NEAR Payments Rail Thesis
Investment thesis · NEAR HACKATHON

Is NEAR already a payments rail?

If NEAR is becoming money infrastructure, its stablecoins should move like payments: lots of small transfers between ordinary wallets, a steady daily rhythm, and no dependence on NEAR's price swings. If it is mostly a trading venue, the flow will run through DEXs and bots and follow volatility. We pulled USDC and USDT transfers straight from NEAR mainnet blocks over days and checked them against six tests we wrote down first.

Verdict

pre-registered tests passed

Six tests, fixed before the data was pulled

Each test is a yes-or-no rule. If NEAR's stablecoin flow were a payments network, all six should pass. The thresholds live in THESIS.md and the code that scores them is scripts/analyze.py.

The evidence

Every number is an estimate from a seeded random sample of whole mainnet blocks, scaled up by the number of blocks in each day. 95% intervals come from resampling blocks.

1. Who is on the other side of a NEAR stablecoin transfer?

Source: NEP-141 ft_transfer events from sampled blocks (FASTNEAR neardata.xyz). Counterparty types: labels/known_accounts.json + contract-code check via NEAR RPC. Script: analyze.py

2. Weekly stablecoin transfers by counterparty

Source: sampled blocks, expanded to full weeks; complete weeks only. Data: daily_estimates.csv

3. Transfer size: wallet-to-wallet vs DEX

Source: sampled ft_transfer events, share of transfer count per size bucket. Script: analyze.py

4. Payments fingerprint: round-dollar amounts

Share of transfers of $1 or more whose amount is a whole dollar (no cents). Invoices, payroll and remittances tend to be round; swap legs almost never are.

5. Daily transfers vs NEAR price

Transfers: sampled blocks, 7-day average. Price: CoinGecko daily close (/coins/near/market_chart). Spearman correlations in summary.json.

6. Stablecoin supply on NEAR vs NEAR price

Source: DefiLlama stablecoins API (/stablecoincharts/Near), circulating USD, all stablecoins incl. bridged. Price: CoinGecko.

7. Biggest counterparties in the sample

Share = sampled volume sent or received by the account / total sampled volume (each transfer counted on both sides, so shares add to 200%).

Where the payment-like dollars may be going: NEAR Intents

Not pre-registered, so read it as exploration. Stablecoins from other chains (USDC on Base, USDT on Tron and Ethereum, and more) sit inside the NEAR Intents ledger as multi-token balances. Their moves never appear as USDC or USDT token transfers, so the tests above cannot see them. Here we read the Intents verifier's own events from the same sampled blocks.

8. Weekly stablecoin flow through NEAR Intents, by type

Source: intents.near events in sampled blocks: dip4 token_diff (swaps), dip4 transfer (to solvers / peer), nep245 mt_mint / mt_burn (deposits / withdrawals). Stablecoin asset ids: intents_stablecoins.json from the NEAR Intents 1Click token list.

9. Which chain's dollars move through Intents

Share of sampled deposit, withdrawal and swap volume by the origin chain of the stablecoin. Source: same as chart 8.

10. Peer transfers inside Intents

Estimated daily count of Intents transfer events in stablecoins between non-solver accounts, 7-day average. Solver = solver-* name or 5+ token_diff settlements without a referral tag.

What would make this a payments rail

    Investment implication

    Method

    • Tokens: native USDC (17208628…33a1, Circle), native USDT (usdt.tether-token.near, Tether) and the Rainbow Bridge versions USDC.e / USDT.e.
    • Sample: for each UTC day we draw a fixed number of block heights at random (seed 20261114) and download the full block from neardata.xyz. Every successful NEP-141 transfer, mint and burn from these contracts is kept.
    • Estimates: day total = sample total × blocks in day ÷ sampled blocks. 95% intervals: 300 block-level bootstrap resamples.
    • Counterparty type: documented labels first, then a code check (does the account have a contract deployed?). Wallet-to-wallet means neither side has code.
    • Rebuild everything: make all (Python 3 standard library only, no API keys).

    Limitations

    • It is a sample ( of blocks), not a census. Unique-user counts are not estimated because a block sample cannot measure them.
    • Exchange hot wallets are not labeled unless documented; they show up as wallets or contracts. A wallet-to-wallet transfer can still be a trader moving funds.
    • Transfers into NEAR Intents can be cross-chain payments, not only swaps. We count them as trading-venue flow, which is the conservative choice for this thesis.
    • Amounts are taken at $1 per token. Window: .