GridEconTokenomics (@Grid_Econ) header
GridEconTokenomics (@Grid_Econ) profile picture

We Build The Token Model That Lets You Lead Before the Market Forces You To React. Tokenonomics, financial modeling & incentive architecture | Web3 & DePIN.

New York
Joined February 2026
Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)Photo by GridEconTokenomics (@Grid_Econ)
118 tokens launched in 2025. 100 are still down. The founders who avoided the graveyard made these decisions in a specific order, months before their TGE.
1
1
1
479
One thing I find fascinating about AI is that demand is truly unlimited, however that demand has to pass through physical servers which have material needs. As this market starts to mature, we are going to find its not a high margin tech business across the board and operating expenses are going to increase in lock step with demand. Check out the Trust to Traction podcast where we get more into it. youtube.com/watch?v=1y9drB16…
87
Excited to see more of this from both @Zelle and @solana
BREAKING: @Zelle just sent $1.25M in its first confidential transfer using ZelleUSD ($ZLUSD) on Solana, live on stage at Breakpoint. The network that moved $1.2T this year is going global.
20
Bullish Exchange’s November 12 listing of USD.AI’s sUSDai gives a deeper market price to the public record of mid-market GPU credit that sUSDai has always carried. sUSDai funds GPU loans to mid-market compute operators. I see the mid-market as the weakest link in GPU credit. These operators have fewer ways to raise capital than Microsoft or Google. Their revenue also depends more heavily on AI demand. USD.AI describes the compute middle market as a segment with large financing demand and a limited pool of capital. Credit default swaps already price stress in AI infrastructure. That market covers only companies large enough to have traded debt. Mid-market operators fall below that threshold. sUSDai’s loan book, NAV and redemption queue have always lived onchain. USD.AI now reports $531M in GPU financing across 20 deals. The Bullish listing adds institutional buyers and sellers, supported by a dedicated market maker. Their trades price mid-market GPU credit alongside the NAV that USD.AI publishes. I see this as the first layer of a pricing market for compute credit, with sUSDai as an early benchmark.
1
25
Kazakhstan’s central bank sold its first tokenized notes last week. The structure shows a country building tokenized finance in stages. On Nov 9, the National Bank of Kazakhstan sold 50bn tenge of 91-day notes through KASE, the national stock exchange. 10bn tenge of the issue went out as tokens, on terms identical to the traditional notes. The National Bank said the dual format let it test the new issuance mechanism alongside the existing market. Kazakhstan already runs a digital tenge on a permissioned ledger, where the government uses it to program and monitor public spending. This first issue settles through the KASE Clearing Centre. The tokens deliver four things on their own. Holder rights flow into the Central Securities Depository registry, so the notes sit inside the existing legal record. Anyone can check the issued supply on a public explorer. Brokers on the KASE platform can distribute sovereign paper in token form. One issue sold in two formats on identical terms. That gives the market a way to price the token format once secondary trading starts. An onchain tenge adds the next layer: same-day settlement, round-the-clock trading, automated collateral use and automatic redemption. Kazakhstan already runs both pieces, and this issue tests the asset side in a live market. That is a measured step in the right direction.
40
Tune in tomorrow for a deep dive into tokenomics with Trust to Traction. We get into all things tokenomics and how DePIN will support AI demand. youtube.com/watch?v=1y9drB16…
1
90
. @Treasure_DAO TIP-63 asks a question every gaming token eyeing DePIN will face. Can a vote turn an existing token into a work token? The proposal mints up to 173M $MAGIC, about 50% of max supply. 100M lands a month before the network launches. The other 73M unlocks against milestones, several of them active agent counts. A work token earns its supply from paid demand. Agent counts measure presence. Project finance releases capital when revenue capacity is certified, not when headcount arrives. Founders and new protocols should tie emissions and unlocks to revenue. Here most of the new supply clears before the first fee is paid. Voting runs Nov 18 to Dec 2.
1
156
AI is exposing how much licensing depends on clean records. Clearing one song for one commercial takes weeks of emails and legal review. Licensing a catalog to a model means pricing and auditing millions of uses. Manual clearance can’t run at that speed. The music labels have already started. Universal and Warner have settled with Udio and signed licensing deals. Warner has done the same with Suno. Each deal turns a catalog into an ongoing stream of AI usage that someone has to track and pay out. The records underneath haven’t caught up. A song needs one license for the master and another for the composition. In 2022 I was preparing to work with a client putting music rights onchain, and some of those records lived in Excel. Regulation is adding pressure from the buyer side. The EU AI Act requires providers of general-purpose AI models to publish a summary of their training content. That puts the origin of training data on the record. Story Protocol saw this early. It launched its network in 2025 as a blockchain for IP, with music, patents and trademarks all in scope. CEO Andrea Muttoni told CoinDesk the team spread itself too thin. Now renamed DATA (@datafdn), it is moving to Sui with $30 million from the Sui Foundation. Its focus has narrowed to tracking where AI training data comes from. Every rights registry built for AI faces the same design question: who gets paid when a model uses the work? A registry that pays rights holders for licensed usage gives them a reason to bring their catalogs.
2
104
Institutions keep choosing Avalanche. @AvaLabs announced today that the clearing firm for NYSE’s planned tokenized stock venue will keep its books on Avalanche. Since November 1, eligible tourists leaving Japan can receive their consumption tax refunds in JPYC, a yen stablecoin, on Avalanche’s C-Chain. Dubai’s DIFC Courts began anchoring court orders and hearing records to Avalanche in October. In South Korea, Hanwha Investment & Securities built a tokenized securities platform on the chain. A clearing firm cannot let its settlement costs move with a volatile asset. Avalanche built its L1 architecture around that constraint. Validators on an Avalanche L1 pay a continuous AVAX fee that does not scale with transaction volume. Institutions get predictable costs and control over who participates. The same design keeps the token’s share flat as institutional volume grows. Ava Labs’ announcement names no role for AVAX in fees, staking, or settlement. AVAX trades around $8.86 today, about 94% below its $144.96 peak. For every chain courting institutional settlement, solving the client’s cost problem pushes the token to the edge of the cost stack. Institutional adoption raises its own design question for token economies. How can institutional volume create token demand while client costs stay stable?
2
124
. @bittensor’s eTAO turned every subnet’s emission split into a number stakers can price. Owners now choose how new alpha divides between themselves, validators, and miners. The chain caps owners at 33% and validators at 50%, and miners keep a 33% floor. Stakers see each choice before they put TAO into a subnet. Their flows move alpha price, and that price drives each subnet’s share of TAO emissions. TAO holders sit outside the experiment by design. Each split change recalculates the subnet’s effective TAO weight, which keeps root stakers’ baseline share intact. The protocol holds its royalty fixed while 128 operators test their own economics. Vanta (SN8) set its owner share to zero. OxMarkets (SN35) and Epago (SN36) set theirs at 33%. Every weekly reset adds another data point on what stakers will pay an operator to run a market.
51
A token can sit on top of a profitable business and still hold the weakest claim in the capital stack. @VeniceAI says it earns over $150M a year. Tokenomist traced October’s burns on Base and found the VVV token received about a tenth of that. Three questions place any token in the stack when revenue settles in an operating company: 1. Is the burn or buyback a formula enforced on-chain, or a policy decided off-chain? 2. Can supply parameters change without a timelock or cap? 3. What happens to the token if the company is sold? Venice answers all three on the record. The company sets the burn budget, and one discretionary transfer made up more than half of October’s burns. A four of six multisig sets the mint rate with no timelock and no cap in the code. Revenue belongs to Venice.ai, and its Terms of Service never mention VVV. The answers come back as policy, yes, and nothing. A token with that profile is a bet on the issuer’s continued goodwill. A credit analyst would file that as an unsecured, subordinated claim. The market values VVV at roughly $1.1B to $2B. The Series A valued the company itself at $1B. Venice has kept faith so far. It holds about a quarter of VVV supply and has cut the mint rate six times this year. That treasury position is what keeps the company aligned with holders. Each of the three answers is a design choice. Before TGE, changing one costs a negotiation. After TGE, it costs capital.
1
65
Aave's planned buybacks sit behind a 30,000 ETH Mantle facility. Revenue reaches AAVE holders after that liability gets serviced. @aave's founder said on October 22 that Aavenomics 3.0 will include buybacks, staking, and a burn. Three unpublished design choices will decide how those mechanisms affect AAVE. Where repurchased tokens land sets the supply effect. A burn retires supply. A staking pool recycles tokens as rewards. A reserve holds them as inventory. The suspension rule sets how the market values future payouts. Written thresholds for bad debt and reserve coverage would turn a pause into a parameter. The revenue waterfall decides how much reaches holders at all. A buyback that scales up as the Mantle facility amortizes gives holders a schedule they can model.
56
. @akashnet stakers just approved moving AKT to Solana with 97% of the vote. The economics that will govern the token on the other side arrive later. Old-chain inflation and staking rewards end at cutover. The new staking mechanism gets announced before that date. The emissions model goes to a later vote. The path for unvested AKT is still being settled. Akash sequenced the chain, the token standard and custody first. Economics follow. That order moves the decision about holder value to a future gate. Lease revenue runs on its own track. At launch, providers keep 100% of lease payments and gas is paid in SOL. Network demand can grow without reaching the token. Utilities face the same sequence when they reorganize under a new holding company. Shareholders approve the scheme of arrangement first. The new dividend policy comes after. Markets price that gap as uncertainty until the terms are published. The cutover proposal is where AKT’s value gets written. Staking yield, emissions and fee capture all land in that vote. Every DePIN network weighing a chain migration will run into this same order of operations.
52
NEAR’s AI strategy now runs through its issuance schedule. @NEARProtocol started in 2017 as Near.ai, a company teaching machines to write code. The founders pivoted to blockchain and launched mainnet in 2020. Six years later, AI is back at the center of the strategy. The way the network earns has changed with it. Through 2024 and 2025, a few rewards apps drove most of NEAR’s transaction count. The apps paid the gas and paid users in tokens to transact. Kai-ching alone sent about 2 million relayed transactions a day until it stopped in March 2026. Daily transactions fell by about three quarters almost overnight. Revenue now comes from products that earn off the base layer. NEAR Intents routes cross-chain swaps and sends its share of fees to NEAR buybacks. At the trailing 30-day rate, that comes to about $28M a year. NEAR AI lets users stake NEAR to pay for private inference. The staking rewards go to NEAR AI as payment, and the principal stays withdrawable. More than 1M NEAR is now staked this way. Both products move the token without moving the transaction count. NEAR’s staking yield comes from new issuance. Every dollar of NEAR AI credit is redirected inflation. Holders who do not stake absorb that dilution. House of Stake is voting now on HSP-029, which lowers maximum issuance from 2.5% to 1.6% over 24 months. Gross staking yield falls from 5.27% to 3.37% at today’s staking ratio. At a constant price, a staker’s inference budget shrinks by 36%. If stake leaves as yield falls, the remaining stakers earn more. That offsets part of the cut. The proposal’s forum thread runs past 60 posts on validator economics, dilution, and ETF staking carry. None of them connect the cut to NEAR AI’s credit budgets. Issuance still outweighs buybacks about five to one. At 2.5%, NEAR issues roughly 32.7M tokens a year, around $145M at today’s price. For any protocol running a stake-to-access model, the issuance schedule is part of the product’s price. Transaction counts miss that demand. Three numbers capture it: capital staked for access, revenue retained by the protocol, and the share of issuance that revenue covers. On NEAR today, the third number is about one fifth. Every future issuance vote on NEAR is also a pricing decision for its AI.
68
Korea wants to know who is moving dollars out of the country. The Bank of Korea plans to draft an amendment this year that classifies stablecoins as a means of payment under its Foreign Exchange Transactions Act. A USDT transfer abroad would then fall under the same framework as a dollar wire. The ledger already records every stablecoin transfer. FX law adds the sender, the purpose, and the identity behind the wallet. Regulators want attribution. Korean exchanges and custodial platforms would take on the reporting role that banks hold for wires. Unregistered offshore platforms face access blocks instead. Self-custody wallets sit outside both. Attribution extends only as far as the regulated Korean businesses that handle the money. coinedition.com/crypto-live-…
48
My conversation with Stefan Furcoi on the Trust to Traction podcast premieres November 13 on YouTube. We talked about how token value responds to real business inputs like usage, emissions, and treasury decisions. I also brought in what traditional finance and insurance already learned about incentives under stress. Save the date here: youtube.com/watch?v=1y9drB16…
50
S&P just built the assessment rails for onchain credit before the credit arrived. Vault deposits hit $10B in September. Most of it is stablecoins lent against BTC and ETH, where collateral and liquidation speed decide repayment. The Vault Risk Assessment weighs portfolio credit quality, liquidity mismatch and curator judgment. Those factors carry their full weight on loans to borrowers whose creditworthiness drives yields and repayment. Tokenized private credit is that market. Liquidity mismatch has the sharpest edge there. Private credit loans run for months or years, and vault depositors can withdraw at any time. The SEC has rewritten money market fund rules twice since 2014 around the redemption and asset maturity gap. I expect private credit to be the next asset class to receive SEC exemptive relief. The assessment layer for the vehicles that would hold it is now in place. press.spglobal.com/2026-10-0…
1
52
I have been thinking about the AI bubble for the past couple of months. Demand has no visible ceiling, and no one yet knows what the capital behind the supply will earn. Here is what I found when I followed the financing, the grid, and the filings.
1
1
1
506
Onchain credit is entering its underwriting era. GSR is putting $100M behind Hare, a new vault business built with Turtle. Hare's focus is credit. It assesses collateral, counterparties, and how positions hold up in a stressed market. After 2008, Dodd-Frank required securitizers to retain 5% of the credit risk they sold. The rule tied the originator's own capital to the quality of its underwriting. GSR's own capital sits in the same vaults as allocator capital. Onchain, that alignment is visible in the vault itself. Tokenization already made the collateral transparent. Credit quality has been the open question. Capital is now moving toward the firms built to answer it. coindesk.com/business/2026/1…
58
. @openstandard is running a token emissions playbook with equity. The company plans to distribute most of its cap table to partners over four to five years. Each allocation tracks the OUSD supply and transaction activity a partner generates. Every allocation calculation maps how money moves through the network. That map is the most valuable output of the growth phase. The revenue and the rewards track different variables. Reserve interest scales with balances, and the rewards pay partly for movement. Phase two depends on a demonstrated link between paid activity and retained float. Founders launching without a fully defined model need two commitments. Set a public end date for the bootstrap phase, as OUSD has. Publish the metrics that will shape the next phase. Published metrics remove the discount markets apply to an open-ended experiment.
51