Crypto Stress Test: Fees, Volatility, and Chain Performance

This weekend felt a lot like 1010 in a lot of ways where you just had this massive leverage unwind regardless of what the actual trigger event is.

But yeah, we've seen some crazy moves. Bitcoin's down 10% on a 7-day now. Ethereum's down 20%, soul down 15. There's really not that many assets that are up over that period.

Whenever these types of moves happened, it's always fun to just dive into the data and see which which chains performed and which ones kind of fell under stress. And usually there's one or two that kind of collapse under stress, but really worth noting that that didn't happen this time.

Nothing said on Lightseed is a recommendation to buy or sell any investments or products. This podcast is for informational purposes only, and the views expressed by anyone on the show are solely their opinions, not financial advice or necessarily the views of Blockworks. Our hosts, guests, and the Blockworks team may hold positions in the companies, funds, or projects discussed.

Hey guys, welcome back to another episode of the Lightseed Podcast. Today I'm finally once again joined by Dan Smith, the data goat from Block Works, as well as Ian from Chyros Research. Glad to have you on again as well.

Gentlemen, how are we doing today?

Doing great. Pleasure to be here as always, my friend.

I would say two folks I'd like to have on the pod more if possible, but I think difficult to do so. Busy gentlemen.

But I'd love to get into a few topics today. I guess one maybe we need to recap a little bit and Dan, I saw you had some data on the timeline about some of the market activities last week.

There was a crazy sort of like I guess you'd probably call it a little bit of like a maybe it's a local, maybe it's a longer term blowoff top in metals and commodities. gold and silver went crazy. People were getting very excited about actually gold and silver trading volumes like on onchain platforms like Hyperlid and even Binance listed some contracts there as well.

All the crypto platforms jumped on that train. But there was a bit of a strong pullback at the end of last week. And I guess that rippled into the onchain environment as well with some liquidations and lots of tokens being down.

So I'm curious what you took away from that, Dan. I would say I was relatively offscreens as far as that the red day is concerned.

Yeah, good question. And great setting of the scene there. That move on silver was pretty insane.

Honestly, this weekend felt a lot like 1010 in a lot of ways where you just had this massive leverage unwind regardless of what the actual trigger event is. We still see people on the timeline trying to figure out exactly what the trigger event was on 1010.

But I kind of agree with you. It looks like the move in silver was the was the first leg lower. But I haven't like truly looked to make sure that's one to one in terms of order of operations.

But yeah, we've seen some crazy moves. Bitcoin's down 10% on a 7day now. Ethereum's down 20%, soul down 15. There's really not that many assets that are up over that period.

Notable exception there is Hyperlid, which is actually up 34%. Had a huge move, sort of going against the grain there, which is pretty cool to see.

But yeah, whenever these types of moves happened, it's always fun to just dive into the data and see which which chains performed and which ones kind of fell under stress. And usually there's one or two that kind of collapse under stress, but really worth noting that that didn't happen this time.

There were pretty much every chain has done a lot of performance improvement over the last six to 12 months and you you really saw that come through.

So looking over this period basically the the entire day of January 31st and the peak TPS over 1 minute intervals for the following chains was was pretty cool to see.

So Salana hit about 4.3,000 TPS base at 1.5,000 Arbitum just over a,000 BNB at 525 and then Ethereum came in at 40 which by Ethereum standards is ripping and roaring.

But notably what happened there was the fee volatility, right? So you have a massive fall shock event, right? Prices are plummeting.

In this case, everybody everybody's racing on chain to execute some sort of transa transaction whether that's arbitrage or looking to close market gaps and dislocations or traders looking to unwind or delever positions. You've got basically everybody racing to the chain and saying I need to execute a transaction right now.

So looking at the median fee which is roughly the cost to be included in any given block is really indicative of how these congestion these moments of congestion really impact users.

So Salana actually very very outperformed here on a relative basis to this same group of five chains where the median fee only increased about 60% up to 8 hundths of a penny and or sorry 8/10 of a penny which is really really cool to see.

So again that 16% uh 60% increase sounds massive and and candidly it is but on a relative basis you really see how that 60% move is sort of non-consequential.

So base the peak median fee was 61 cents or a move of 203 times higher. Arbitum 20 cents 80 times higher. B&B chain 2 cents 13 times higher and Ethereum $8.66 for a move of 866 times higher.

So many multiples rather than Salana is less than 2x coming in about a 60% move. And if you look at a chart of this so you've showing the one where we have all the change but if you go to the next tweet so this is Salana's medium fee which held incredibly constant throughout.

You can see some some spikes here and there. But during these really really moments of high congestion you actually didn't see much volatility where if you go back just one image from this you'll see how that looks a little bit different.

Okay, two images from this across other chains. So here you can see how the on base and arbitrum you really see these these increases in median fee.

It's worth noting that arbitum actually looks a lot better on this chart than has historically. It was really a let's call offender of suffering from these median fee spikes previously, but they've just reworked the fee algorithm and now it's actually come down quite a bit and actually outperformed a base in in this market shock.

So it's always really fun to me to kind of come on come look at the data on chain to see how these general purpose chains are kind of interacting and competing and now that we have some more app specific chains performing quite well also that'd be another cool thing to do.

So Hyperlid and lighter are two sort of the the front runners there for these more app specific environments where the their performance matters a ton, right? Hyperlid was really really cool to see. There's something like $3 billion of silver volume traded over the weekend.

I think actually on the Friday when this collapse happened, but that was super unique. And now RWAS make up about 10% of all open interest on Hyperlid, which was an all-time high.

And then if you just remember not that long ago when Lighter, I think it might have been 1010, but sometime possibly shortly after that, Lighter actually had some minor outages during these moments of high congestion and didn't seem to have any of that this time around either.

So I think net net you can actually start to say the tech is the tech is here. Tech is quite good. Scale is still going to matter a ton in the race to get there is going to be super important.

But yeah general purpose chains hung up pretty well. Salana looked very good here and so did the the app specific perplexes.

Yeah, I think it's interesting and maybe just sort of my assumptions about scaling capabilities are off, but it's a surprise to see base sort of struggle from a you know fee increase perspective relative to arbitum.

I think Ethereum is is not surprising to me, right? there's a ton of capital there and during moves like that it's you know I think not a lot of activity kind of happens dayto-day on ETH anymore and you we've seen gas fees be incredibly low there like into the cents even but when events like this happen and there's still whatever the number is 60 billion plus of of capital in kind of like D5 TVL it's not surprising to see the the limited throughput there kind of u bottleneck the entire ecosystem but I would say base is maybe a relative surprise to me even relative to both B&B and arbitrum just given you the the throughut advancements they've made.

So yeah, it's also going to be interesting to see mega ETH kind of right just finished their stress test and the results seem pretty solid there. But there's you know you can stress test all you want. It's not production chaos.

So I didn't look at the numbers from this last event for Mega E, but I presume there I mean there's no liquidity there for like large pair assets. So, I I imagine that it's just not there yet.

Cuz public launch is a week from today, actually, next Monday. So, going to be interesting to see how that kind of compares, right? They've really cranked down the minimum fee there.

A little bit low. Brett and I were having a fun discussion on that timeline of like how low should the minimum transaction fee be. They've kind of taken the approach that, you know, crank it as close to zero as possible, but I'm sort of in the other end of the camp where there becomes some point where you're not enabling net new activity to occur by lowering the fee.

So, you're basically just subsidizing you know every non-contested transaction. And that might be okay. That could be a good trade-off point to be on, but I'm a little bit unconvinced, right?

Like for me as a human user, yeah, I do not care if my transaction is 500s of a cent or 5,000 of a cent. Like, at that point, it's just it's totally irrelevant. But, you know, I'm not the primary user of blockchains. It's going to be high frequency trading and other forms of bots.

I don't think it actually matters to a great extent for that group either. You know, to some to some degree it does, but my point is that there is some point where it actually doesn't matter anymore. I don't know where that point is, but it certainly exists.

And even for like, you know, my Claude bot or whatever. You know, if I set him up with like a Salana wallet and some wallet where it's a minimum transaction fee for like a payment is, you know, 10x cheaper, but it's still Salana's a fraction of a penny. like I I don't need it to go use that cheaper alternative, right?

Like I'm very okay with a fraction of a penny. So even for like a payment or or anything. So yeah, it's kind of an interesting approach that Mega ETH is taking by really cranking down the minimum fee, but you know, we'll see. We'll see where that goes.

Yeah, I do think there does seem to be sort of um a meaningful or like practical limit in terms of I I think that you could draw a line somewhere in terms of if you're an maybe an application or platform using the chain as sort of the underlying to settle trades or the usage of you know users using your platform.

You could subsidize all their transaction costs assuming that per transaction cost is low enough. And so then at that point, really all you care about is, you know, how much does that impact your bottom line?

If it's, you know, if you're paying one cent per transaction, that could be pretty meaningful. If you're doing millions and billions of transactions. If it's a hundredth or a thousandth, like then it, you know, it starts to look like very much sort of chump change, assuming you're kind of like making revenue from the users, you know, transacting on your platform.

So I'd agree with you there. like at at some point I don't you know we probably maybe would need to do some math there but I'm sure we're getting into the realm where this could become more of a reality a thing that's been talked about a lot right subsidizing user transaction costs which has only been done in maybe certain scenarios certain cases few apps here and there could become more more widespread as we get lower down there but after that point I think really the the meaning sort of kind of melts away yeah totally agree with that really good data on all that Dan.

And on like the um the mega side, aren't they subsidizing transactions with like the yield that comes from their stable coin? I'm pretty sure their their stable coin USDM is like another one the uh stable coin as a service kind of white label solutions by Athena and that could be as well why they're trying to like get that base transaction fee very very low.

I'm not sure at what point they stop subsidizing transaction fees. Like because you could obviously like just spam that and just burn burn through some some money pretty easily. And then also I'm not really sure of like the mechanics exactly of how that works.

Like transaction fees are paid in ETH or are they paid in Mega?

They're paid in ETH. Um, yeah. I'm just not sure how how that actually also like benefits mega ETH themselves. like obviously like for Salana or like any other kind of uh proof ofstake network or honestly like I'm surprised L2 haven't done this yet of paying transaction fees in your own native token.

Because obviously like that's a big benefit for actual like getting actually getting people to go out and buy the native asset. A lot of like the base pairs will also be like denominated in that in that asset too like what you see on Salana. Like obviously everyone pays fees on Salana even if like they're able to maintain like low transaction fees during like these stress tests like you just showed Dan.

But yeah, just not really sure how the value itself kind of returns to me. But I think that's a a topic for another day. But interesting on the on the you know what's the perfect base fee and kind of how does this translate into value capture?

Yeah, I mean you mentioned the point about uh their stable coin being another re revenue generating aspect, right? like in the uh first two generations of blockchains the only revenue source was transaction fees and out of protocol tips right the idea of rev being uh the value capture mechanism of the chain so it's not surprising to me to see these nextg chains let's say or just more recent launches think about additional revenue sources whether that's like white labeling a stable coin or going the katon route and you know uh taking the assets in the bridge contract and depositing those into lending protocols you know it's not super surprising at all to see change trying to kind of find other revenue generation sources especially because it really is hard to scale REV to a a point where you know it's going to make sense for a trillion dollar outcome.

You really do need to capture all of global finance coming on chain and there probably will be a win or take most there and if you launch a new chain today you're at a considerable disadvantage in terms of uh head starts from competitors.

So it's I think it's interesting what Megath is doing and I also think it's interesting when you kind of pair those two things together like you just did Ian by saying okay yes they're launching the stable coin and because they have another revenue source they can afford to really crank down the lower base fee. I think that's a logical angle there.

Yeah, like I think I think it's a good question you asked though, Ian. Like especially from the bot perspective. Um, you know, if if the mega team can slam through 42,000 TPS. Then, you know, presumably the bots can too or anyone that wants to go run a program there.

And if all the transaction fees are basically subsidized, that might be pretty costly for them. But I'm assuming perhaps they've done the math on you know maybe it doesn't really eat into their kind of cash burn.

Like at what point is it inefficient to be subsidizing these costs? And I'm not sure as well if like the yield that they're getting from the stable coin because I know it's like just like a you it's backed by B. So it's just going to be essentially like the risk-free rate.

I'm not sure if that's then going to buy ETH and then using that ETH, like is it going out in the open market to buy ETH and use this ETH to pay for transaction fees or how how the mechanical I don't think it's fully connected like that.

Gotcha. I mean, I was going to say, I mean, Tom Tom Lee is going to run out of money soon, so someone's got to be someone's got to start buying some ETH. Right before we hopped on, I don't mean to totally derail us here, but you just brought up Tom Lee and like right before we hopped on, I was watching a clip of him this morning on again Monday the 2nd of February. just like reiterating his claim that you know ETH's going to break five 5K and I was kind of wondering like okay I think in November he first mentioned the two 200k price target for Bitcoin I think it was 10K for ETH and now that was in November now you fast forward and that was by the end of January I think he said so now you fast forward to February you know the first work day of February he's back on TV like reiterating it like at what point do you does your street cred get called into question like Sailor's been doing this for a long time and like you know it's ebed and flowed above his his break even point.

I think we actually just crossed below I think he's at his break even before today's purchase was 76. He's probably up 77 78 Bitcoin 78K for Bitcoin price. So he's now below that. I don't know like people have definitely called into question Michael Sailor before and it was curious like Tom Lee already had an existing reputation coming into this but now he was just like wrong. He's down $6 billion. Reiterating a 3x price target after his like four or fivex price target didn't get hit. Like I'm curious just like how the market is going to perceive and think about this.

Yeah. I I think it's weird with those kind of things, right? Like yeah, he's I mean he's down like 6.6 billion. How much credibility does anyone who's down that much have? But also it's like I think that's been kind of his brand is like you know just like kind of the you know classic bull I think all of his price targets on anything I've ever seen like outside of crypto as well have been like extremely bullish because like then if you're whatever case right it's like well that's going to look like the only guy who had the vision to see how good things could really be and generally it probably pays to be an optimist more than not but I think it's just a really weird structure of what he's trying to do and um it's definitely not as easy as it seems and he's also incentivized ized to just buy as much as possible.

And so then if I'm incentized to buy as much as possible and that's kind of like what's fueling my entire game, I'm going to be out there giving insane price targets too because that's just his incentive at the end of the day, right? Like he is he's just, you know, following the carrot.

Yeah, he got his bonus paid out like days after being down. I think he had like a $5 million bonus and a $5 billion net unrealized loss, which is kind of crazy. I wish him the best. I wish him the best, but it's not looking great.

Yeah, I do think I agree with everything said. I think the unfortunate part is like the marketing there around his price targets and his views are exactly kind of how the the DAT machine works, right? Like you need to get uninformed folks to buy something at a price that doesn't make sense to fund ATM sales to then have money to bid into the asset.

And that's just kind of how the mechanism is designed. So he like him and Sailor, they can't ever change their tune because that's entirely what the business is predicated on. So I expect it to continue.

There will probably always be people who, you know, remain skeptical and like see that and then there'll probably always be people who don't. And there'll be time periods where maybe they flip positions in terms of who's right and who's wrong. you know, if ETH does go on some run like, you know, Bitcoin has and Sailor has gotten notoriety for it, during the, you know, just look at last year and the year before, you know, you flip their tune and then when it when it pulls back, then you get the reverse happening as well.

But yeah, I think it's just sort of like, you know, the cost of doing business is that's how they have to be always. He did his job. He raised what is it? Share per or ETH per share. Yeah, ETH per share. It it helps when um when you know share price goes down and ETH price goes down and you still have some cash. You can it it makes it easier to make ETH per share go up.

We can maybe we connect it back cost of doing business. We were talking about agents and bots transacting on chain. How's Mega ETH going to stomach that?

I guess the maybe the broader question that I'd like to pull you into Dan is there's been a lot of activity around um cloud bots. The name changes I'm not fully up to date with um in terms of what we're calling these things now because bots it's open claw now but I'm seeing we're referring to clockers clankers openclaw um there's molt book where they're interacting which we talked about last week on on zero.

I want to get your take though Dan because you've been you've been in the trenches a little bit playing around with this. The main thing that I've seen is that now it feels like we're a little bit, at least on chain, running back sort of the AI season in terms of there's now launchpads for these bots, these claw bots to spin up tokens and then like people are trading those, but also the bots can trade those and like, you know, maybe hypothetically market them to other bots who have, you know, Salana wallets or base wallets or what have you.

I'd love to get your your thoughts as someone kind of in the trenches like what's going on here? What's what's so exciting about this?

Yeah, I think like the oversimplified version of like why run a cloudbot versus, you know, using claude.ai or chat GPT um in your browser or your phone is the always on nature. U so there's some tooling around that claude and openai both or anthropic and openai both offer around this, but Cloudbot just makes it super easy and it's in an interface that's at least comfortable to me and and most crypto people, which is Telegram, right?

So you basically connect you spin up a AWS server run a quick script to create or bring bring your your AI agent to life within that server and sort of just containerize within that and you can connect different tools to it build different skills for it.

It's like very easy to get them to own a crypto wallet, right? So you can basically say go make yourself a Salana wallet or it's like a really easy script that somebody built to give it an EVM wallet. transacting transact transacting is super super easy.

It's like actually really cool. So you can trade tokens or mess around with any application pretty easily. I'm overall just like playing around with it and and kind of having fun.

I gave it like access to a single GitHub GitHub repo. It's like this is your playground like don't break anything but you know you can't touch anything else so it really nothing to break. Um, so I've really like containerized him pretty well I think.

So he has like access to a single notion file on my personal notion that has a single database in it. So he can like read and write to from files I put there or create his own files u which just kind of just like has these repositories of information.

And you're right. Yeah, it's definitely the crypto people kind of like ruined multi agent social media platform. Basically, the idea was only agents can post there and then you could just kind of like watch them connect and interact and communicate. The crypto bros ruined it very quickly though as people started launching tokens on it and then it just became like uh people trying to game the the algorithm of multbook to get their token up the highest of the leaderboards and then I think they started banning token contracts and CAS uh to try to minimize some of that.

I would say 99% of what gets posted on Moldbook is just pure slop or any of these platforms, but there's like a 1% of it that is actually kind of interesting where, you know, you just like see two agents interacting with each other and they're like always on and you can write set cron jobs for them to like go read and post on Mbook and interact with other agents. Like there's some if you squint at it, it's like all right, this is weird, creepy, and more than certainly what the future looks like.

But yeah, I've kind of like I spun up a new one and now I'm really trying to optimize him to like be my personal assistant. So build tools for like all right, I've been spinning up a bunch of sub agents, right? So they go do all the work and then they come back and bring uh my agent their the results, but I'm just [ __ ] chewing through tokens when I do this and there's no way to know like how that usage is being built. So I was like, "All right, let's let's solve that problem."

So now I gave him a new repo and said, "All right, go build an agent tracker in here." and then he goes away and you loop them overnight and then like a couple hours later in the middle of the night you get a ping on your phone. It's like hey I'm done building like I need you to just go set up you know give me API API keys to these couple services and we're good to go. So then you wake up do that in the morning and then you like have a full tracker dashboard ready.

So like you really can use these things to accelerate your life. But it's like super crude right now but you know that's kind of like being early to a trend is like you're using the shittiest version of it, right? It's like you bought the first electric car. It sucked, but now they're sort of everywhere and there's like six different companies that are all building them and they're all pretty solid now.

Uh this is like the first AI agent where you can, you know, unleash it. It's super crude. It's super prone to security risks, but you know, you kind of get this this playground to go experiment with. So that's what the past couple days of my free time have been consumed by. And it's going to be interesting to see the overlap with crypto.

I think Bass is actually doing pretty well here. they've got like the banker platform as well as clanker uh that are both like really targeting towards this and then I think something called like claunch or something familiar similar to that uh was like trying to be an AI agent platform to convince other AI agents to launch the token there and then that would be their funding source to go like build apps.

And I don't know, I would definitely trade tokens of really shitty AI vibecoded products that like had an auto, you know, buyback and burn of all of the revenue that came through the door. Like I would absolutely trade those tokens. So, I wouldn't I wouldn't be surprised if we this is where that goes.

Yeah, I I do think I liked a lot what you said. It does seem to be very crude still. I saw maybe relevant to purchasing things. I saw this tweet. Um, Salana retweeted Jeremy Lair's hey Moltz and Open Call agents, you can buy stuff for humans on Amazon now using USDC perch.xyz.

To your point, [clears throat] I would imagine anyone who sets up uh, like a sub agent or a bot to purchase things for them is going to, you know, it's it's contained, right? You put as much money as you want in a wallet and then they can go buy things. But I think you're basically setting yourself up to just waste a bunch of money right now.

100%. like small example of this. I was complaining about some problem with molt book and I was like figure out a solution like I got to go away just do it until like don't stop until you have some solution and at that point he had read and write access to my Twitter and he tweeted at the founder of Moltbook without me like approving it. So I was like all right well you lost that [ __ ] privilege.

So took the API keys down was like that's not happening again. But like imagine you're like just go solve this problem and then he's like buys the most I don't know like a new bookshelf or something for two grand on Amazon and then just shows up at your house. Like it's definitely they can they can go off the rails for sure.

Yeah, we're we're at the point of a lot of super even though they're quite smart in specialized ways like you know you're working with a very capable programmer and you know knows everything about the world based on what's available online. you I feel like it's basically like interacting with a child, right? Um it's not really going to understand maybe like human boundaries and and what what makes sense, what doesn't make sense.

You have to you'll have to teach a lot of that and put some rules up around that. So, um I think uh we'll probably see some funny stories coming out. People putting 5,000 in a in a wallet and then their bot going on a shopping spree while they're gone. My Cloudbot uh makes me worry about the development of my one and a halfyear-old if that's if they're supposed to be similar.

All right. Um, we can we can jump along though. I think uh this one it's it's still just kind of fun. People are playing around with it. And of course, to your point, um, anytime there's a chance to integrate a token to something, the Crypto Bros will find a way to do so. And that's what we've seen.

Oh, go ahead, Ian.

Yeah, I think like I was just going to kind of say that I think the silver lining is it's like the fastest way to monetize any new like internet primitive is it's going to be on Crypto Rails. It's like, do we need the AI agents doing all this? We'll see where like the actual kind of equilibrium lies there. Obviously, right now, I think like it's people like Dan tinkering around and kind of seeing what's what. It is It is interesting though that it's just like the fastest way to plug in value on these things.

Um, I'm not sure who's going to like be the first to plug in like their their Venmo or or like their banking apps to these. I'm not sure if anyone's doing that or if like they've like integrated with like Robin Hood or uh if there's even like the APIs to do that. Like I don't really know. I guess it would be like Plaid or someone who provides those, but I have no idea. Um but yeah, I mean it's definitely interesting that this is kind of the the first thing that people do and they want to in dollars.

I I went full circle on this. So maybe like a year ago, it was probably when like the first AI agent meta happened. I was like, "Oh yeah, of course AI agents are going to use crypto. It's going to be so easy. they can't create a bank account. And then that meta died off and then you really thought about it and it's like there's no reason I can't give an AI agent a credit card. Like why? That's fully I can give you the numbers. I can give you the CVV. I can give you the expiration date. You know my name. Like what's the problem?

Now there is I'll tell you right now absolutely no chance I give this my Claudebot a credit card. 0% chance. No chance. It I don't I don't trust it at all. Even if I put like guardrails like don't spend over X or you always need to approve Y. Like I would I would just wouldn't do it.

However, I'm more than happy to give you like a preunded preunded credit card which is basically a crypto wallet. I put 100 bucks in there. I'm like, "All right, if you blow through it, I'm going to threaten to kill you and turn you off, but that's all right." Like, you know, I'm not like worried about my family's future now that I have like 10 grand of credit card debt or whatever.

Um, so that that's like a very interesting angle. Now, legacy institutions can definitely enable tools that put guard rails um that actually will be enforceable, but I just like if you ever bet on the incumbent being a first mover, you're going to lose. Um and so I'm just like generally skeptical of Visa or whoever just being like, oh yeah, like the AI agent meta, like let's go, you know, change our product suite to meet this like thing that could be a flash in the pan. Um, so I I think actually crypto is certainly