Traevon

Introducing a breakthrough in DeFi pricing infrastructure: Dynamic PT Oracle, developed in collaboration with @objectivedefi. A new pricing engine designed for PTs and how they trade in the real world. The first use cases are launching on @eulerfinance. Learn more 🧵

Capital always flows to yield. TradFi proved it for 100+ years. Now, crypto is set to replicate it at internet speed. Today, we publish the Yield Bearing Assets & Stablecoins Report 2025. Only 8–11% of crypto yields vs 55–65% in TradFi. A structural breakout is underway 🧵

With all the focus on risk exposure in lending markets lately, we wanted to share a bit from our work with Euler and how its modular architecture isolates risk by design. First things first: all Euler DAO markets across all chains, including Euler Prime and Euler Yield, have zero exposure to Stream assets either directly or indirectly. These core markets that most users interact with are completely unaffected. There is exposure to Stream assets in a few permissionless vaults. However those vaults are isolated islands within the protocol, not systemic risk to all Euler users. This is one of the core design principles of Euler v2. The Euler Vault Kit allows anyone to permissionlessly deploy entire lending markets and configure them however they like. Critically the EVK places guardrails such that risk decisions in one vault don't cascade across the protocol. Each vault operates as an independent lending facility with its own risk parameters and collateral types. What does this mean practically? If a vault accepts Stream assets as collateral and that collateral loses value, only lenders in that specific vault face exposure. Lenders in other Euler vaults have zero exposure because they're completely separate contracts. This is fundamentally different from monolithic lending models where all lenders share exposure to all accepted collateral types. In those systems, one bad collateral decision affects everyone. The core Euler markets continue operating normally because that's exactly how isolated vault architecture is supposed to work. No assets are "locked" at the protocol layer. The recent volatility across DeFi resulted in temporary utilisation spikes as users reassessed risk. In these cases, the mechanisms of interest rate models kick in: high utilization drives up borrow rates, incentivizing repayments and new supply, which naturally rebalances the market. We're already seeing this play out and utilization on key Euler vaults is normalizing. Lending protocols built on permissionless infrastructure accept a plurality of risk opinions -- some more conservative than others. Euler's infrastructure ensures risk stays contained. That's the tradeoff you must embrace when building truly open financial infrastructure.

Taming DeFi's Ouroboros A deep dive into quantifying recursive lending risk and why your "diversified" yield vault might be a ticking time bomb 1. Intro Managed vault strategies increasingly allocate to venues that provide leverage against shares of those same vaults. This creates circular dependencies, what I call the ouroboros, where losses propagate back to their origin, amplifying drawdowns beyond what isolated risk would predict. In this short technical exploration I attempt to quantify when feedback loops activate, how much they amplify, and which metrics signal dangerous configurations. 2. Defining the Ouroboros Most lending risk is linear. A 10% collateral drop causes proportional losses scaled by loan parameters. The ouroboros introduces threshold-dependent feedback. Below a critical price drop, risk behaves normally. Above it, losses loop back and amplify. The ouroboros emerges when a vault allocates to venues that lend against the same vault's shares as collateral. When those shares drop enough to trigger liquidations, the lenders take losses, which flow back to the vault through its holdings. The initial shock gets magnified. 2.1. A Minimal Loop The simplest ouroboros is a two-vault cycle. Consider a managed USDC vault, Vault A, that allocates 20% to Vault B. Vault B provides borrowable USDC liquidity to A/USDC, completing the dependency cycle. As long as Vault A holds its NAV, the feedback loop stays dormant. Once Vault A loses enough value to trigger liquidations in the lending market, Vault B (as the lender) takes a haircut, which flows back to Vault A through its 20% allocation. Notably, this behavior is dormant up to an activation threshold, above which feedback mechanisms engage and amplify the initial shock. 3. Quantifying Feedback Let's trace what happens when Vault A loses 15% in NAV. 1. Vault B has 50% allocated to loans against A shares at 86% LLTV. 2. Liquidations trigger. The liquidator sells A shares with 5% slippage. 3. After settling, debt was $86 per $100 of original collateral, but only $80.75 is recoverable (5% liquidation cost). 4. The shortfall is $5.25 per $86 debt, roughly 6.1% loss on those positions. 5. Since B allocated 50% to these loans, B's NAV drops 3.05%. 6. Vault A holds 20% in B, inheriting 0.61% additional loss on top of the initial 15%. This additional loss is Loss Given Default, LGD(d) = max(0, 1 - (1 - h)(1 - d) / LLTV). LGD remains zero until d crosses the activation threshold, d > 1 - LLTV / (1 - h). This is roughly 9.5% for these parameters. Below this, liquidations recover all debt and lenders see no losses. Above it, losses materialize and feedback engages. In this scenario the 15% shock amplified to 15.61%. The loop coefficient (allocation weights multiplied: 20% * 50% = 10%) sets the structural feedback path, i.e. 10% of LGD echoes back to Vault A per feedback loop. The danger scales nonlinearly: higher allocations increase the loop coefficient, while larger shocks push LGD toward 100%. With B at 70% and A at 40%, a 15% shock becomes ~16.8%. A 25% shock could amplify past 30% as LGD approaches maximum and the full loop coefficient engages. 4. Measuring Stability The loop coefficient captures feedback in our two-vault system. For larger networks with multiple cycles and cross-holdings, we need a system-wide metric: the spectral radius (ρ), which measures how much any perturbation amplifies as it propagates through all network paths simultaneously. When ρ < 1, shocks dissipate. When ρ > 1, shocks amplify without bound (in DeFi, a death spiral). The spectral margin (1 - ρ) measures distance from instability. In calm conditions, our network has no active feedback edges and all conditional loans sit dormant. The spectral margin is 1.0. Under 15% stress, those edges activate via LGD. The feedback paths engage, pushing ρ from 0 to 0.08. Spectral margin drops to 0.92. This drop is the manifestation of latent instability within the network. Higher allocations in loops, higher LLTVs, or worse liquidity all push ρ higher. Even with positive spectral margin, shocks can still amplify significantly through repeated passes. The margin tells us we’re far from runaway cascades but it doesn’t mean that we’re safe from meaningful losses. Beyond spectral margin, two metrics complete the picture. The self-echo multiplier measures how much a direct shock to a single vault amplifies after all network feedback returns to it. In our example, Vault A has a self-echo of ~1.04x under stress: a 10% direct hit to A becomes 10.4% after the loop completes. In larger networks with multiple paths back, this can be much higher, revealing which vaults sit at the center of the most dangerous feedback cycles. While self-echo isolates individual vault risk, the network loss multiplier captures aggregate damage: total system loss divided by the initial shock. Our 15 to 15.6% example has a network multiplier of ~1.04x. In historical events with cascading liquidations across multiple vaults, this metric reveals how much the ouroboros structure amplified losses beyond what isolated failures would have caused. 5. Conclusion Like the archetypal snake, DeFi's ouroboros lurks in the shadows. Each managed vault abstracts away complex DeFi strategies, often distributed across wallets and chains. These same vaults are presented with "institutional-grade" gift wrapping to unsuspecting lenders. Because recursive lending inflates both TVL and net yield, participants in the network often choose to turn a blind eye. The ouroboros is deceptive because it's emergent: no single vault creates the risk, yet the network as a whole amplifies shocks. The loop only activates under stress and by the time the feedback becomes visible, you're already in the spiral. The next time you lend to a managed vault, take a look at its portfolio.

Over the weekend we learned that our new CMO crossed a line. His actions were taken without my knowledge or the knowledge of our team. Coordinated doxxing campaigns have no place at RedStone. We verified the facts and made a decision: we cannot continue working together, effective immediately. Below is my perspective. We have a small marketing team and wanted to bring on a seasoned CMO. Peter joined us in September with years of professional marketing experience. Figuring out communication strategy in a competitive market is part of the job. But it has to be done following the company values. RedStone team are builders, not drama seekers. That's not who we are, and it's not what we stand for. We mutually agreed with Peter to part ways effective immediately. Peter acted outside our boundaries without the knowledge or support of our team. We have our differences with Chainlink, but we have many similarities too. We believe the world will run on crypto infrastructure, that oracles are an essential part of the system, and that Chainlink helped to bootstrap DeFi. We respect their skills and achievements. The oracle space needs healthy competition that drives progress, not toxic behavior. CatfishFishy and I can have our exchanges of views, but ultimately, we're all playing to grow the crypto pie. That is what matters long term. It's your choice to be anonymous or public: that's a personal decision. But targeting someone's identity to cabal against them is not the crypto industry I want to build. Our top priority remains building the most reliable and useful oracle technology for the crypto ecosystem. That's where our energy goes. After this, I'm only more certain what I and RedStone strive for and where our integrity lies - on supporting builders with the highest quality and reliability of blockchain oracles data. We've grown RedStone from 15 to more than 50 people over the last year. Our team has shipped Bolt, Atom, Credora DeFi Ratings, dozens of network launches, and tokenized asset feeds for the world's biggest financial institutions. We've helped over a hundred clients expand and reach their potential. We built RedStone to make crypto a better place by offering best-in-class products and building honest relationships. That's how we measure our success. Here's RedStone's Official statement: x.com/redstone_defi/… We're committed to staying on that course and keeping our values intact as we scale. Our foundation and motto doesn't change: By builders, for builders.

🔥 Don’t miss the $100M Bonus Pro campaign with 100% deposit bonus 👉 lbank.com/activity/bonus… 🏆 Trade like a champion with @LBank_Exchange, now the official regional sponsor of @AFA 🇦🇷⚽️ To celebrate, we’re giving away 2 official Argentina jerseys 👕🎉 How to enter: ✅ Follow @LBank_Exchange, @AFA & @Degen_Radar ✅ Like + RT this post & tag 2 friends ✅ Use #BlueWhiteOnLBank

It’s been an honor to collaborate on this piece with the giga-brains at @Sumcap. Hyperliquid.

@arno39 @Dune @HyperliquidX tbh, I was just wishing for HyperEVM data on Dune in company meeting yesterday, and then it showed up in the evening. Were you guys eavesdropping on us? 😂

HyperEVM finally nuked its way onto @Dune: $1.9B TVL 60k weekly actives $4.35M fees burned 2k+ NFT collections All @HyperliquidX data, raw and uncut. Go touch some dashboards

.@Theo_Network's thBILL grew from $0 to over $40M TVL in 7 days on @HypurrFi. Explore how HypurrFi's E-Mode enabled this breakthrough: 🧵

@MSilb7 and a picture is worth a thousand words.

dashboards are just pictures

17.41% for depositing $USDT on @HypurrFi. Happy lending.

ICYMI: @HypurrFi has maintained market-leading APR stability on WHYPE for 30+ consecutive days. Their volatility metrics outperform most other HyperEVM protocols - creating measurable alpha for both retail and institutional allocators. Why this matters for your strategy:👇

Everything you need to know about the @eulerfinance x @LineaBuild collaboration 🤝 ▪️ Isolated ETH markets: $WETH, $wstETH, $ezETH, $weETH ▪️ LST + LRT looping strategies ▪️ Risk-managed by @gauntlet_xyz ▪️ Growth-led by @objectivedefi Modular. Risk-aware. ETH-aligned.

Euler Yield just crossed $700M in TVL Since its inception, @eulerfinance's stablecoin-focused market has delivered industry-leading yields, proving true product-market fit. This recent milestone comes after a concentrated effort for risk-adjsuted growth: 🧵

Will ChatGPT-5 Boost the Next AI Coin Pump? .@OpenAI is expected to launch ChatGPT-5 this August with major improvements like: 👉Smarter reasoning 👉Full multimodal capabilities (text, image, voice, etc.) 👉Stronger AI agents 👉A significant upgrade in overall intelligence ⁉️Will this upgrade boost AI-related tokens like previous OpenAI announcements did? TBH not sure, but I believe so for the following reasons: 1⃣ Historical Trends: Every time OpenAI announced a big update, the overall AI token market outperformed 2⃣ Real Utility Matters: Over time, the market tends to reward projects with real use cases, not just hype 3⃣ Growing AI Ecosystem: A smarter ChatGPT-5 could drive demand for AI infrastructure and agent platforms. That means more features, more users, and more adoption all of which can help drive token prices Let’s take a look at some AI coins worth watching 👇

LIVE NOW Me-OW! Spend without selling your crypto across 100M+ global merchants to pay for your milk, kibble, litter, & anything else you tap to pay or buy online. We're purring bonus points on spend for HL's first live Credit Card w @BrahmaFi’s Swype. Program details below.

ICYMI, right now on @HypurrFi: $kHYPE: 2.3% APY (2.2% from @kinetiq_xyz + 0.11% native), 5x Points (up from 3x) $HYPE: 7.65% APY, 10x Points $USDT: 15.16% APY, 7x Points $USDHL: 20.33% APY, 5x Points Juicy.

@jayowtrades @objectivedefi LFG broooo!

Happy to announce that I’ve joined @objectivedefi as a Risk Analyst. It’s been a month in, and I’m loving the ride so far. We help teams assess risk, design better incentives, and make smarter protocol decisions using onchain data. Objective Labs partners with top DeFi protocols to turn data into action. From risk assessment to growth planning, we help teams scale with confidence and clarity. Excited for what’s ahead. Let’s build.

@Hyper_Strategy @kinetiq_xyz @HypurrFi purrfect

We’re expanding our strategic collaboration with @kinetiq_xyz and @HypurrFi to maximize yield for more HYPE We partner with the best to get our treasury the best yield in HyperEcosystem. Here's how we're doing it Deposit your $HYPE into Kinetiq's kHYPE vault kinetiq.xyz/khype -Earn ~2.2% APR - Collect Juicy Kinetiq Points Then get them kHYPE over to app.hypurr.fi/buddies/GBOND and deposit it - Get a 3x Juice Points booster - Stack an extra 0.20% APR - And an extra 5% HypurrFi Points We are effectively Staking our Hype, earning yield and points then depositing for additional points and yield. For all you DeFi degens out there and new users, Put your Hype to work, accumulate more Hype. This is just the beginning of HyperStrategy. HyperStrategy