Here is a sentence I keep repeating to myself whenever a new onchain product shows up: the first week is noise, the first month is the tell. Morpho just opened USDC lending against five Coinbase tokenized stocks on Base, and the early tape already has that odd mix of tiny balances and sudden jumps. Holders have pledged $104,401 in stock tokens and borrowed $54,652 in USDC so far. That is not a flood. It is also not nothing.
I have watched a lot of “real world assets meet DeFi” headlines fade after the press cycle. This one feels different, not because the dollars are huge, but because the plumbing is unusually complete. Price feeds exist. Custody language exists. Rate markets exist. Liquidation math exists. Someone can keep Apple exposure and still pull dollars without selling the position. That is the whole pitch in one line.
Why Stock Backed Usdc Loans Matter Now
Tokenized equities used to sit in a strange corner of crypto. You could look at them. Sometimes you could trade them. You could rarely do anything useful with them. Lending changes that. Once a token can sit as collateral, it stops being a novelty ticker and starts acting like a balance sheet item.
The five live names are Apple as AAPLc, Alphabet as GOOGLc, Nvidia as NVDAc, Meta Platforms as METAc, and SpaceX as SPCXc. Coinbase has ten stock tokens on Base. Amazon, Microsoft, Strategy, SanDisk, and Tesla still have no Morpho markets. That gap is not a footnote. It tells you the rollout is staged, curated, and slower than the marketing tone around tokenization usually suggests.
Holders can keep the stock token and still reach for dollars. That sounds simple. In markets, simple products are the ones that last.
I’ve found that people overestimate launch day and underestimate week three. These markets went live around September 7. Outstanding loans stayed under $600 for days. Then, on September 16, borrowing leapt from $503 to $42,105 in about two hours. By Friday morning it sat at $54,652. That spike is the first real human signal in the data. Somebody wanted liquidity and did not want to sell.
How The Coinbase Stock Tokens Are Built
The first four products, Apple, Nvidia, Meta, and Alphabet, are described as one for one interests in underlying shares held in segregated custody. They are not cartoon synthetics that merely track a price. They represent beneficial interests in securities. That distinction matters when a lending protocol starts treating them like collateral instead of a meme with a chart.
Coinbase later added six more names in September, including Amazon, Microsoft, Strategy, SanDisk, SpaceX, and Tesla. Only SpaceX from that second wave currently sits inside Morpho. In my experience, second-wave assets often wait for oracles, legal packaging, and curator comfort before anyone opens a borrow market. That seems to be the pattern here.
The issuer is an entity set up in the Abu Dhabi Global Market. The products are aimed at eligible investors outside the United States and are not registered under U.S. securities law. That is the legal fence around the whole experiment. You can debate the fence. You cannot pretend it is not there.
Market contracts on Morpho do not themselves police geography. Compliance sits in the tokens and their issuer controls. The markets are unavailable to U.S. persons and people in other restricted places. If you skip that sentence, the rest of the product story becomes a fantasy.
What Morpho Actually Turned On
Morpho now lets eligible holders post those five stock tokens and borrow USDC through variable rate pools or through Midnight, the protocol’s fixed rate layer. Steakhouse Financial curates all five markets and sets parameters such as collateral rules. That curator role is easy to ignore until something breaks. Then everyone suddenly cares who chose the numbers.
Two Steakhouse High Yield USDC vaults supply 98.9% of the dollars in the largest market. For Apple, those vaults put in $24,540 of the $24,805 deposited. Another vault contributes $262. Two wallets add less than $3 combined. That is concentration. Concentration is fine until it is not.
- Five stock tokens accepted as collateral on Base
- USDC available on variable rate markets and Midnight fixed rate markets
- Steakhouse Financial setting market parameters
- Chainlink based pricing through Morpho’s V2 oracle adapter
- No U.S. person access by design of the token layer
Perhaps the most interesting aspect is how ordinary the product looks once you strip away the branding. Collateral in. Stablecoin out. Health factor in the middle. Crypto has spent years trying to make that loop feel boring. This is one of the first times listed-name equity tokens are sitting inside that loop with real price feeds.
Liquidation Limits Are Not The Same For Every Name
Apple, Nvidia, Meta, and SpaceX markets use a 62.5% liquidation loan to value ratio. Alphabet sits higher at 77%. That is a meaningful gap. A higher threshold means more borrowing power and less room when the token wobbles. I would not treat those two settings as interchangeable, even if the interface makes them look like siblings.
On Apple, the market applies a 12.67% liquidation penalty at that 62.5% line. No liquidation has been recorded since deployment. The largest borrower holds 113.46 AAPLc against a 20,360 USDC loan with a health factor of 1.17. That is not a panicked position. It is also not a fortress. One ugly gap versus the feed and the story changes.
| Token | Liquidation LTV | Current Borrow Picture |
| AAPLc | 62.5% | Active variable rate use, no liquidation yet |
| NVDAc | 62.5% | High utilization near model target |
| METAc | 62.5% | Utilization above target, much higher rate |
| GOOGLc | 77% | More borrow room, tighter shock buffer |
| SPCXc | 62.5% | Included despite later token batch |
Chainlink first launched feeds for Apple, Nvidia, Meta, and Alphabet after the Base debut. Each feed aims at the total return value of the token, not a lazy last-trade print. Morpho now prices collateral through its Chainlink V2 oracle adapter. On Friday the Apple feed showed $335.49 at 1:42 p.m. UTC, while one dashboard printed the token near $337.52. Tiny gaps are normal. Tiny gaps plus leverage are how liquidations get born.
Variable Rates Took All The Action
This is the part that surprised me. Midnight offers fixed rate, fixed term books. For each of the five tokens there are 19 USDC markets, or 95 fixed rate markets in total. Maturities run daily through September 30, then hop to October 30, November 27, December 25, and March 26, 2027, plus an open ended book per token. Outstanding units in those 95 markets: zero.
Every dollar borrowed so far sits in variable rate pools. Apple, Alphabet, and Nvidia are running at 90% utilization, matching the interest model target. Borrowers pay 5.62%. Lenders earn 5.06%. Meta is at 97% utilization, past the target and onto the steep part of the curve. Borrowing there costs 17.52%. Lenders earn 16.98%. That is a completely different product wearing the same interface.
Why skip fixed rates? Maybe size. Maybe habit. Maybe nobody wanted to lock a term when the whole market is still finding its first $55,000. I lean toward habit. Variable pools are familiar. Midnight is newer, even if the idea of a locked rate is older than crypto itself.
All current stock backed borrowing is sitting in variable rate markets. The fixed rate books are built, labeled, and empty.
Morpho launched Midnight on Base in July and later expanded it. The design lets borrowers and lenders set conditions without living entirely on a floating curve. On paper that should appeal to anyone borrowing against a stock they do not want to sell. In practice, the empty books say users still prefer the pool they already understand.
The Market Is Still Tiny Next To Trading
About $11.4 million of the five supported stock tokens is outstanding on Base if you combine contract supply and Morpho oracle prices. The $104,401 sitting as Morpho collateral is less than 1% of that stack. Lending is a side room. Trading is the main hall.
Decentralized exchange volume across Base tokenized stocks hit $730.9 million over the 30 days ending September 12. Daily volume printed a record $100 million. Aerodrome handled $557.1 million in the measured window. Another major venue processed $139.3 million. People will trade a tokenized stock long before they pledge it. That is human. Trading is a click. A loan is a relationship with a liquidation engine.
Morpho itself is not a small venue. Deposits on Base sit around $4.03 billion, with $10.42 billion across supported chains. Base also lists other large lending protocols as places where eligible holders can borrow against these stock tokens. Competition is already in the room. That is healthy. It also means rate and risk parameters will get compared in public, quickly.
MORPHO last printed near $2.40, up 11% over 24 hours, with a market value around $1.68 billion. Bitcoin was up 4.9% over the same stretch. Ether was up 4.3%. I would not read the token bounce as proof the stock loan books will scale. Tokens rally on attention. Loan books scale on repeat borrowers who do not get liquidated.
What Borrowers Are Actually Trying To Do
The clean use case is cash without a sale. You like the equity. You need dollars. Selling creates tax questions, timing questions, and the risk of buying back higher. A loan keeps the token in the wallet and puts USDC in hand. If the stock rips, you still have it. If the stock slumps, the protocol does not care about your thesis. It cares about the health factor.
- Keep the tokenized share exposure instead of selling into the market.
- Borrow USDC for spending, hedging, or another onchain position.
- Watch utilization, because Meta already showed how fast the rate can jump.
- Leave a buffer above liquidation LTV instead of hugging the line.
- Remember the token itself still carries issuer and access constraints.
There is a sloppier use case too. Someone can treat the stock token like cheap leverage on a name they already like. That works until correlation, oracle lag, or a weekend gap shows up. I have a bias here. Borrowing to avoid a sale feels more durable than borrowing to press a bet. Durable products attract quiet size. Levered bets attract screenshots.
Risks That Do Not Fit In A Launch Thread
Start with collateral quality. These tokens are designed around real share interests, which is better than a synthetic tracker with no claim. Better is not the same as simple. You still have issuer structure, custody arrangements, eligibility rules, and the chance that an application refuses a wallet. DeFi likes to pretend assets are just addresses. These assets are addresses plus paperwork.
Then comes oracle risk. A feed can be honest and still disagree with a spot print for a few minutes. At 62.5% LTV, minutes matter. At 77%, minutes matter more. No liquidation so far is a nice fact. It is not a guarantee. The sample is young and the borrowed stack is small.
Liquidity risk sits under that. If you need to sell the stock token after a partial liquidation, you are selling into a market that has already proven it can print huge volume and still remain a specialist venue. Depth is not the same at 2 p.m. and 2 a.m. Tokenized stocks do not inherit the full cash-equity tape just because the ticker looks familiar.
Rate risk is the sleeper. Three markets sit on the target utilization line. Meta already overshot and paid for it. A few extra borrowers can shove a curve into double digits. That is fine if you planned for it. It is ugly if you treated 5.62% like a promise instead of a snapshot.
And yes, there is smart contract and curator risk. One firm setting parameters for all five books is efficient. It is also a single point of judgment. Efficient systems fail in clusters. I do not say that to be dramatic. I say it because every lending market I have trusted eventually taught me where the clustering was.
Why Five Markets And Not Ten
If you only read headlines, ten tokenized stocks should have meant ten loan markets. That is not how cautious curators work. Feeds, legal wrapping, borrower demand, and vault appetite all have to line up. SpaceX made the cut from the later batch. Amazon, Microsoft, Strategy, SanDisk, and Tesla did not. I read that as a filter, not a snub.
There is also a branding problem hiding in the product set. Some names are household. Some names are crypto-adjacent. Mixing them in one interface can trick a borrower into using one risk setting for every ticker. Alphabet’s 77% line already breaks that lazy habit. Good. More differences would be even better.
Will the missing five show up? Probably, if volume stays loud and oracles stay clean. I would rather they arrive late than arrive sloppy. Tokenization does not fail because it is slow. It fails because someone ships a loan market before the collateral is boring enough to survive a bad Tuesday.
Where This Fits In The Tokenization Story
For years the pitch was simple. Put stocks onchain and the world will come. The world did not come. Traders came. Now lenders are peeking through the door. That sequence feels right to me. First you need a token people will hold. Then you need a venue where holding it does more than sit in a wallet looking pretty.
Self custody is part of the lure. The tokens are meant to live in wallets and still talk to supported apps on Base. That is the cultural bridge. Traditional brokerage margin lives inside an account you do not fully control. This version lives closer to the asset. Closer is not safer by default. It is just a different control map.
The restricted-access design will annoy maximalists. It should. Open lending against closed eligibility is a contradiction you can feel. It is also how this particular structure was able to exist at all. I would rather see a gated real claim than an ungated fake one. That is a personal preference, and I will own it.
A Practical Read On The First Numbers
Do not romanticize $54,652. Do not dismiss it either. The dead period after September 7 said curiosity was low. The two-hour jump on September 16 said at least one actor came in with intent. The later grind to $54,652 said a few more followed. That is how loan books start. They do not start as press releases. They start as a handful of positions that survive the first week.
Early Morpho stock-loan snapshot: Collateral pledged: $104,401 USDC borrowed: $54,652 Share of token supply posted: under 1% Fixed rate books in use: 0 of 95 Liquidations recorded: none yet
Utilization is the number I would watch next. Apple, Alphabet, and Nvidia hugging 90% means the rate model is doing what it was asked to do. Meta at 97% means the model is already teaching a lesson. If more collateral arrives without more USDC supply, borrowers will pay up. If vaults keep dominating supply, lenders will stay concentrated. Both facts can be true at once.
Who This Product Is For And Who Should Sit Out
It is for an eligible non-U.S. holder who already wants the equity and needs dollars more than they need a clean exit. It is for someone who can live with onchain liquidation math and will not treat a health factor of 1.17 like a sofa. It is for a lender who understands that the yield is coming from stock-backed demand, not from a mysterious surplus.
It is not for anyone who thinks a familiar ticker erases crypto operational risk. It is not for a U.S. person hoping the interface forgot the rules. It is not for a trader who only discovered the token this morning and now wants maximum LTV because the chart looks friendly. Friendly charts are how people meet liquidation bots.
I’ve found that the best borrowers in new collateral markets are almost boring. They size small. They leave headroom. They care more about not selling than about squeezing the last unit of leverage. If this market grows, those are the people who will grow it.
What I Would Watch Over The Next Month
First, whether Tesla, Amazon, and Microsoft get books of their own. Second, whether any Midnight market finally prints a loan. Third, whether supply stays locked inside two vaults. Fourth, whether a first liquidation arrives on a quiet day or a loud one. Fifth, whether borrowed USDC gets used as dry powder or as fuel for the same names on the same chain.
There is also a softer signal. If collateral posted climbs while borrowed dollars stay flat, holders are parking tokens without needing cash. If borrowed dollars climb faster than collateral, someone is pressing the LTV. The second path is more exciting. The first path is healthier.
I keep coming back to that September 16 burst. Markets talk when they jump. They also talk when they go quiet afterward. A grind from $42,105 to $54,652 is not euphoria. It is a book trying to decide if it is real. Give it time. Then look again at utilization, empty fixed rate shelves, and whether the missing five names were delayed for a good reason.
Tokenized stocks do not become financial infrastructure because a protocol flipped five markets on. They become infrastructure when a holder can borrow, repay, borrow again, and never feel the need to explain the product to a friend like it is a magic trick. We are not there. We are, finally, closer than the last dozen headlines suggested.
If you hold one of the five tokens and you are allowed to use these markets, the decision is not “is tokenization the future.” The decision is narrower. Do you need USDC more than you need unused borrow room? Can you respect a 62.5% line, or a 77% line, when the feed and the token print disagree for an hour? Answer those without theater and the rest of the story gets easier to read.
That is the unglamorous version. I prefer it. The glamorous version will be written when the numbers are bigger. The useful version is the one that admits $104,401 in collateral is a start, 95 silent fixed rate markets are a warning, and a two-hour borrowing spike is the first proof that somebody, somewhere, wanted dollars without selling the stock.
