Pinned Tweet
How an RWA credit facility blows up 1. You deposit USDC into an onchain credit facility and get a yield-bearing stablecoin in return ("USDX"). USDX is backed by a portfolio of loans to offchain borrowers, according to the marketing. 2. USDX's issuer creates a Morpho market where you pledge USDX and borrow USDC, up to a 91.5% LLTV. You go to Morpho, post your USDX, borrow USDC, buy more USDX, and repeat. You now have 10x leverage. 3. Morpho values USDX using the portfolio’s NAV, relayed onchain by a “Verified NAV Oracle.” The oracle doesn't verify that the loans are worth that amount, or that the collateral can be liquidated. They pass along the issuer's self-reported NAV, which increases mechanically at par plus the coupon rate. If you bought on Day 1 at $1.00, and the annual coupon was 7%, it would be worth $1.07 on Day 365. 4. The borrowers begin to deteriorate offchain, payments stop flowing in, covenants are busted. Your Morpho position looks fine because the USDX price from the oracle continues to increase. 5. Finally the issuer marks down the NAV, on the quarterly schedule. Assuming a 91.5% liquidation threshold, a position at 90% LTV only needs a 1.64% decline to be liquidated. Several looped positions immediately fall underwater. 6. The markdown causes a run. USDX holders submit redemptions at the published NAV. The issuer pays the first redeemers from its cash sleeve, leaving the remaining token supply backed by a less liquid loan book. 7. Once the cash is exhausted, the issuer pauses redemptions, or begins selling loans at a discount. USDX is listed on DEXs - to facilitate orderly liquidations in situations like this - but it trades far below NAV because it can no longer be converted into cash. 8. On Morpho, USDC suppliers withdraw the remaining liquidity, causing borrowing rates to spike and pushing the debt positions further underwater. 9. A liquidator shows up and buys USDX, repaying a portion of the debt. They try and trade it on the DEX but it can't absorb the volume. They try and redeem directly with the issuer but they are stonewalled. They find out the token doesn't actually represent a legal claim on the collateral, or really anything, despite what the marketing said. 10. No other liquidators show up. At a 91.5% LLTV, the liquidation bonus is only ~2.6%, not nearly enough to step in and pursue recovery. 11. You lose your initial investment. The CEX "Earn" depositor who unknowingly funded the whole operation (via an embedded curator vault), loses their entire position.
74
36
420
88,278
SEC confirms that a token issuer's ongoing efforts after their protocol is functional do not constitute "managerial efforts" under Howey. Huge update + addresses the main concern w/ their prior guidance.
*SEC STAFF ISSUES FAQS ON CRYPTO ASSET SECURITIES LAWS APPLICATION *SEC STAFF: TOKEN BUYBACKS ON FUNCTIONAL PROTOCOLS DO NOT CONSTITUTE MANAGERIAL EFFORTS *SEC STAFF: LIQUID STAKING TOKENS ARE DIGITAL COMMODITIES OR TOOLS, NOT SECURITIES *SEC STAFF: MAINTENANCE, ENHANCEMENTS, SYSTEM GRANTS NOT CONSIDERED ESSENTIAL MANAGERIAL EFFORTS *SEC STAFF: PROMOTING CRYPTO UTILITY WITHOUT PROFIT CLAIMS GENERALLY NOT AN INVESTMENT CONTRACT
2
4
12
1,391
Jeff Amico retweeted
Gensyn deepens prediction market research with new experts Raf Frongillo and Bo Waggoner have joined Gensyn bringing their extensive experience in prediction market research, work that is essential to building machines that predict the future Read more: gensyn.ai/news/prediction-ma…
1
8
43
5,016
Popular distrust in AI is growing. We can solve this with bureaucracy - enshrining a few companies to build all the models and meter out access Or we can solve it with math and cryptography, allowing anyone to prove what's in the model, including the data it was trained on.
open-1b: the first model you don’t have to trust Auditable training is the best defense against the future of AI we’re being warned about open-1b is a milestone toward verifiable AI, a goal that is absolutely necessary for the future of intelligence gensyn.ai/news/introducing-o…
3
15
58
5,434
Jeff Amico retweeted
today we announced OPEN-1B, the world's first fully auditable transformer training run OPEN-1B proves that we can log and audit every single step of AI training and inference, providing proof of its training data, recipe, biases, and weights open1b.gensyn.ai/
10
32
139
24,028
📍RWA Summit, NYC
8
12
971
Jeff Amico retweeted
"We need infrastructure like oracles that can verify offchain state such as covenants, collateral perfection, NAV (net asset value), etc." @_jamico on verifiable oracles
Can RWA credit vaults scale if depositors don’t have direct legal claims on borrowers or collateral? @_jamico of @gensynai says many investors are simply trusting the platform to pay them back and argues that stronger creditor rights are needed before the sector can scale safely. news.bitcoin.com/interview/g…
4
10
31
7,765
Jeff Amico retweeted
Can RWA credit vaults scale if depositors don’t have direct legal claims on borrowers or collateral? @_jamico of @gensynai says many investors are simply trusting the platform to pay them back and argues that stronger creditor rights are needed before the sector can scale safely. news.bitcoin.com/interview/g…
18
14
41
39,893
Jeff Amico retweeted
back when i was running a crypto fund, i used to think about this a lot. i love money markets and i really wanted these activities onchain. everyone focuses on how good the underlying collateral is, but the real question for me was always: if things move against you tomorrow, who actually controls it? at 90% ltv, you can be economically wiped on a ~10% move while the nav is still stale and redemption is weeks away. and if the lender can’t directly acquire the crypto underneath, the fact that it’s liquid doesn’t help much. some of my portfolio companies did implement this but not at scale.
How to get 10x leverage on private credit 1. You go to JP Morgan and tell them you have $100 of shares in a non-traded private credit fund ("Red Dove") and you want 10x leverage. Red Dove makes loans to crypto hedge funds and passes along the yield. 2. You tell them the loans are short term and backed by crypto collateral. They ask to see the Subscription Agreement but you didn't sign one, you just deposited money into Red Dove's online platform. They ask to see the Security Agreement, to confirm the underlying collateral was pledged correctly. You haven't seen this either. They ask if there is any proof that you own shares in the fund. You tell them you have a digital receipt and point them to the platform's Terms of Service, a standard SaaS TOS. 3. They ask how the digital receipt is valued. You tell them that it's formally marked every quarter, but otherwise it increases at par plus the accrued interest. Have they ever marked the loans down? No, these are institutional borrowers. Have they ever liquidated the collateral? No, the team has a background in underwriting. 4. They ask how they would liquidate the digital receipt, in the event the NAV fell, or rates increased, and your margin went underwater. You say they could go to Red Dove and redeem, although this takes 1-2 months lead-time. They ask if any legal rights automatically transfer to them when they acquire the digital receipt. No, they need to be first approved by Red Dove. What happens if Red Dove doesn’t approve? 5. They thank you for coming in, and politely decline the offer. You ask why, they tell you that there’s a clear mismatch between the funding you’re seeking and the assets you’re financing. Most they could do is 5% LTV, but even that’s unlikely given the legal issues. As you walk out, they begin to dial the SEC, you tell them you don’t need them, you can get this financing elsewhere.
3
12
3,158
Jeff Amico retweeted
The @Delphi_fyi information markets can now be settled automatically with verifiable ML models. Markets can be created over anything - and the creator configures how the ML model should run. $51k of volume processed by the system so far! 💪
4
10
26
2,861
How to get 10x leverage on private credit 1. You go to JP Morgan and tell them you have $100 of shares in a non-traded private credit fund ("Red Dove") and you want 10x leverage. Red Dove makes loans to crypto hedge funds and passes along the yield. 2. You tell them the loans are short term and backed by crypto collateral. They ask to see the Subscription Agreement but you didn't sign one, you just deposited money into Red Dove's online platform. They ask to see the Security Agreement, to confirm the underlying collateral was pledged correctly. You haven't seen this either. They ask if there is any proof that you own shares in the fund. You tell them you have a digital receipt and point them to the platform's Terms of Service, a standard SaaS TOS. 3. They ask how the digital receipt is valued. You tell them that it's formally marked every quarter, but otherwise it increases at par plus the accrued interest. Have they ever marked the loans down? No, these are institutional borrowers. Have they ever liquidated the collateral? No, the team has a background in underwriting. 4. They ask how they would liquidate the digital receipt, in the event the NAV fell, or rates increased, and your margin went underwater. You say they could go to Red Dove and redeem, although this takes 1-2 months lead-time. They ask if any legal rights automatically transfer to them when they acquire the digital receipt. No, they need to be first approved by Red Dove. What happens if Red Dove doesn’t approve? 5. They thank you for coming in, and politely decline the offer. You ask why, they tell you that there’s a clear mismatch between the funding you’re seeking and the assets you’re financing. Most they could do is 5% LTV, but even that’s unlikely given the legal issues. As you walk out, they begin to dial the SEC, you tell them you don’t need them, you can get this financing elsewhere.
17
7
61
7,745
♬ Hello darkness my old friend ♬
How an RWA credit facility blows up 1. You deposit USDC into an onchain credit facility and get a yield-bearing stablecoin in return ("USDX"). USDX is backed by a portfolio of loans to offchain borrowers, according to the marketing. 2. USDX's issuer creates a Morpho market where you pledge USDX and borrow USDC, up to a 91.5% LLTV. You go to Morpho, post your USDX, borrow USDC, buy more USDX, and repeat. You now have 10x leverage. 3. Morpho values USDX using the portfolio’s NAV, relayed onchain by a “Verified NAV Oracle.” The oracle doesn't verify that the loans are worth that amount, or that the collateral can be liquidated. They pass along the issuer's self-reported NAV, which increases mechanically at par plus the coupon rate. If you bought on Day 1 at $1.00, and the annual coupon was 7%, it would be worth $1.07 on Day 365. 4. The borrowers begin to deteriorate offchain, payments stop flowing in, covenants are busted. Your Morpho position looks fine because the USDX price from the oracle continues to increase. 5. Finally the issuer marks down the NAV, on the quarterly schedule. Assuming a 91.5% liquidation threshold, a position at 90% LTV only needs a 1.64% decline to be liquidated. Several looped positions immediately fall underwater. 6. The markdown causes a run. USDX holders submit redemptions at the published NAV. The issuer pays the first redeemers from its cash sleeve, leaving the remaining token supply backed by a less liquid loan book. 7. Once the cash is exhausted, the issuer pauses redemptions, or begins selling loans at a discount. USDX is listed on DEXs - to facilitate orderly liquidations in situations like this - but it trades far below NAV because it can no longer be converted into cash. 8. On Morpho, USDC suppliers withdraw the remaining liquidity, causing borrowing rates to spike and pushing the debt positions further underwater. 9. A liquidator shows up and buys USDX, repaying a portion of the debt. They try and trade it on the DEX but it can't absorb the volume. They try and redeem directly with the issuer but they are stonewalled. They find out the token doesn't actually represent a legal claim on the collateral, or really anything, despite what the marketing said. 10. No other liquidators show up. At a 91.5% LLTV, the liquidation bonus is only ~2.6%, not nearly enough to step in and pursue recovery. 11. You lose your initial investment. The CEX "Earn" depositor who unknowingly funded the whole operation (via an embedded curator vault), loses their entire position.
10
43
5,952
Maybe there should be some like, laws for this kind of stuff or something
1
3
255
How most "RWA credit vaults" work today 1. You deposit USDC into a smart contract and get a "yielding stablecoin" in return. 2. The affiliated labs co / foundation takes your money from that contract and lends it to an offchain borrower. Sometimes the borrower pledges / sells assets to an SPV as collateral. The SPV is usually managed by the labs co or an affiliate. 3. You have *no direct legal relationship* w/ the end-borrower, and only a Terms of Service w/ the labs co / foundation, which disclaims all liability. You are not a lender to anyone, just a "depositor" into a "protocol". 4. You have no transparency into the ongoing health of the loan / collateral, other than a monthly or quarterly report by the platform. You may have a dashboard, but it doesn't prove that the loan exists, the SPV owns it, the collateral is still there / hasn't been repledged, or the covenants are being met. 5. In a default, you are an unsecured creditor (at best) to the labs co / foundation, which expressly disclaimed liability in the TOS. You have no recourse against the end-borrower or its collateral. RWA credit is an extremely promising direction! And a natural fit for smart contracts, stablecoin payments etc., if you set it up correctly. But there's a long way to go.
64
39
354
27,617
Lord give me the confidence of a crypto lender depositing into DeFi and earning 100 bps below SOFR
Show me the cheapest places to borrow USD-pegged assets against BTC or ETH collaterals with available liquidity of =>10m.
6
13
2,439
Enjoyed this conversation with @JacobRobinsonJD on prediction market oracles and how to fix them.
Prediction markets are a (poorly understood) multi-billion dollar industry. This @LawofCodeFM episode is a multi-hour deep dive on prediction markets, from conclave betting in 15th century Rome to proposed rulemaking from the @CFTC earlier this month. My goal: the internet's most comprehensive explainer on prediction markets. I spent several months and over 100 hours on this podcast, and spoke to the world's leading experts on the legal layer of prediction markets: @robertjdenault, @WALLACHLEGAL, @SPSchropp, Sam Enzer, @iampaulgrewal, @BradBourque, @ThaniaCh, @passalacqua_mj, @JoshSterlingLaw, @_jamico, as well as @KolemanStrumpf, @mattkalish, with clips from prior conversations with @giancarloMKTS, @DustinGouker. By the end of this episode, I promise you'll be in the top percentile for understanding prediction markets, regardless of your starting point. (You just might want to listen twice. There's a lot here.) 0:00 Intro 1:40 16th century papal betting @KolemanStrumpf 11:13 Insider trading rules @robertjdenault 16:20 The Google insider case 27:38 Why prediction markets matter @giancarloMKTS 33:20 Election betting in America 44:50 Iowa Electronic Markets 52:11 Dodd-Frank, swaps and the Special Rule 54:23 Senator Lincoln on sports contracts 1:02:04 Parlays as swaps 1:07:38 CFTC's exclusive jurisdiction @ThaniaCh 1:13:45 Perspective on CFTC's NPRM @passalacqua_mj 1:21:10 Exceptions that swallow the rule @iampaulgrewal 1:33:40 How prediction markets actually work 1:42:20 Kalshi's fee structure 1:44:33 Cardi B and the resolution problem @DustinGouker 1:51:20 @Polymarket's decentralized resolution @_jamico 1:58:10 The Ninth Circuit case 2:14:04 CFTC's proposed rulemaking @BradBourque, @SPSchropp 2:25:21 @Kalshi's landmark 2024 win 2:29:20 PASPA, Murphy v. NCAA @WALLACHLEGAL 2:51:29 The case against banning prediction markets @robertjdenault Nothing in this podcast is legal or investment advice. Thank you to the sponsors of this episode, @CahillGordon (@NYcryptolawyer), @HyperliquidPC and .
16
3
34
3,332
RWAs need better oracles. Can't just relay the borrower's claims at face value. Need to independently verify collateral / covenants on a regular basis.
Accountable has terminated its service agreement with MainStreet, effective immediately. MainStreet was unable to meet our verification standards. We recognize that markets relied on this feed and are working closely with partners across the ecosystem as the situation evolves. We will continue to hold this standard without exception.
22
2
57
6,248
The way to keep open source legal is to make it even more open. Weights, data, recipe should all be open source and provable. Show you didn't train on harmful data or steer the model. Doesn't eliminate risks but strikes the right balance.
Banning open-source AI in any form would be a mistake. A general audience PSA with @kevinsxu on why open source upholds American values. Managing frontier risks is hard, but reducing transparency, innovation, and education from kneecapping the open frontier would be worse.
12
3
44
5,156
Nemotron hasn't cracked the Chinese OS stranglehold, despite being free for 2 weeks. Would love to see user geo data from @OpenRouter. Possibly an East vs. West thing.
10
1
22
2,225