Unchained Podcast – Ep. 984

The Chopping Block: Hyperliquid vs. Tarun, ADL Transparency & The Coming Perps Arms Race
Date: December 17, 2025
Host: Laura Shin
Panel: Haseeb Qureshi, Tarun Chitra, Vlad (LiDAR), Tom (absent; brief mentions)

Overview

This episode centers on the recent controversy in the decentralized perpetuals (perps) market—specifically, the fallout and industry-wide debate around Tarun Chitra’s recent research paper on auto-deleveraging (ADL) and the subsequent public sparring with Hyperliquid’s team and investors. The panel discusses:

• What ADL is, why it matters, and how different exchanges implement it
• The arms race among perps platforms (Hyperliquid, LiDAR, others)
• Transparency and fairness in DeFi risk management
• The evolution of fee models and how innovation is mirroring/contrasting with TradFi
• The ongoing debate around token/equity, especially in the context of M&A (Axelar, Tensor, Aave drama)
• Insights into the future roadmap for DeFi platforms, including cross-margining and RWAs

Main Topics & Key Discussion Points

1. What is ADL and Why Does It Matter?

(00:00–12:26)

• Definition & Genesis:
  ADL (Auto-Deleveraging) is a process triggered in high-volatility markets when traditional liquidation methods can't resolve insolvencies—as a result, profitable traders may have some of their profits socialized to cover losers.

  “ADL stands for auto deleveraging. This is a process that happens when basically a market is moving really crazy fast... must forcibly close out positions in order to deleverage the market.” — Haseeb (01:21)

• Historical Context:
  ADL originated in centralized exchanges (Huobi, BitMEX), then was adapted for perpetuals. The process and documentation is often opaque.
• Transparency Problem:

  “A lot of people... DM'd me and said, hey, these ADL repeated events have happened before at centralized exchanges, but no one knew because you can't really see it... There's no public data. You only see yourself getting ADL’d.” — Tarun (00:00 & 19:10)

• Research Backlash:
  Tarun’s paper critiqued Hyperliquid’s ADL implementation for being “over-aggressive,” estimating $650m in over-liquidated P&L (but later revising the figure after an admitted math error).

  “The paper said $650 million, it’s probably more like $100 [million], but there’s still this amount that was over-liquidated by this algorithm.” — Tarun (17:10)

• Broader Lessons:
  • Measuring and improving fairness/predictability for traders
  • Need for public data and open-source logic
  • Comparison to bankruptcy in TradFi: if ADL fires repeatedly, the system design is off
• Notable Moment:
  The controversy erupted during Tarun’s flight—he landed to find a Twitter firestorm (04:57).

2. Platforms’ Different Approaches to ADL

(12:26–20:12)

• LiDAR (Vlad):

  • Strives for less aggressive, more trader-friendly ADL using insurance fund (LLP) first; only then targeting traders.
  • “If the algorithm is too complicated, that creates other problems in that you can’t really prove it in… real time. So our algorithm actually is also pretty, pretty heuristic based... It’s just less aggressive.” – Vlad (14:18)
  • On October 10, 2025, LiDAR absorbed losses via LLP, cushioning traders.
  • “LLP was down $20 million that day, but that also means all the other traders collectively made $20 million… perps are zero sum.” – Vlad (15:23)

• Hyperliquid:

  • Less transparency; source code and detailed parameters are not public.
  • Haseeb notes credit that at least trade data is visible.

    “That’s to Hyperliquid’s credit. Maybe you can’t read the exact ADL algorithm, but you can see all of the ADLs.” — Haseeb (20:23)

3. The Public Clash: Hyperliquid vs. Tarun & Industry Sparring

(02:34–21:26)

• Social Media Drama:
  • Hyperliquid founder Jeff tweets: "'Those who can, do, those who can’t fud.'" — (03:27)
  • Paradigm's Dan Robinson (a Hyperliquid investor) enters debate, writing dueling threads.
• Meta-commentary:
  • “I think it was a good lesson in like how research has changed in crypto… now it’s like, immediately assumed that you invested in [a competitor], so you must be fudding my bag… The idea that all research is marketing… is somewhat disheartening.” — Tarun (04:57)
• Resolution:
  Dan Robinson later agrees to participate in the V2 of the paper, marking a détente.

4. Product Innovation & the DeFi Perps Arms Race

(27:31–41:45)

• Fee Models:

  • LiDAR offers zero-fee trading for most retail; pro traders pay for speed/API access.
  • “There are three reasons for this fee structure. One is philosophical, kind of democratizing finance… [retail] is the harder part of the marketplace to attract.” — Vlad (29:52)
  • Comparisons drawn to Robinhood’s disruption in equities.

• On-chain Competition:

  • Hyperliquid launched cross-margining and new asset support.
  • Real-world asset (RWA) perps (e.g., FX, stocks) are becoming popular—surprising market leaders (Nvidia and FX perps are outperforming mainstays like Tesla).
  • "I really enjoy watching the RWA perps because the leaderboards show me things I never would have expected to see on chain." — Tarun (40:19)
  • Vlad: Upcoming features include tokenizing exchange positions, composability with other protocols (Aave, Morpho), powered by ZK tech.

5. The Token & Equity Debate: M&A and Incentive Alignment

(47:11–62:30)

• Recent Cases:

  • Axelar Devco acquired by Circle: token did not benefit, price fell.
  • Tensor acquired by Coinbase: token rose.
  • Aave Labs capturing front-end fees.

• Drama & Critique:

  • Is it fair that DevCos can be acquired, benefiting only equity holders while token holders lose?
  • “There’s no like token holder lawsuit rights, which I think is true. Maybe that’s like the real problem.” — Tarun (51:23)

• DAO Governance vs. DevCo Retention:

  • Haseeb's suggestion: DevCo docs could give token holders right-of-first-refusal in such situations, but “DAOs hate paying for talent.”
  • “One of the worst parts of DAO governance… they hate paying above market salaries… every single individual line item is overwhelmingly picked apart.” — Haseeb (61:11)

Standout Quotes:

• Haseeb (Retail vs VC):

  “Normally when a startup is down 85%, it’s time to call it quits… As an investor, it's your job to be mindful of the human capital… Retail investors are like: you created a token, you lost me money, keep working on this until I’m finally back in the money…” (59:11)

• Tarun (Legal Risks):

  “I think there’s lots of lawsuits that exist in the world that might be frivolous, but take seven years to adjudicate.” (63:12)

6. TGE (Token Generation Event) & Market Dynamics

(42:42–45:58 & 64:02–65:24)

• LiDAR’s Forthcoming TGE:
  • Vlad is realistic about market expectations and focused on long-term innovation—not short-term token price.

    “It’s a marathon, not a sprint… goal isn’t just to maximize value on day one, but to have something that's healthy and continue iterating from there.” — Vlad (42:42)

• On-Chain Activity Distortion:
  • Pre-TGE points farming and market behaviors can be misleading, so “rocket ship” launches shouldn’t be the benchmark.

7. DeFi Metrics: TVL as North Star

(45:05–47:11)

• TVL (Total Value Locked) remains, in LiDAR’s view, the most robust metric for gauging real value and trust—not just volume or open interest.
• “TVL is a metric that represents capital that customers really trust us with.” — Vlad (45:57)
• Haseeb agrees, calling it the “70 IQ, 140 IQ meme.”

8. Closing Thoughts & Meta-reflections

(65:01–End)

• Tarun: Somewhat demoralized by the social-media reaction and questions the value of public research if always taken as “fud.”
• Haseeb: “You can’t just be like, this is the beginning of your expert witness arc.”
• Ends on a positive note:

  “$100 million in LLP!” — Tarun (65:47)

Notable Quotes & Moments

| Timestamp | Speaker | Quote | |-----------|---------|-------| | 04:57 | Tarun | “I think it was a good lesson in like how research has changed in crypto… The idea that all research is marketing… is somewhat disheartening.” | | 17:10 | Tarun | "The paper said 650 million, it’s probably more like 100, but there’s still this amount that was over-liquidated by this algorithm." | | 20:23 | Haseeb | “That’s to Hyperliquid’s credit... you can see all of the ADLs. You can see the record of this person was closed at this price at this time on this chain.” | | 29:52 | Vlad | “There are three reasons for this fee structure. One is philosophical… democratizing finance.” | | 40:19 | Tarun | "I really enjoy watching the RWA perps because the leaderboards show me things I never would have expected to see on chain." | | 42:42 | Vlad | "It’s a marathon, not a sprint… we'll kind of start in a healthy place and then trend upwards from there." | | 45:57 | Vlad | “TVL is a metric that represents capital that customers really trust us with.” | | 59:11 | Haseeb | "Retail investors are like, you created a token, you lost me money, keep working on this until I’m finally back in the money, otherwise I'll hound you till the end of your days..." | | 65:47 | Tarun | "$100 million in LLP!" |

Important Timestamps

| MM:SS | Topic | |-----------|--------------------------------------------------------| | 00:00 | ADL context, transparency problem | | 04:50 | Social media firestorm over the paper | | 12:26 | Vlad explains LiDAR’s approach to ADL | | 17:10 | Tarun clarifies analysis error and transparency | | 20:23 | Public auditability, even if code isn’t open source | | 27:31 | Fee models and the Robinhood analogy | | 34:12 | Roadmap: cross-margining, tokenization, ZK sidecar | | 40:17 | Surprising RWA/FX trading patterns | | 47:11 | Token vs equity: Axelar, Tensor, Aave | | 62:30 | Should DAOs have a say in DevCo retention? | | 64:02 | Vlad on upcoming TGE and expectations | | 65:47 | Tarun’s positive closing remark |

Tone & Vibe

• Witty, deeply insider, at times irreverent (“shitstorm”, “fudding my bag”)
• Philosophical as much as technical: frequent use of analogies (bankruptcy, pharma, Robinhood)
• Candid and self-critical—especially Tarun on the fallout from his research

Summary for the Uninitiated

This episode provides a deep but accessible look at the technical, social, and incentive-driven issues at the bleeding edge of DeFi’s perps boom. ADL—a largely unexamined but critical risk management mechanism—came under the microscope thanks to a candid research paper, and the resulting community drama ends up revealing quite a bit about DeFi’s growing pains: the fight between transparency and information asymmetry, the shifting approaches to exchange fee models, and the inherent conflicts when venture investing, community governance, and open-source philosophy all collide.

For builders, investors, or anyone curious about the future of trading and risk on-chain, this episode is a tour de force on the clash of incentives, tech, and culture that will shape DeFi’s next act.

Follow LiDAR at LiDAR.xyz for TGE updates and product launches. For more reflections, catch the next “Chopping Block” in a couple of weeks.