TWAP Clapper

Thank you for letting us know Will. We are aware of the issue and are in contact with the impacted users, actively developing a plan with them to mitigate the issue. In our effort to adjust the LTV, we tasked a smart contract engineer to make the necessary changes. However, it was discovered that the liquidation threshold was altered unexpectedly without prior notification to our team, leading to the current issue. Going forward, we will set up a governance contract/timelock and forum for all future upgrades to ensure that discussions are planned ahead of time.

Hi Guillaume :) You are right LVR is only a means, not an end, what matters is AMM LP profitability, and LVR is just a metric that can (maybe) help to unravel AMM LP protif-and-loss in a systematic way. This LVR metric is indeed opinionated as it seeks to separate out market risk by analyzing LP PnL _assuming a perfectly hedged LP position_. Very likely, the legitimate discrepancy of "whats the right way to look at things" stems from this specific framing for analyzing LP PnL (assuming the perfect hedge). Note that market risk can still be analyzed, it is another important risk, but being able to separate out the different risk axes is likely good/useful, and the paper is considered a breakthrough precisely because of this separation and the (potential) clarity that it brings. Moreover since market risk can in theory be hedged perfectly, it is quite useful to know that unhedge-able costs remain for CFMM LPs (on the ethereum blockchain) and what they are. Before the paper, this was not as clear or well understood. The two unhedegable costs of LPing: gas fees you pay to do transactions, and this last interesting term that appears, the infamous "LVR" term, which is essentially the running value leak to arbitrageurs that comes from your pocket as an LP thanks ONLY to how CFMMs and ethereum currently work. LVR is important bc it's a cost that AMM LPs are paying and it is NOT due to your LPing abilities (constructing and adjusting hedge, picking liquidity distributions), it is completely a function of how the AMM mechanism is designed (and the block production mechanisms are designed). If you can solve this structural disadvantage for DEX LPs with innovative DEX application design, it will be important. LPs won't notice "wow LVR disappeared!" they will just notice "I turn a profit much easier/more often as an LP on this DEX!" It's because you plugged the leaky hole in the system that made LPs job an even larger uphill battle (turning a profit as an LP will still be tricky but you won't have to overcome this extra burden anymore). AFAICT the only legit objection to LVR framing is that LPs don't perfectly hedge their positions, and hedging can also come with high costs or just be plain impossible in real (not idealized) markets, so the market risk shouldn't be left so easily by the wayside. Here I could agree! Though I would still see the LVR framework as useful as we can try to quantify the risks of not hedging at all and/or the cost of maintaining your hedge, to get a more complete picture (LVR doesn't go away, just these additional aspects appear). I also agree that profitability is what matters. Whether you "reduce LVR" (aka reduce running costs of LPing) or "increase swap fees" (aka increase revenues you collect from LPing) these both amount to the same thing for your PnL. However naively increasing swap fees doesn't seem to work that great because you disincentivize the flow you want just as much as you disincentivize the flow you don't want (arbs). Other methods seek to tax arbitrageur-trades more than non-arbitrageur-trades, which should be strictly better than just raising swap fees indiscriminately to better compensate LPs overall (because unifrom fee hikes will supress all-around volume EVEN MORE than the more interesting arbitrage-identifying solutions)

🚨🚨 WARNING 🚨🚨 The author (@Gas404Official) of the medium post linked to in the above tweet is connected to the butterfly-effect.eth wallet which was responsible for the original exploit. Writeup in the document linked below: drive.google.com/file/d/179w…