Wanted to run a proper tweet where I cover many questions regarding LONG. Bookmark this, it’s going to be quite in depth. Before I start, I want to clarify that this is an educational effort. One of our core values with LONG is not to look down on our users but actually equip them with the right knowledge and prove another type of playbook can exist in the space. There is no reason for us to be defensive(esp not on tech) LONG proves itself every day and we will keep doing so. I also think there is a fine line between slightly disingenuous FUD vs critical thinking, so pay attention to it as well. #1 Why is LONG optimizing for liquidity as the moat with stock pairs and generally? Main problem in the space = not lack of motion but lack of stability (that drives rotation and lack of conviction) Deep liquidity solves two things: bundling and supply control are very expensive + the market can absorb extreme periods of volatility. Deep stock liquidity creates a black hole or a magnet where a LONG pair effectively becomes a secondary market for the tokenized stock and keeps a high % of circ (demonstrated in many pairs) This is the biggest pure “DeFi” flywheel and is similar to how network effects evolve around lending protocols, DEXs like Uniswap and so on. Stock liquidity is a moat that enables users to become “market makers” and share the upside of any downstream effect coming from it (increased trading activity on the stock itself, any new utilities like lending, and simply being a source for arb) #2 How exactly are stock pairs correlated with their underlying stocks? DEX pairs have a sell side and a buy side. With stock pairs, the buy side is the new token and the sell side is the tokenized stock. When the stock goes up, the USD value of stock liquidity goes along with it. This is exactly how majors were able to bootstrap the trenches in the early days of Sol + ETH. The major or stock going up = the Fed printing new money supply or giving stimulus. In practice: - If the stock went up by 20%, there is now 1.2x more stock liquidity in the pool, so selling the same token amount will give 1.2x more in USD value(rule of thumb) - This also means the impact on the chart is becoming smoother - What makes it more impactful is the depth of liquidity. If it’s 10k worth of total stock liquidity, it wouldn't matter because a 5k trade will drain the pool completely -LONG pairs are not just super liquid generally, they are also ranked as the largest sources of stock liq for the tokenized stock pools(AI is the 2nd largest source of NVDA on RH) #3 Two sides of the same coin: How do arbitrage and price coupling actually play out onchain when the stock price rises? Something VERY important to keep in mind. AI priced in NVDA and NVDA priced in AI are two sides of the same coin. You can’t have a significant depeg between the implied price of NVDA in AI vs NVDA in USDG, and the same applies to AI in NVDA vs AI in USDG. Take the following scenario: NVDA just went up by 5%, the oracle updates immediately, and now there are 2 sec for the new NVDA price to update onchain. An arb race starts: 1. Arb starts with buying AI on the AI/USDG pool (front-running on a stale NVDA price) 2. AI is being sold on the AI/NVDA pool, receiving NVDA 3. The arb bot now holds X NVDA they bought at a discount 4. The arb profit depends on how fast they can sell it on the updated NVDA/USDG pool + an optimized sell size This is a bit of a simplistic flow because, in effect, this is happening 24/7 and AI/NVDA is effectively a coordination mech to keep prices in sync. The more NVDA liq there is on the AI/NVDA pool, the more predictable this arb can be and the more “utility” builds into AI/NVDA just by being a large source of liq. This is not simple reflexivity but more of a compounding long-term game. Exactly the same effect takes place when NVDA dumps, but in this case it’s actually helping the AI (USD price) absorb volatility in a much smoother way (happened a few weeks ago when NVDA had the 1st 7d down streak since ‘22 and AI actually went up) one can think about what might happen when we drop the assumption that the price of onchain NVDA or any stock follows the stock market, and the price of onchain NVDA is actually driving the arb offchain (I’ll leave it as an exercise for the readers) #4 The LONG term game is dist >> fee capture/dividend The only way for an asset like AI to go up is having consistent flows of net new marginal buyers that are willing to buy at a higher price. This is typical growth. Every single incentive we can place on top of it gets stronger when the asset becomes bigger. A good way to think about it is stock dividends. An early-stage stock that starts handing out dividends over reinvesting into higher growth is simply sacrificing these gains. Having an asset like AI at 1b would be correlated with the ability to capture more value back to holders, whether through “dividends”, voting rights or NVDA accumulation. It won’t necessarily be a 50% APY but more similar to 1-3%, with real size (NVDA did 6b in dividends last quarter and it was just $0.25 per share) #5 Why do AI pools with USDG and ETH have so much vol and is it good? Part of it goes back to my prev point regarding arb, the other part is also all sorts of AI pairs. Having an AI/AI-pair makes it cheaper to go from USDG→AI→pair vs USDG→NVDA→AI→pair. A few immediate positive effects: 1. We already internalize this effect with AI pairs as fees remove more AI from circ regardless of whether it was routed via the main AI/NVDA pool. It also locks more AI in uncorrelated pools 2. More vol = more fees to external LPs = higher incentive to provide more liq to back AI on any pool There is no perfect fee/hook that can fully eliminate it (and it might be undesirable) for two reasons: 1. If we were to relaunch AI with 0.15%, anyone can still set up a pool with a 0.1% fee. These undercuts are very common and it’s a race to the bottom type of situation 2. You need extremely centralized and active LP management. I don’t think anyone would have wanted LONG or any launcher to have the option to just rug the entire LP #6 Is it possible to 10x NVDA accumulation or have more fee and vol capture ? Yes! Take, for example, some of our more active LP actions: we’ve added 200k worth of NVDA as a sell wall on the AI/NVDA pool. Yesterday we did something similar with 200k worth of LongX assets. Adding these into the community vault, for example, would have increased the total NVDA worth by 2x. There are dozens of other ways to do it with more sophisticated mechs. And it reminds me of the early days of the vault when users asked why we didn’t just use a buyback vault instead. This goes back to my prev point: the potential of monetization via fees is capped by the size of the asset. Anything we can do today to grow AI will pay 10x more in the future when we would want to start rolling it out. #7 “If you don’t know where the yield is coming from, you are the yield” Why is LONG not supporting reflections out of the box? Note that all of the prev points about how easy it is to undercut high-tax pools, how liquidity capture is the moat, and how organic non-incentivized growth is the real key apply even more strongly to the typical reflection mech. We think the sort of DeFi summer APY maxxing is a not part of our vision of stock pairs. We want users to buy early and hold, not because they can farm fees (which can be done on any yield-style protocol even a stable pair) We want them to buy and hold because they want to align with the stock and grow a movement around it. Generally: High-yield products in crypto have ended up dying The PMF for yield is actually super solid yield (the biggest vault on RH is USDG with 3% APY) I think this is a bit like creator fees. Incentivized vol over sustainability, and it’s so damn easy to just buy 20% at low FDV and have a no-lose option forever, then dump to move to the next as vol decays. Hope you were able to go through it! :) LONG.