on the risk: the agent economy at scale is still mostly ahead of us, NEAR's own throughput numbers are aspirational, and plenty of chains have claimed the "AI" mantle and delivered nothing. What separates NEAR is that this story has an actual revenue mechanic attached: every agent transaction through Intents is volume, volume is fees, fees are revenue, revenue is buybacks. The narrative and the cash flow run through the same pipe.

Tokenized Stocks Are the Tailwind

There's one more structural reason this matters over the next 2 to 3 years, connected directly to a broader, already-established trend: real-world assets moving onchain.

Tokenized stocks are coming, but most of that shift will be invisible. It'll show up as a behind-the-scenes upgrade at brokerages people already use. Custodians adopt tokenized rails to extend margin and settle around the clock, and the end user feels nothing. The stock in a brokerage account quietly becomes tokenized and the holder never knows. That's how most of the volume arrives: invisibly, through incumbents.

But the part that stays visibly onchain stays fragmented across many chains and venues, and somebody has to solve cross-chain settlement for it. That difficulty is well documented: teams building onchain aggregators for exactly this have run into the same wall. The hard part was never the front end, it was settlement. If a user holds USDT on Tron and wants to trade on Solana or Base, the choice is either pre-swap and pre-bridge ahead of time, eating the cost and risking a wrong guess, or sit through 10–15 minutes of cross-chain plumbing once a decision is made. There's no clean answer with the old tooling.

Intents is a clean answer: a best-execution, native-asset, cross-chain settlement layer is precisely the missing piece for a world full of fragmented tokenized assets. The same megatrend already has an issuance-side expression through tokenization platforms like Securitize (a firm that issues blockchain-based versions of traditional securities); NEAR stands out as the most credible settlement-layer version of it currently available.

The Numbers

Now for the part that matters most, because everything described above is a genuinely good business, but the price of a token doesn't care how good a business is until the business actually grows into that price.

Here's exactly where NEAR stands today, based on the protocol's own revenue dashboard: market cap ~$3.1B, circulating supply ~1.3B NEAR, inflation cut from 5% to 2.5%, with the supply now fully vested (no more scheduled token unlocks), all-time fees $55.1M against all-time revenue (NEAR's cut) of $7.8M, year-to-date fees over ~5 months of $33.25M against year-to-date revenue of $4.6M, and last-30-day fees of $4.5M against last-30-day revenue of $774K.

Two definitions before any of this means anything. "Fees" is the total amount paid across the whole system, including the cut that solvers and partner apps keep. "Revenue" is the slice NEAR itself actually keeps, and that's what funds the token buybacks. The number that actually matters for valuing the token is revenue, not fees. Everything below is built on revenue.

Three things jump out of that table.

First, annualize the year-to-date revenue (project its pace out across a full 12 months) and it comes to roughly $11M a year. Against a $3.1B market cap, that's about 280 times revenue. That's not a typo. On a price-to-revenue basis, NEAR ranks among the most expensive assets covered in this research, by a wide margin.

Second, fees from just the last 30 days annualize to about $54M, almost matching the protocol's entire all-time fee total. The jump since near.com launched is real, and its effect on revenue is visible.

Third, the share of fees NEAR actually keeps as revenue is climbing: about 14% all-time and year-to-date, but about 17% in just the last 30 days. That means NEAR is keeping a bigger cut of every dollar moving through the system, exactly what would be expected as more volume shifts to near.com's own interface, at 100% capture, instead of through partner apps, at roughly 50%. The take rate is expanding.

The messier part matters too, because false precision helps no one. Revenue from just the last 30 days annualizes to about $9.3M, actually a touch below the year-to-date run-rate. Revenue is lumpy month to month, and there was a launch-related spike that flatters the average. Revenue is not yet cleanly compounding every single month. What is true: the business took a real structural step up this year and now operates at a scale it's never hit before. The forward bet is that adoption keeps that engine running. That's a thesis, not a guaranteed trend line.

So the right question isn't "does today's revenue justify today's price?" It doesn't, and it's not close. The better question is: what has to be true for today's price to actually make sense?

Reverse-engineering it: starting from the $3.1B valuation, here's the revenue NEAR would need at different multiples, where a "multiple" just means how many times a company's revenue investors are willing to pay for it. At 100x, it would need $31M in revenue, about 2.8 times today's ~$11M. At 50x, $62M, about 5.6 times today's. At 25x, $124M, about 11.3 times today's. At 15x, $207M, about 18.8 times today's.

Here's how to read it. If the market keeps paying an aggressive growth-stage multiple of 100x, NEAR only needs to roughly triple its revenue to justify today's price. If the market settles into a still-rich-but-more-reasonable 25x, the kind of multiple expected from a fast-growing exchange, NEAR needs to grow revenue more than elevenfold. At today's price, that means underwriting something like a tenfold or better increase in revenue over the next 2 to 3 years, just to stand still at a "reasonable" multiple. That's a very high bar, which is exactly why the position has been sized at only 2-3% of a tracked allocation so far.

This is usually the point where most write-ups on NEAR, including earlier coverage, quietly pick one multiple, plug it in, and move on. That approach doesn't hold up here, because the multiple assumption is doing most of the work in the price, and it's also the hardest variable to predict with confidence. Volume, revenue, and the take-rate climbing as near.com wins more of its own traffic can be reasonably modeled. What multiple the market will pay for that revenue in the future cannot be modeled with any real confidence, and pretending otherwise is where this kind of analysis usually goes soft. Before getting into forward scenarios, the multiple itself is worth taking apart piece by piece.

What Multiple Is Actually Fair?

A "multiple" isn't a fixed fact about a company. It's a price the market agrees to temporarily lend an asset, for as long as its story holds up. NEAR's roughly 280x today isn't one number. It's four separate layers, stacked on top of each other. Gravity, at 15x–25x, is what's being paid for cash flow and the buyback, and nothing takes it away: it's the floor. Franchise, at 30x–40x, is what's being paid for best-in-class growth and a rising take rate, and a credible competitor is what takes it away. Scarcity, at 50x–70x, is what's being paid for being the only large-scale way to express this theme, and a real rival or a cooling theme is what takes it away. Attention, at roughly 200x on top of all that, is what's being paid for trading flow, a thin available supply, and a loud narrative, and tightening liquidity or the crowd simply leaving is what takes it away.

The bottom two layers, Gravity and Franchise, are multiples NEAR can actually earn through real execution and cash flow. The top two, Scarcity and Attention, are essentially rented. They exist because there's currently nowhere else for theme-chasing crypto money to go, and because nobody credible is competing at NEAR's level for cross-chain coordination.

When NEAR bottomed out near $1, it was still trading at 80x–100x revenue on an even smaller revenue base than today. It has never once traded at a "sane" multiple, because the market never priced it as an exchange. It priced it as a Layer 1 blockchain, essentially a lottery ticket, and an empty chain with a buyback attached is still mostly a lottery ticket. The repricing this thesis is betting on is the market finally valuing NEAR for what it actually became this year: a prime brokerage that happens to also own the blockchain it settles on.

Now the part that actually matters for risk: the moat itself is worth being precise about, since it's the most debatable part of this thesis. Services like Circle's CCTP, LayerZero, and Wormhole move a token from one chain to another; that's transport, and transport commoditizes. NEAR Intents competes one layer above that, on best execution, and best execution is a solver marketplace, the one structure here that compounds a moat, since more solvers mean better prices, better prices pull more flow, and more flow pulls more solvers. That's a two-sided network effect, the most durable kind there is, but only if it scales before someone with deeper pockets builds a rival. near.com itself, the consumer app, is the replaceable piece: an interface, defended by distribution and habit rather than technology, the way Revolut and Robinhood won on distribution, not on being uncopyable. So the durable moat, if it exists, lives in Intents, not the front end.

That's why this analysis doesn't lean on a single multiple assumption. Instead, a few different scenarios are modeled below, letting the numbers speak for themselves, with the "Base" case treated as the central estimate.

These forward scenarios model a 2-to-3-year horizon, matching the broader tokenization timeline referenced earlier, and hold token supply roughly flat at about 1.3 billion, since ongoing buybacks and the now-reduced 2.5% inflation rate roughly cancel each other out. If buybacks end up outpacing new token issuance, supply actually shrinks and every number below gets better, so "flat" is the conservative assumption.

NEAR is currently doing about $3 billion in trading volume a month, comparable to 1inch, at $3.91B, and ahead of CoW Swap, at $2.6B, two established crypto exchanges, and still growing. Applying NEAR's roughly 0.15% gross fee rate and 17% revenue capture rate gets to the $9–11M of annual revenue it earns today. Each scenario below assumes a different level of future monthly volume, benchmarked against real comparable platforms: the bear case keeps NEAR below Aerodrome at roughly $14.5B a month, the base case puts it just above Aerodrome and at about half of Uniswap's volume, and the bull case assumes Uniswap-level scale at roughly $36B a month. Against a $1.55 entry price, the bear case works out to about $32M of revenue over 2–3 years, a 32x multiple, a ~$1.0B valuation, and an $0.80 price, down 48%. The base case works out to about $115M of revenue, a 106x multiple, a ~$12.2B valuation, and a $9.40 price, up 506%. The bull case works out to about $300M of revenue, a 130x multiple, a ~$39B valuation, and a $30 price, up 1,835%.

In plain language:

The volume assumption is defensible. NEAR already runs at mid-tier exchange scale, and it does the one job single-chain exchanges structurally can't: moving native assets between blockchains. Reaching a bit above Aerodrome's volume over 2–3 years, before even counting tokenized-asset trading, isn't a stretch. If this thesis fails, it probably won't fail here.

The multiple is rented, not earned. The base case quietly assumes the market keeps paying something like 100x revenue three years out, a scarcity multiple, not a mature-exchange one, holding only if NEAR stays the only real option and capital is still short on places to chase this theme. Price that same ~$115M of revenue like an ordinary competed business at 50x–70x, and it's worth $5–$6, not $9.40. That gap isn't the business. It's premium nobody controls.

The bear case is severe because both assumptions fail together. The $0.80 bear scenario still assumes a 32x multiple, the market still treating NEAR as a credible exchange. The real floor is lower: if volume stalls and the multiple also compresses to the 15x–25x of an ordinary cash-flow business, that revenue is worth closer to $0.40. That's the gap that matters most: the rented premium evaporating before revenue grows enough to hold the price up on its own. A fall back under $1 would be the first real warning sign worth watching.

Putting the three pieces together, the bet has a clean shape: volume that can be underwritten with confidence, a multiple that's essentially rented, and a position that's a wager that the rented premium holds up long enough for real revenue growth to catch up and support the price on its own merits. The single number worth watching most closely isn't the price. It's the revenue-to-fees capture rate, because that's the earliest signal that near.com is winning real usage, and that the business is growing into its price rather than the other way around.

What Would Change This View

In roughly the order these matter, from most to least:

The multiple is the entire risk. At 280x revenue, what mostly drives the price is which multiple the market is willing to pay, not what the business does. A sentiment shift can compress that multiple faster than revenue can grow to catch up. That's the single biggest reason the bear case is so severe.

Intents aren't exclusive to NEAR. Solver competition and intent-based execution are an idea, not a permanent moat. CoW Swap did it first, on Ethereum. Others can and will build competing cross-chain intent systems. NEAR's edge today is execution quality, integrations, and being early to the hardest chains, but a model this promising will attract well-funded competitors, and long-term defensibility is a real open question.

near.com has to actually win everyday users. The entire base and bull case rests on it growing beyond people already comfortable onchain. Crypto is full of good products that never escaped the crypto-native bubble. The card, the wider asset selection, and the overall experience all have to land. If near.com stays a power-user tool, that's the bear case.

The AI volume might not show up. The bigger second act depends on agents transacting at real scale through Intents, a bet on the future, not revenue that exists today. If that economy is slower, smaller, or routes elsewhere, the larger half of the story never arrives, leaving only the consumer app to value.

Privacy invites regulatory scrutiny. Confidential Payments is a feature and a liability at once, meant to be compliance-friendly and provable, not a mixer, but privacy infrastructure still attracts regulators, and one bad ruling could force changes.

Cross-chain is a large attack surface. This depends on safely moving value across 35-plus chains via multi-party computation, solvers, and bridges, a large, complicated area to secure. A serious exploit anywhere in it would hit confidence and volume at once.

Position Sizing

NEAR went from a chain with little reason to own to one of the more interesting businesses in all of crypto, and it got there by doing the one thing base-layer blockchains almost never manage to do: it built its own killer application and routed that application's revenue back into its own token. On top of that, it sits exactly where the next five years of crypto look to be heading, at the intersection of AI and privacy, with a founder who literally helped invent the technology the entire AI industry now runs on, and with a privacy model that protects a user's whole financial life rather than just one volatile coin.

This position is currently sized at roughly 3% of a tracked research allocation, grown from an initial 2% entered around $1.55 on 14 May 2026.

Everything covered in the first three-quarters of this piece argues for owning a lot more of this: the flipped revenue model, the AI-and-privacy overlap, the founder's pedigree, and its role as a settlement layer for the broader tokenization trend. On the quality of the thesis alone, this is a name that could reasonably be sized as a high-conviction position.

The position stays smaller than that, for one simple reason: the current numbers don't support it yet. The valuation, at 280 times revenue, is genuinely expensive, but a 2-3% position that goes to zero only costs 2-3% of the tracked allocation. The same position, playing out the way the bull case describes, still contributes meaningfully, without betting a large share of a portfolio on a market multiple nobody controls. Revenue is left to do the work of catching up to the price, and the number worth watching most closely is the revenue-to-fees capture rate, because that's the earliest sign that near.com is actually winning the flow.