A third of all sUSDe is levered. That is demand talking. Looping is how DeFi holds @ethena's yield: post collateral, borrow, buy more, repeat. Yield bearing dollars currently sit at $15B (down from $21B peak) onchain, and looping is the growth engine behind the biggest ones. sUSDe's looped share doubled in a year to ~30%. So the real question is not whether to loop Ethena's carry. It is what collateral you loop. This is what @LidoFinance EarnUSD's new allocation answers. > @strata_markets splits sUSDe into two claims. $srUSDe gets paid first, floored at the @aave benchmark. jrUSDe takes first loss and currently earns about 10% for it. About $7M of junior capital sits under $60.5M of senior today. > @pendle_fi fixes the rate. PT-sUSDe maxes at 13.33%. PT-srUSDe at 13.01%. Thirty basis points is the full price of $7M in first-loss capital standing in front of you. > @aave prices the protection. Same risk methodology, same day, same maturity: PT-srUSDe gets 91.16% LTV against PT-sUSDe's 90.45%, because first-loss capital tightens the worst case from 12.27% to 10.22%. Protection converts directly into borrowing power: 16.1% looped, and $7M stands in front of you. > @twynexyz unlocks the ceiling. Aave liquidates at 93.16%. Twyne lets you choose up to 98%, funded by PT holders lending unused borrowing capacity. Same 2 point buffer, 11.3x becomes 24.4x, 16.1% becomes 23.3%. And the senior buys something the raw PT cannot: eligibility. @twynexyz's boosted channel only exists on PT-srUSDe, because first-loss capital is what makes a 98% liquidation threshold underwritable. Four layers, four jobs. Strata prices the risk and provides the first loss cushion. Pendle fixes the rate. Aave supplies the leverage. Twyne unlocks the idle capacity. The market already decided loops are how Ethena yield gets held. Lido just showed the institutional way to hold the loop: senior, fixed, protected. Fixed rate on top. First-loss capital underneath. Best risk-adjusted yield. Note: leverage figures are Aave E-mode maxima. Pendle's one-click loop tool deliberately caps lower (~7.5x, wider liquidation buffer). Twyne extends the ceiling to 24.4x via delegated credit.

8.5% APY Senior 15% APY Junior On a product with almost a full year of onchain track record and no down days. Pretty compelling no? Managed by @hyperithm, tokenised by @MidasRWA and tranched on @strata_markets

Privacy is making a huge comeback, and the reason is simple yet very powerful: everyone needs it. Retail wants it for sovereignty. As surveillance tightens, a transparent ledger stops being a feature and turns into a liability. That's the Zcash run, shielded by default, a Grayscale ETF filing, Paradigm and a16z writing checks. It topped Monero as the biggest privacy coin in May, then gave some back. The tech is still maturing. The direction isn't. Institutions want it for a colder reason: signaling. Every onchain position is copyable the second you submit it. For a desk with size, transparency isn't openness, it's a tax. You're handing your book to every competitor and front-runner on the chain. Last cycle's answer was binary. Mix it, cut the link, disappear. Tornado Cash. That era basically ended in a courtroom. The contract got de-sanctioned because code isn't property, but the developer still caught a conviction and a retrial. The lesson stuck: anonymity with no selective disclosure is a dead end for real capital. This wave is different. The industry traded privacy ideology for privacy architecture. FHE, ZK and MPC don't hide you from everyone. They hide you from the market while staying auditable to whoever holds the keys. Private to competitors, reportable to regulators. @GSR_io and @zama already cleared the first confidential OTC trade between KYC'd desks. The Ethereum Foundation stood up a dedicated privacy cluster, institutional task force attached. Coinbase is calling it institutional demand. Which brings it back to @Morpho. @zama's cUSDC keeps your position encrypted through FHE, then routes into the same @SteakhouseFi Prime vault. Same collateral, same markets, same risk curve. Nothing about the credit changed. Your size, direction and timing go dark to the market, the underlying supply stays auditable. Not a mixer. Selective disclosure built into the standard. Very simple understanding of what this actually is: not a new product, a new deposit surface. Steakhouse didn't build a confidential strategy, they bolted a confidential entry point onto their flagship vault. The yield was always there. Privacy changed who can reach it without leaking. That's the Morpho thesis. It didn't build the encryption, it didn't build the strategy. It stayed the neutral credit layer Zama plugs privacy into and Steakhouse plugs curation onto. 🦋 The base stays minimal and transparent, everything composes on top. Morpho's already framing it as the first of many experiments. The moat is not the mechanism. It's distribution. Privacy just opened a new lane into the same credit.

Privacy is making a huge comeback, and the reason is simple yet very powerful: everyone needs it. Retail wants it for sovereignty. As surveillance tightens, a transparent ledger stops being a feature and turns into a liability. That's the Zcash run, shielded by default, a Grayscale ETF filing, Paradigm and a16z writing checks. It topped Monero as the biggest privacy coin in May, then gave some back. The tech is still maturing. The direction isn't. Institutions want it for a colder reason: signaling. Every onchain position is copyable the second you submit it. For a desk with size, transparency isn't openness, it's a tax. You're handing your book to every competitor and front-runner on the chain. Last cycle's answer was binary. Mix it, cut the link, disappear. Tornado Cash. That era basically ended in a courtroom. The contract got de-sanctioned because code isn't property, but the developer still caught a conviction and a retrial. The lesson stuck: anonymity with no selective disclosure is a dead end for real capital. This wave is different. The industry traded privacy ideology for privacy architecture. FHE, ZK and MPC don't hide you from everyone. They hide you from the market while staying auditable to whoever holds the keys. Private to competitors, reportable to regulators. @GSR_io and @zama already cleared the first confidential OTC trade between KYC'd desks. The Ethereum Foundation stood up a dedicated privacy cluster, institutional task force attached. Coinbase is calling it institutional demand. Which brings it back to @Morpho. @zama's cUSDC keeps your position encrypted through FHE, then routes into the same @SteakhouseFi Prime vault. Same collateral, same markets, same risk curve. Nothing about the credit changed. Your size, direction and timing go dark to the market, the underlying supply stays auditable. Not a mixer. Selective disclosure built into the standard. Very simple understanding of what this actually is: not a new product, a new deposit surface. Steakhouse didn't build a confidential strategy, they bolted a confidential entry point onto their flagship vault. The yield was always there. Privacy changed who can reach it without leaking. That's the Morpho thesis. It didn't build the encryption, it didn't build the strategy. It stayed the neutral credit layer Zama plugs privacy into and Steakhouse plugs curation onto. 🦋 The base stays minimal and transparent, everything composes on top. Morpho's already framing it as the first of many experiments. The moat is not the mechanism. It's distribution. Privacy just opened a new lane into the same credit.

I've been keeping a close eye on the fed and macro for weeks, and all the suits are locked onto one thing: Wednesday, 2pm et, Kevin Warsh's first meeting as chair. If you're building in defi, you should be watching closer than they are. Not for the rate, that's over 99% hold, a non-event. Watch it because warsh will play key role on how much liquidity flows back into defi and onchain credit over the next two quarters. The fed funds rate is the anchor under every yield in this market. Where he points it, our whole curve follows. Here's what most people seem to be mispricing. They've decided warsh is trapped. - Inflation at 4.2%, a three-year high. - Jobs running 140k a month this year against 10k last. - A committee turned against him: april's hold passed 8 to 4, the most fractured fed vote in 34 years. The dovish chair handed a hawk's economy. But headline is 4.2% and core, which strips out food and energy, is 2.9%, up just 0.2% last month. - The inflation isn't broad. it's oil. - Over 60% of the spring spike was energy alone. - And oil just broke: the iran truce landed sunday, brent is back at 83, a three-month low. We have kind of seen this already. - 2022, oil rips to 120 after ukraine - CPI peaks at 9.1%, the fed hikes into it. - Then oil rolls over, inflation halves in a year, no wage spiral. - They fought a spike that was already dying. That's warsh today, a hot headline about to fade. Watch one number at 2pm: the 2026 dot. it sat at 3.4% in march, one cut below today. if that cut survives, he's holding his dovish lean against the data. if it drops, the hawks took the room. And he knows the second move. - Lower the short end, shrink the 6.7 trillion bond book, - Exempt treasuries from the slr so banks can hold them without the capital hit, and let them absorb what the fed sheds - Easing that prints as tightening. We ran a version of this in oct 2023: treasury shifted issuance short, liquidity came back through the side door, risk ran to new highs. nobody called it QE. Now bring this onchain. - When the short-term rates grinds lower, the risk-free leg under every onchain yield compresses with it. - T-bill-backed rwa and stablecoin yields fall, and capital does what it always does when the safe rate drops. - It climbs the curve, into onchain credit, into structured product, into the institutional-defi stack. - That's the adoption tailwind A hike does the reverse: the safe leg widens, competes directly with our yields, and capital stays home. The rate is priced. The path is the trade. The path is liquidity. And the liquidity flows onchain. Nobody stays hawkish through a falling oil price. They just find out last.

i’ve been fortunate enough to be mentored by some of the most experienced designers in tech early in my career. (thanks to the lads at @lifiprotocol & @ParagonsDAO) these learnings never made me compromise on the creative trust that i owed to every founder and team i worked with. for the last 3 years, every project came through referrals from clients and friends (ilu), so i never really felt the need to launch. but in a post ai era, i believe that peer-to-peer learning would be one of the most important aspects of growing as a designer or even as a studio. we've spent the last year experimenting obsessively with ai tools, workflows, storytelling systems, and new ways of creating. so it feels like the right time to share learnings and build in public. launching @dapptales publicly - a creative studio specialising in tasteful storytelling for tech companies. and here's a little video about what we're best at :))

This post is spot on. Everyone is converging on the same architecture which blurs the lines of "stablecoin" and "lending market". Sky has been operating this model for a while, but it's the most flexible design so defi protocols are arriving from slightly different angles.

DeFi and TradFi are collapsing into one capital-allocation layer, and the biggest venues in finance are folding into a handful of onchain allocators who hold the deepest liquidity like @SkyEcosystem, @aave, @ethena. Every other structured product will be the foundational layer. None of them is just a stablecoin or a lending market anymore. They're becoming crypto's wholesale banking layer. Whoever assembles the deepest liquidity, the sharpest allocators, and real distribution into one system wins. Two ways I see teams building it. 1) modular and decentralized: a central balance sheet issues the liquidity and sets the risk envelope, while specialized, externally-run desks plug in, each with its own mandate and limits, all drawing on the shared book. A universal bank. One balance sheet, many desks. @SkyEcosystem shipped this first. @aave is bringing it to the deepest liquidity in DeFi. 2) centralized then converging: one team allocates the whole book itself, then externalizes as it scales. @ethena. Sky has run the model for ~18 months. Sky Core issues USDS and owns the risk framework. Sky Agents borrow USDS up to a governance debt ceiling, deploy their own strategy, and compete on risk-adjusted return that funds the Savings Rate. - @sparkdotfi owns DeFi yield (~$12B). - @grovedotfinance owns institutional credit (~$2B). - @OseroHQ own fintech/ institutional plugins - 4) Onchian Prime Brokerage via Spark Prime + @ArkisXYZ - Now expanded to other leaders like @maplefinance, @Securitize, @centrifuge, @daylightenergy, @USDai_Official, @RiverFND via @obexincubator Each the best in its lane, all on one balance sheet inside one risk envelope. The allocator OS Maker pioneered, $9.2B USDS later. @aave V4 shipped the same architecture with V4. A Liquidity Hub holds the capital and the accounting. Spokes draw a credit line against it and set their own collateral, risk, and liquidation rules. - A team launching a spoke inherits Aave's liquidity on day one instead of bootstrapping deposits. - Think of it as a supranational bank allocating capital to regional facilities. @StaniKulechov is right that it's powerful. Hub = Core. Spoke = Agent. A credit line is a debt ceiling. @ethena is the centralized path, mid-pivot. Perps are down to ~11% of USDe's backing; the rest is institutional lending and CLOs. - It still allocates from the center, but it's already routing USDe into @sparkdotfi's Liquidity Layer and tranching the yield through @strata_money on top. - Centralized today, but looks more like converging on the same hub-and-spoke tomorrow. Here's what's actually flowing through all of it. Crypto-native yield compressed, so the real spread now comes from real-economy credit, structured and distributed onchain. The biggest names in TradFi credit are already here: - @JHIAdvisors ($480B) put its ~$27B AAA CLO ETF onchain as JAAA via @centrifuge. It now sits in both Sky's and Ethena's reserves. - @apolloglobal ($785B) tokenized its diversified credit fund as ACRED via @Securitize, with Coinbase Asset Management and Kraken among the buyers. - @HamiltonLane ($956B) put its senior credit fund onchain via @Securitize, live across five chains. - @Figure has originated $20B+ of private credit, the largest non-bank HELOC lender in the US, now public on Nasdaq. - @galaxyhq structured a ~$50M onchain CLO with @grovedotfinance. - @3janexyz flips the direction: onchain dollars funding US fintech lenders. A $10M warehouse with LendSwift, ~$8.5M of Slope's SMB receivables, and a $50M forward-flow line, all through bankruptcy-remote SPVs. Securitization, rebuilt onchain. And it's early. As per, @RWA_xyz, tokenized RWA is ~$31B, onchain private credit at $5.5B, against a $3T TradFi private credit market. The pipe is tiny. The flow just started. The hub is the bank. The agents are the desks. The inventory those desks now trade is Wall Street's credit. Sky built the first one. Aave is building it. Ethena is following soon.

This revenue arc is 9 years of shipping relentlessly, sometimes being wrong, fixing it, winning, and compounding. 1) Started with ETHLend in 2017. P2P loan matching on Ethereum, $16.2M ICO. Right idea, kinda dead structure. Order books need two sides, and lending had none. Most requests just sat there. 2) So they killed their own product and started working towards aave thru the bear of 2018-2019. In 2020 rebrand to @aave, pivot to pooled liquidity. Best decision in the protocol's history. 3) Jan 2020, V1. aTokens + flash loans. IIRC it was a primitive that existed in no financial system before, invented here. TVL went zero to $100M+ in six months. Then DeFi summer made pool THE model. 4) Dec 2020, V2. Collateral swaps, credit delegation. Revenue: $177K → $252M in one year. TVL past $20B by late 2021. Nothing in fintech compounds like this. 4) Jan 2022, Arc. Everyone forgets Aave already tried "institutional DeFi" with permissioned V2 pools, @FireblocksHQ whitelisting about 30 institutions. 5) It flopped tho. Ig 3 main problems ( 2 side gating, limited QC support, not actual demand). Also celsius was on the whitelist, lol. Maybe.. compliance belongs at the asset level, not the venue level. Hold that thought. 6) March 2022, V3. Isolation mode, caps, e-mode. The CRV bad debt got hard-coded into risk architecture. This was insane ngl and till date v3 holds aave most tvl. 7) Then deployed to 21 chains. 8) Bear years. Revenue $137M, then $105M. Many wrote obituaries. They shipped hard. 9) July 2023, $GHO. Aave's own dollar. Rough start. Spent its first months under peg. GSM + rate discipline clawed it back. 10) Slow, unglamorous, worked. And is working pretty goood for aave. 11) Why GHO is the only number that matters long term: on the core book Aave keeps the reserve factor, 10-20%. On GHO it keeps ~100% of the stability fee. Owning your dollar vs renting Circle's. 12) 2024-25, the comeback. - Revenue $456M to $907M. - TVL $11B to $30.25B all-time peak. 13) Aug 2025, Horizon. Arc take two, done beautifully. - Compliance at the issuer level. - @Securitize, @vaneck_us, @WisdomTreePrime bringing RWA collateral, NAV feeds via @chainlink + @LlamaRisk. - Largest RWA lending market onchain, ~$580M heading to $1B. 13) Umbrella. The safety module rebuilt as a standing default fund. Staked aTokens absorbing bad debt by rule, not by vote. 14) Feb 2026: first protocol in history to cross $1 TRILLION in cumulative lending volume. 15) Then April 2026 tested everything. - Kelp/LZ exploit dropped ~$200M of bad debt on the book. ~$8B exited in days. - Umbrella absorbed, the ecosystem came together @LidoFinance, @ether_fi, @ethena, @StaniKulechov himself. Put $300M+ behind the rest. Depositors near whole. Everybody wants aave to win. Defi wins. 17) March 2026, V4 on mainnet. - Three hubs, eleven spokes. Unified liquidity, isolated risk, GHO mintable from any spoke. 18) GHO grew a second engine (should check out @josefabregab's recent posts). - GSM reserves earning even when nobody borrows. - In March, GSM fees overtook borrow fees — >50% of GHO's monthly revenue. also h/t @Token_Logic. - sGHO at 4.25%, Aave App taking it to people who'll never know what a health factor is. Aave is winning. DeFi is winning.

$175M, def one of the largest raise defi has ever seen. The idea started small but landed pretty strong. 1) Morpho optimizers, a p2p matching layer on top of aave and compound, optimizing better rates out of pools that already existed. Clever, but boxed in by someone else's design. 2) So they built their own. Morpho blue, one minimal, immutable, permissionless primitive for lending. 3) That's what kicked off the broader modular lending stack. 4) Anyone could spin up isolated markets and build on top. 5) The curators turned it into an industry. @gauntlet_xyz, @SteakhouseFi alone manage/ curate over $1B each. Now OG teams like @wintermute_t also joining in with armitage and hundreds more, running hundreds of vaults, owning the risk and providing results. 6) This is the stage that really pulled tradfi in. The "beauty of vaults". 7) This is the defi distribution era now, the defi mullet. tradfi ux up front, morpho's noncustodial rails underneath. 8) @coinbase earn and btc loans, steakhouse curating. @krakenfx btc vaults. trezor stablecoin earn. 9) Can find your way to morpho markes via almost all leading venues, from @TrustWallet to @utila_io. 10) @deel paying contractors in latam. and with the clarity and genius acts sorting what's compliant onchain, the mullet stops being a workaround and becomes the default. 11) $11B+ in deposits. Wait not done yet. 12) Tradfi assets came back the other way, onchain as collateral. @Figure ( a @Nasdaq listed lender), putting home equity loans on morpho and vaults curated via @SentoraHQ. apollo's acred credit fund. eventually everything from s&p 500, treasuries. 13) @gauntlet_xyz now running rwa markets, the fastest-growing market iirc. 14) I also heard the @Vault__Summit on the @NYSE floor was dope af. Still not done. 15) Now the coolest unlock, midnight. Fixed rate, fixed term, the thing every institution always asked for, because you can't fund a 5-year plan on a rate that moves daily. 16) Capital stays productive on blue until the offer's taken, so liquidity doesn't fragment. minimal primitive, curators build, distribution scales. 17) Unlocking crazy new primitives and upgrading the existing inefficient tradfi ones. 18) The mega modular playbook for credit. This is the whole thesis. (And i didn't even get into v1/ v2, tokenized private credit quietly becoming the biggest rwa market on morpho, compliant KYC-gated vaults via @KeyringNetwork, and a lot more). 19) Three years ago it started off more like a "lending primitive." Now it's the open credit network. The $200T market. 20) Stablecoins became the money rails. Morpho's becoming the credit rails. Banks don't get replaced, they get re-plumbed. Congrats to the entire morpho team @PaulFrambot, @MerlinEgalite, @talkintokens, @maccanomics, @Crotts__

attended my second @Vault__Summit this year (cannes, NYC) at a legendary venue: the NYSE 🇺🇸 key takeaways: > regulation is the product; SEC & CFTC news shapes every product decision in real time > vaults have found PMF but regulatory clarity is still murky; "vault" could mean very different things depending on the use-case > billion dollar question: is curation equivalent to traditional asset management? > lots of discussions around risk management, lack of insurance infra and non-custodiality > non-correlated yield sources in high demand; asset-backed financing, reinsurance quietly gaining traction > there's only a handful of OG crypto teams left; massive consolidation happening > people are genuinely excited about fixed-term, fixed-rate loans; will see how morpho midnight and related products play out > some crypto VCs still alive; though mostly deploying into non-crypto stuff > some crypto people tried to attend in shorts and got turned away; degen vibes are officially dead > RWAs & tokenization are the most interesting things happening onchain rn; @Securitize , @plumenetwork and @OndoFinance leading the pack > ton of work left to make RWA assets composable with rest of defi shoutout to leading teams like @Morpho , @gauntlet_xyz , @kpk_io, @LlamaRisk who are leading the onchain asset management revolution crypto is dead. long live crypto.

Last week @Morpho dropped the midnight whitepaper + open-sourced the code, so i've been sitting with it and talking to a few builders. Something clicked that i don't see many people mapping. First, a recap on morpho midnight. On a high level it's a primitive. think what blue was for variable rates, midnight is that for fixed. It sits on top of deep variable markets (blue) and turns a loan into a tradable, fixed-term unit. Not a consumer product, it's infra. Minimal primitive, everyone builds on top. Classic morpho. So the same way blue needed them, midnight needs curators, solver networks, fintech/ux integration to actually scale. The obvious one: Package a real lender. capital sits in a variable vault, quotes a fixed rate into midnight, gets matched to a borrower. you're a bond dealer running a matched book, pairing someone who wants to lend fixed with someone who wants to borrow fixed. @TenorFinance is doing this, built straight on midnight. the lender holds the fixed risk because the lender wanted it. The non-obvious one, and the part i've been thinking more on is how protocols can build or leverage on top. And recently i've gone kinda deep on @iris_credit, this is where i think it fits pretty clean (and where it also breaks). IRIS is an intent-based model. A borrower just signs "i want fixed, this size, this term," and solvers compete to deliver it. The solver isn't boxed into one venue or one rate. it funds floating wherever's cheapest and actively manages that liability over the life of the loan. That's just how banks work. They don't fund a fixed mortgage by finding a matching fixed deposit. Treasury desks manage the liability stack over time, reshuffling funding to keep cost down. tldrl; Midnight is supply-side rails. IRIS is a demand-side management layer. Think of midnight is the zero-coupon bond. IRIS as the treasury desk that funds it. - Midnight makes the fixed rate a property of an instrument, t-bill-like at a high level. - IRIS makes it a property of a managed service, like a bank issuing a fixed-rate mortgage. So, where I feel they connect and what does IRIS actually get by leaning on midnight. 1) Right now the solver quotes fixed but funds floating, so it carries the basis. midnight gives it a fixed-rate funding source. it can match-fund a loan straight off midnight and kill the basis at origination instead of wearing it. 2) A reference. even today midnight prints an observable fixed price per tenor, a curve solvers can actually quote against instead of guessing the term premium. Underwriting gets tighter because the thing you're pricing is finally visible. 3) Depth. Midnight aggregates fixed-rate liquidity that IRIS solvers can tap, so the solver isn't the only balance sheet warehousing the risk. More places to lay it off = thinner bonds = tighter quotes passed to the borrower. And it runs both ways. IRIS is the demand aggregator midnight is missing, it concentrates scattered borrower intent and routes it down. Midnight gives IRIS the fixed leg it doesn't have. complements at the infra layer. Also, interestingly, @iris_credit and @TenorFinance seem to fit the two sides of one book. IRIS aggregates the borrowers, tenor packages the lenders, and the two meet on midnight. Where the tenors line up, a solver can fund off a @TenorFinance vault's fixed offer instead of going floating, essentially locking its cost instead of carrying the basis. a selective move, not the whole book. FWIW, the basis doesn't disappear, it just moves down a layer. - And at the bottom, someone has to actually want to hold that fixed risk to maturity, the natural lender defi still doesn't have. @TenorFinance can package that lender, not manufacture one. - And depth thins fast past the front of the curve, exactly where IRIS lives. Anyway, very much thinking out loud. Would love someone to rip it apart. Needless to day, fixed rate in defi isn't a simple product that can be solved easily, it's a whole stack reassembling onchain piece by piece.

Onchain lending was never going to reach the masses in it's raw form. Vault-based structured products are how it actually scales. Just not for the reason that worries you. First, ofc a vault isn't bulletproof inherently. It exists to isolate risk so one blowup doesn't take the system with it, and to hand a user a finished product instead of parameters to babysit and rebalance alone. That abstraction isn't DeFi betraying its mission. It's DeFi finally getting a MEGA distribution layer. It's why @coinbase routes USDC through @SteakhouseFi on @Morpho, why @krakenfx earns through @veda_labs (s/o to @sunandr_ and team), why capital that never touched a lending market now sits inside one. The packaging is the product. And the part this critique skips. Someone always does the underwriting (or risk management). No lending market exists where nobody picks the collateral and prices the risk. The hardcoded oracle and the allocator you point to were a real curator failure on Resolv, and the same exploit was contained in minutes by curators who'd done theirs. On Morpho the "underwriters" or more precisely risk managers are individual curators like @gauntlet_xyz, @SteakhouseFi, @SentoraHQ, @kpk_io, @Re7Capital, etc. On Aave it's the DAO. Don't get me wrong I'm very bullish on Aave and the team, but rsETH is the cleanest proof that even a monopoly "curator" with maximum skin in the game is still a curator, and still a single point of judgment. Aave priced rsETH. It did not price Kelp's bridge. Up to $230M of bad debt, billions draining sideways through the shared pool until stables hit 100% utilization. The response was textbook-fast, frozen within the hour. The pooled structure still socialized the stress, because in a shared book one mispriced asset is everyone's problem. Roughly 40% of LayerZero apps run that same config; it was simply never priced. Which is also why Aave V4 unbundles the monolith into hub-and-spokes: the DAO holds the perimeter, specialized spokes set precise markets, a risk-premium layer finally prices collateral. To me, even the biggest team in lending seems to be moving toward more of a curation-like unbundled approach, rather than away form it. Because most capital does not want to underwrite its own risk, and never has. That is the entire reason banks, asset managers and underwriters exist, and why structured credit is a multi-trillion-dollar business. They have failed too. The answer in 2008 was never "everyone audits their own mortgage." It was better disclosure and better rails. That's the real work where defi has to step up. We have this shaping up pretty decently i'd say with risk ratings from teams like @CredoraNetwork, @xerberus, @StakingRewards, etc and an recently an EF bounty for the aggregation of all that, Another thing that helps. @andrewhong5297 and the @herd_eco team recently shipped a product that recursively maps every manager's position in a token into one balance sheet. More visibility, so any curator or risk manager can do their job better. Strip the abstraction and you don't get purer DeFi. You get a worse bank with a smaller balance sheet. Every lending market has a curator-adjacent role. Aave's is a DAO, a bank's is a committee you'll never meet. The only difference is that onchain you can audit yours in real time. Bury the risk in a pool or price it in the open. That was always the only choice.

Last week @Morpho dropped the midnight whitepaper + open-sourced the code, so i've been sitting with it and talking to a few builders. Something clicked that i don't see many people mapping. First, a recap on morpho midnight. On a high level it's a primitive. think what blue was for variable rates, midnight is that for fixed. It sits on top of deep variable markets (blue) and turns a loan into a tradable, fixed-term unit. Not a consumer product, it's infra. Minimal primitive, everyone builds on top. Classic morpho. So the same way blue needed them, midnight needs curators, solver networks, fintech/ux integration to actually scale. The obvious one: Package a real lender. capital sits in a variable vault, quotes a fixed rate into midnight, gets matched to a borrower. you're a bond dealer running a matched book, pairing someone who wants to lend fixed with someone who wants to borrow fixed. @TenorFinance is doing this, built straight on midnight. the lender holds the fixed risk because the lender wanted it. The non-obvious one, and the part i've been thinking more on is how protocols can build or leverage on top. And recently i've gone kinda deep on @iris_credit, this is where i think it fits pretty clean (and where it also breaks). IRIS is an intent-based model. A borrower just signs "i want fixed, this size, this term," and solvers compete to deliver it. The solver isn't boxed into one venue or one rate. it funds floating wherever's cheapest and actively manages that liability over the life of the loan. That's just how banks work. They don't fund a fixed mortgage by finding a matching fixed deposit. Treasury desks manage the liability stack over time, reshuffling funding to keep cost down. tldrl; Midnight is supply-side rails. IRIS is a demand-side management layer. Think of midnight is the zero-coupon bond. IRIS as the treasury desk that funds it. - Midnight makes the fixed rate a property of an instrument, t-bill-like at a high level. - IRIS makes it a property of a managed service, like a bank issuing a fixed-rate mortgage. So, where I feel they connect and what does IRIS actually get by leaning on midnight. 1) Right now the solver quotes fixed but funds floating, so it carries the basis. midnight gives it a fixed-rate funding source. it can match-fund a loan straight off midnight and kill the basis at origination instead of wearing it. 2) A reference. even today midnight prints an observable fixed price per tenor, a curve solvers can actually quote against instead of guessing the term premium. Underwriting gets tighter because the thing you're pricing is finally visible. 3) Depth. Midnight aggregates fixed-rate liquidity that IRIS solvers can tap, so the solver isn't the only balance sheet warehousing the risk. More places to lay it off = thinner bonds = tighter quotes passed to the borrower. And it runs both ways. IRIS is the demand aggregator midnight is missing, it concentrates scattered borrower intent and routes it down. Midnight gives IRIS the fixed leg it doesn't have. complements at the infra layer. Also, interestingly, @iris_credit and @TenorFinance seem to fit the two sides of one book. IRIS aggregates the borrowers, tenor packages the lenders, and the two meet on midnight. Where the tenors line up, a solver can fund off a @TenorFinance vault's fixed offer instead of going floating, essentially locking its cost instead of carrying the basis. a selective move, not the whole book. FWIW, the basis doesn't disappear, it just moves down a layer. - And at the bottom, someone has to actually want to hold that fixed risk to maturity, the natural lender defi still doesn't have. @TenorFinance can package that lender, not manufacture one. - And depth thins fast past the front of the curve, exactly where IRIS lives. Anyway, very much thinking out loud. Would love someone to rip it apart. Needless to day, fixed rate in defi isn't a simple product that can be solved easily, it's a whole stack reassembling onchain piece by piece.

was an absolute pleasure to work with @0xyanshu for over a year now he's got two complementary strengths: growth & bizdev when he joined the team, he was in his early innings of bizdev - strong network but still learning how to close deals fast forward, now he's completely capable of handling client relationships and closing deals on his own which reflects how quickly he has learnt and he's already really good at growth and marketing - his active posts on the CT timeline says it all in summary, he's a young, energetic and well-networked guy any high-performing team would be glad to have on their team