YT YouTube2Text

YouTube2Text → Transcripts → Where Is The Crypto Alpha Hiding This Cycle?

Where Is The Crypto Alpha Hiding This Cycle? — Transcript

by Forward Guidance · 6,745 words · 944 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Nothing said on Ford guidance is a
0:02recommendation to buy or sell any
0:03investments or products.
0:11All right, everybody. Super excited
0:12today to be joined by some of my
0:15favorite allocators in the industry here
0:17to talk a bit about where the alpha is
0:20in this cautiously optimistic but
0:22potentially early bull market that we
0:24have going on here. Um, yeah, why don't
0:26we go on the line here? Just feel free
0:27to introduce yourself. Um yeah, maybe
0:29mention your your style of fund and what
0:31you focus on and we'll start with you
0:32Arthur.
0:33>> Yeah. Uh I'm Arthur. I'm the founder and
0:35CEO of Defiance Capital. Uh we are
0:37crypto liquid funds that's been
0:38investing in the space since 2020. Uh
0:41classic liquid venture style. We buy the
0:44crypto of um the ad that we have a lot
0:47of promising potential. Um and yeah,
0:50stay in for the ride. Yep.
0:52>> Hi, I'm Rajie. I'm a GP of Framework
0:54Ventures. for SF-based investment firm
0:56um that mostly does uh early stage
0:59venture and some liquid token investing.
1:02>> Yeah, I'm smack uh I'm a partner at
1:04Compound. We are a thesis driven
1:07research ccentric investment firm
1:08typically early stage venture um and
1:11more recently spending time on uh
1:13publics in equities and and tokens as
1:16well. So generalist fund based in New
1:18York probably a third of the fund is
1:20cryptoreated.
1:22>> Awesome. Um yeah, why don't we get into
1:25where the emerging sectors are and
1:26what's interesting. Um obviously the
1:29first one is AI has obviously been the
1:32the you know a lot of excitement there
1:34over the last year and starting to see
1:36that shift towards excitement in crypto
1:38and obviously the the intersection of AI
1:40and crypto is there's there's a lot of
1:42interest and excitement there. So just
1:44love to hear a bit about uh that
1:46specific niche sector, how the three of
1:47you are are thinking about it, how
1:49you're thinking about the alpha there
1:50and uh the opportunity set. Arthur, why
1:53don't you go ahead?
1:54>> Yeah, so I think the initial phase, I
1:56think since 2023, um I think that was a
1:59bit of struggle. There was a lot of uh
2:01force fitting of a crypto and AI theme,
2:03but I think over the past one year, we
2:04start to see more and more things that
2:06make sense. Uh I think this is probably
2:08not a very well-known fact but uh the
2:12developer of the largest open-source uh
2:14AI agent framework uh Hermes is actually
2:17built by uh news now research that was
2:20actually kind of building something on a
2:22crypto AI side as well and I think that
2:23now they are like being considered a
2:25very legitimate AI team as well and so
2:28those are one of the example um we also
2:30saw like uh venise AI which is like
2:33essentially like a crypto version of
2:34open router where you can route uh
2:36inference on a uh with privacy focus uh
2:39through it and I think that that
2:40actually has been really taking off. I I
2:42think that obviously if you compare them
2:43to open router like um the the token
2:45span and inference is still not as big
2:47but the growth chart is still a hockey
2:49stick. So I think that this is some of
2:50the prime example that we are seeing
2:52right now. Yeah.
2:53>> Yeah. So I think there's a few things.
2:55Um one is that um AI is obviously uh
2:58transforming the way that companies
3:00operate. So, we're seeing sort of um uh
3:04operational leverage that uh we haven't
3:06seen previously sort of for early stage
3:08companies where they're able to make a
3:09lot more progress without needing to
3:11hire a lot more people um or potentially
3:13spend on on headcount. Um on the crypto
3:16side, I think in particular, we've been
3:18really excited about financing
3:19opportunities. We think um crypto is
3:21really good for capital formation um and
3:24uh sort of with the whole AI capex
3:26buildout, there's lots of different
3:27areas that that need to be financed. One
3:29of our portfolio companies, USDA, is
3:31doing like mid-market cluster financing.
3:33So, if you, you know, want to um if
3:36you're you're a promising startup or
3:37even a public company, um, and you want
3:39to buy, you know, $150 million worth of
3:42GPUs, um, you can go to them and
3:45potentially get more attractive rates
3:46with like less of a down payment
3:47relative to what you might see
3:49elsewhere.
3:50>> Yeah. I I So, we've been investing in
3:52traditional AI and ML since 2016. And so
3:56for us, one of the things that when we
3:58look at kind of crypto/ AI or or where
4:01that overlap is happening, a lot of it
4:03comes down to like the talent level of
4:06the builders there. And so I think
4:08what's kind of interesting is we've seen
4:11some of the overarching ideals of kind
4:15of crypto start to permeate more into
4:17these AI systems and this this AI kind
4:19of uh narrative, which is a lot around
4:21permissionlessness. Um, and so to me,
4:26I think we think that if crypto AI is
4:29going to compete on efficiency today,
4:31that's probably not the place that makes
4:33the most sense. Um, at least today. I
4:35mean, maybe that's not true in the
4:37future, but I think there's a lot of
4:39kind of underlying societal trends that
4:43um that make the things behind crypto
4:45more appealing. And that's kind of
4:47around coordination. I think some of the
4:50economic models could be different. And
4:51then again like permissionlessness
4:53um compared to the you know centralized
4:55AI and and some of the centralized
4:57frontier labs and what they're doing.
4:59>> Yeah. I think one thing I would like to
5:01add a little bit is like um crypto
5:04usually all the crypto AI project kind
5:06of enforce encryption and privacy by
5:09default and I think that's a big
5:11difference compared to all the AI
5:12project where they are really in a stage
5:14where they just do everything first and
5:16then don't really care about the
5:17privacy. I think that um the in the
5:20world in general I think is going to
5:21care more and more about this and I
5:23think that they will found they will
5:24find a solution that already exists in
5:26all the crypto AI project. Yeah.
5:29>> Um yeah I'm always hesitant when I hear
5:32people talk about this idea of uh you
5:35know crypto was never m you know it was
5:37never built for humans to utilize. you
5:39know, there's there's this I don't know
5:40if it's just like forcefitted in in
5:42retrospect, this idea of okay, crypto is
5:45actually going to be used by AI agents
5:46into the future. Curious how the three
5:48of you think about that premise. Is that
5:49just sort of, you know, hindsight bias
5:51or is there something there?
5:53I think it's so I've always thought like
5:55crypto is especially good at
5:57coordination and then kind of like
5:58economic incentives and as blockchains
6:01have got more efficient or more
6:03performant I think there's the argument
6:06I would say on the crypto AAI side is
6:08cool machine to machine transactions and
6:10communication
6:12um maybe centralized systems that you
6:15use a credit card and you can have
6:16agents talk to each other but there's
6:17some minimum floor to actually do those
6:20transactions and I think the argument
6:21for crypto AAI is okay, cool. Can we
6:25smush the actual cost of doing this
6:28towards a level that's more competitive
6:30that a centralized provider can't
6:32compete with? Um, I think that's TBD on
6:35if that is actually possible. But, um,
6:38to me it comes back to like coordination
6:40more so than anything else. Like that's
6:42especially what crypto has been
6:43historically good at. Uh, and then also
6:46like the global nature of the
6:47permissionlessness. I think people are
6:49like Arthur said much more concerned
6:51about privacy and kind of like where
6:54their data sits and who owns it. Uh and
6:57yeah, I think there's there's definitely
6:59an underlying trend of push back on
7:01these kind of frontier labs and AI like
7:03the traditional AI model. Um so I I do
7:06think there's like opportunity for
7:08crypto somewhere in in those wedges.
7:10>> This episode is brought to you by
7:12Fidelity Crypto, a platform built
7:14in-house with the same discipline.
7:15Fidelity applies to everything, so you
7:17can invest in crypto with confidence.
7:19Free crypto backed by industry-leading
7:22security. Get started at
7:23fidelity.com/crypto.
7:25Crypto is offered by Fidelity Digital
7:28Assets, NA, is not insured by FDIC or
7:32SIPC, and includes risk of significant
7:34loss. Fidelity Brokerage Services LLC,
7:38member NYC, SIPC.
7:43>> Yeah. One idea we've been toying with
7:45recently is like sort of with all of
7:46these agent swarms that have been out
7:48there um the idea of like a distributed
7:52um permissionless ledger um becomes
7:55potentially very interesting. You can
7:57imagine um for for better or worse um
8:00you know if a bunch of agents need to
8:02communicate with each other and need to
8:03make sure that um their messages are are
8:06are durable or available um blockchains
8:09actually might make a lot of sense
8:10whether you're cramming it into um
8:12something uh with high data availability
8:14whether you're using persistent um file
8:16file storage um I think it would be a
8:18very interesting experiment for somebody
8:20to like crawl um arweave or filecoin and
8:23see if like uh any of these agent
8:24swarmms are already using it
8:28I think that I don't have a strong view
8:31on this but I do think that there's
8:33obviously a two-sided argument right so
8:35people say that oh AI agent are just
8:36going to be able to use all this like
8:38blockchain based payment and like you
8:40know transaction a lot more easily than
8:42human and then you obviously people say
8:44that you can also use the API what's the
8:45difference I do think that the
8:48permissionless nature means that it's
8:49easier for a rogue AI AD to exploit so
8:54it's it doesn't really ing the best
8:56scenario. But when you think about it,
8:58like if a AI agent go rogue, is it going
9:00to use a bank API or a fintech API is
9:03going to use crypto? I think it's
9:04definitely going to be crypto. So, um I
9:07think that
9:09there's just so much possibility. I
9:11think you cannot write it off, but I
9:12think I think that the permissionless
9:13nature combined with the AI like they're
9:15just going to be going to create a lot
9:17of like the um uh the the the possible
9:21um result that we are not able to
9:22foresee. But I think that there's
9:23definitely something that can be done
9:24here. But I think it's still too early
9:26to see. Yeah.
9:26>> Mhm.
9:27>> I think also just like one other point
9:28around this is if you think the future
9:30is going to be more
9:34local models or smaller models, not
9:37everything being done in these like
9:39massive kind of frontier models and more
9:41of the work or more of the inference
9:44gets done at distributed versions of
9:46that. you know, I don't whether or not
9:49there's value to be had there or
9:50captured from crypto companies or
9:53projects is I think an open question,
9:54but directionally
9:56it's I think it is again like in the
9:58ideals of some of the early kind of core
10:01beliefs of crypto around you know
10:04distributed networks and um privacy to
10:07some extent. Um so yeah, I mean TBD
10:12though the value capture is always like
10:13an open question with a lot of these
10:15things. Um, all right. I want to look
10:17into
10:18privacy, but through the lens of the
10:21modern crypto fund allocating towards it
10:23and using that as as a jumping off point
10:25because I mean the three of your funds
10:26kind of represent this idea of this uh
10:28this barbellness of you know you have
10:31the you have the venture book, the more
10:32illquid book and then you also are
10:34actively on the on the liquid side of
10:35things. And that's it. It's it it's
10:38quite an interesting, you know, when you
10:40just look at the form factor of what
10:42crypto funds have done well and and are
10:44still here and are still allocating, it
10:46seems like that barbell is a nice place
10:47to live at. Um, and then you put that
10:50into, you know, this privacy narrative
10:51has been something that's really taken
10:52hold recently. And what's interesting
10:54there is that a lot of the the
10:56opportunity set there are assets that
10:58anybody can go buy in the market. You
11:00know, like Zcash has been around for
11:02ages. um we've seen near near intense
11:04start to go off. Um there's obviously
11:07you know anything related to to the zero
11:09knowledge proof world of there. So you
11:11know there those are assets that you can
11:12go out there and buy in the market
11:13today. But allocating in size is is a
11:17difficult situation there compared to
11:19the allocation through a venture book
11:21where you can size into it very easily
11:23and not be as worried about liquidity.
11:25Obviously you have to worry about it at
11:26the liquidity event. But, uh, yeah,
11:28using that that that case study there,
11:30I'm curious how the three of you think
11:31about why we've ended up in this style
11:35that seems to be just, uh, you know, the
11:37best way to go in crypto these days.
11:39Start with you, Arthur.
11:40>> Yeah. Um, I think that this is, uh,
11:43never really been a big issue for if you
11:46have hell conviction. I think that, um,
11:48we have seen people being able to
11:50accumulate like I think a few percent of
11:53any protocol if they want over a long
11:55period of time. And I think that uh
11:56given the execution um um tools
12:00available, I don't think it's really an
12:02issue except if you are like more than
12:04500 million or a billion dollar and
12:05above. Um but that say I think that you
12:08know that's that's why I think
12:09historically a lot of people just like
12:10venture because you don't have to think
12:11about it that much and then once you
12:14like you know start to go into the
12:15execution part you start to compete the
12:17market makers and then it's a lot more
12:18tradingish focus. But I actually think
12:21that um for people who are because I
12:23think that ultimately for venture you
12:24are underwriting the volatility of the
12:26outcome. It's just that you do not see
12:28the volatility on a day-to-day basis. I
12:30think that just for a lot more um like
12:32the people who are allocating it on a
12:33liquid side, they just have to be
12:35willing to not pay attention to the
12:37day-to-day volatility as much uh while
12:39they are building the position. So I
12:41think that we basically budget for like
12:43uh 5 to 10% uh variance in terms of our
12:47final entry price when we are building
12:48any position. And it's also for exit as
12:50well. I think that sometimes like it's
12:52easiest to think that you're not going
12:53to time it too well um because you're
12:55not uh your age is not in a short-term
12:57trading the day-to-day market movement.
12:59So just budget yourself for like a 5 to
13:0110%. I think 10% is probably a bit too
13:02much. We usually want to get like
13:04between 5% variance. But yeah, I think
13:06that if you're okay to pay 5% stippages,
13:08you shouldn't really have an issue
13:10building a position. And that's that's
13:11generally been our approach.
13:14>> Yeah. Um I mean the way we look at
13:16things is like regardless of whether
13:18it's um a liquid asset or or sort of on
13:20the venture side, we are looking for
13:22venture style outcomes for for the
13:23things that we invest in. And that often
13:25means uh being willing to um hold an
13:28asset over a very long time horizon and
13:30sort of uh realize some of the
13:32volatility and and and hopefully
13:33convexity sort of sort of with the
13:35asset. Um I think you yeah like sizing
13:39into assets is is always a challenging
13:41thing especially if um the liquidity is
13:43not great. um um the team may or may not
13:46be willing to OTC. Um there's also a lot
13:50more crowding these days in the market
13:51where things become consensus a little
13:53bit more quickly and I think people or
13:56the funds and and market participants
13:58that that that are here are a bit more
14:00discerning around like what is working,
14:02what isn't, what actually like makes
14:03sense to invest in and what doesn't. So
14:06um the prices sort of when you get
14:07excited about something can move very
14:09quickly because it's very likely other
14:10people are doing sort of similar
14:11homework. So a lot of it is just being
14:13comfortable and understanding at least
14:15for us what the upside case looks like
14:16and um similar to Arthur like you know
14:19if there's some variance um uh sort of
14:21like in our entry price like we we are
14:23okay with with that.
14:25>> Yeah we've we've so
14:27historically um we were primarily only
14:31kind of private venture seed and preede.
14:33I think there was a point a couple years
14:35ago where we said um hey we're still
14:38excited about the asset class and the
14:40technology and we're not meeting enough
14:44founders that we could get to a yes on
14:46uh in terms of like early stage
14:47investing and so we said hey we should
14:49be spending more time on the liquid side
14:51mostly as a function of wanting exposure
14:54to the asset class um which again if
14:57you're kind of periodically only doing
15:00private preede and seed stage like maybe
15:04we only do four or five transactions a
15:06year and so it's like the asset class
15:09moves faster than that and so we shifted
15:11kind of the belief and I think part of
15:13this is um just like the overall market
15:16maturing as well is okay I think there's
15:18more probably investable assets today
15:20than there were five or 10 years ago and
15:24because of that I think shifting how you
15:26think about the asset class matters but
15:29on the sizing side we again like even
15:31when we think of structure For us, it
15:33was very much cool, what do we think is
15:34like the proper size that we could own
15:36assets to Reie's point where our view is
15:39these should be venture style returns
15:41whether they're liquid assets or not.
15:43And so like what is the size that we
15:45think we could reasonably deploy uh and
15:49generate those type of returns and and
15:51do it in a way that's not kind of beta
15:53because I think that is the other thing
15:54in crypto is like it can just turn into
15:56a levered beta asset class which is um
16:01yeah which is like kind of not as
16:03interesting to us. And so our view is
16:04like okay cool can we generate alpha
16:06still and like what's the size
16:08appropriate to do that? And a lot of it
16:09does come down to the volume of these
16:12things traded, like what assets we
16:13actually want to own. And then yeah, the
16:16VA is so high that I think there's a
16:18little bit of like, okay, if we want to
16:19build a position over time, if you have
16:21some patience, like you kind of have to
16:23believe that the volatility is going to
16:24work both ways. Um, but yeah, obviously
16:28sometimes you see these assets just
16:30trend super hard and
16:32>> you kind of have to be comfortable with
16:33that, I think.
16:34>> Yeah. Take a disciplined approach to
16:35crypto investing with Fidelity Crypto, a
16:38platform built inhouse using over a
16:40decade of crypto experience. At
16:42Fidelity, you can trade crypto and
16:44stocks together, backed by
16:45industryleading security. When you can
16:47see your accounts in one place, it's
16:49easier to make smarter decisions. Get
16:51started at fidelity.com/crypto.
16:54Crypto is offered by Fidelity Digital
16:56Assets NA is not insured by FDIC or SIPC
17:01and includes risk of significant loss.
17:04Securities offered by Fidelity Brokerage
17:06Services LLC member NYC SIPC.
17:13What do you guys what do you guys think
17:14about this? you know, Rejie, like what
17:16you're saying about this idea of
17:18dispersion leads to crowding outness
17:20really quickly because there's only a
17:22handful of assets that, you know,
17:23suddenly everybody latches on to because
17:25there's just a plethora of, uh, assets
17:28out there that, uh, you know, the
17:30fundamentals aren't as strong. Um, so
17:33intuitively, you would think dispersion
17:35is great for active managers because,
17:37you know, as long as you're selective
17:38and in the right assets, um, you know,
17:40you can go to your LPS and say like,
17:42look at this alpha that we're
17:43generating. Um, you know, it's not
17:45necessarily just levered beta where
17:46we're just, you know, everything's going
17:48up at the same time. But to your point,
17:49the crowding out of it can also be
17:51difficult to manage. I'm curious, it
17:53feels like this is the first bull cycle
17:54that we're going into where there is
17:56significantly more dispersion. It's not
17:58just everything all up, all down. How do
18:00how are the three of you guys managing
18:01that?
18:03>> Yeah, I can start. Um, I think some of
18:06it is like, you know, time timing timing
18:08does matter. If everybody has very
18:10similar ideas, sort of how you time it
18:11uh probably matters a little bit. Um uh
18:14timing comes in a bunch of forms. It
18:16comes around like how you think about
18:17entry. It comes about like you know what
18:19your time horizon kind of looks like.
18:20There's things that might look really
18:22attractive and end up being really
18:24consensus and crowded over a short
18:26period of time but over a longer period
18:27of time maybe are are a little bit less
18:29consensus. So I think that's sort of um
18:32some of how we have looked at it. Um
18:35uh yeah I think I think we're still uh
18:39sort of seeing how all this is going to
18:40evolve. that feels very different than
18:42um previous cycles. Um and there's still
18:44a lot to learn there.
18:47>> Yeah. I mean I I would say like it
18:50shouldn't be easy. I think that's like
18:52something that in the past for crypto it
18:53was kind of like
18:55not a high level of sophistication, not
18:58um there was like almost this influx of
19:00capital and so there was a period where
19:02it was kind of easy. you could just
19:03throw a dart and no matter what you
19:04bought it went up and I think the
19:07maturation is a little bit of okay cool
19:09it's becoming more difficult to get to
19:12assets early um size them accordingly
19:17mostly a little bit of that is
19:18sophistication but I I think like you
19:22have to you have to in my view part of
19:24it is okay do we believe like what are
19:27the highle things we believe the future
19:29looks like 12 months from now 18 months
19:31from now 24 months from now. And I think
19:33crypto especially like it almost seems
19:36to happen quicker. So like you could
19:38have a view of hey we want to own this
19:39asset class and here's probably where we
19:41would like to exit it and that cycle is
19:45much quicker than kind of public
19:46equities or or other asset classes. So
19:48like all of the moves can happen in 3
19:51months or 6 months and you have to be
19:52comfortable saying okay is this just
19:56crowding out? is this like is the thing
19:58we believe happening quicker or are more
20:01people starting to believe it and so I
20:02think a little bit of it is you know
20:04what would we need to see once we've
20:06entered the position of like when should
20:09we be exiting so I I actually think this
20:11is the best uh time to um for us in the
20:15recent years because you know um for
20:19active managers we want uh dispersion
20:21right and this is what we are getting
20:22paid for right so I think that this is
20:24also a year that where I think on the
20:25street I
20:27probably more than 80% of the liquid
20:29funds I know hedge fun liquid funds I
20:31know on crypto they outperform Bitcoin
20:33ETH and so because I think year to date
20:35they are still down and I think year to
20:37date most of the crypto funds I know is
20:38like up like from 20 to 100% um and and
20:43this is also a good way for us to
20:44justify why you know you should get an
20:47expert to manage the the book for you
20:49and I think that yeah this is a
20:51extremely healthy phenomenon and I think
20:52that it's also kind of um how fast the
20:56market have rewarded like the good
20:57fundamentals. I think that um you know
20:59previously there was a little bit of a
21:01more of like a order book positioning
21:04driven like oh like hyperlquid it was
21:05been a consensus name for more than a
21:07year and previously when the name become
21:09very consensus in crypto it tend to
21:12start you know not doing well and peak
21:14but this is like I think and also
21:16obviously depends on the market cap but
21:18hype has been like a huge market cap
21:20coin for like quite a while and and is
21:22very consensus and yet it still continue
21:24to go up and outperform so I think
21:26that's actually very healthy like it's
21:27Just because u an asset is consensus um
21:30amount of smart money doesn't mean it
21:32can't outperform anymore and as long as
21:35it has great fundamentals great growth
21:36ahead. So I think this is something that
21:38is is kind of like what I've dreamed
21:40about since like four five years ago and
21:43finally playing out. Obviously that we
21:44wish there's a lot more project like
21:45hyperlquick that we can invest in but I
21:47think that this is like a incredibly
21:50encouraging. I don't think that the
21:51market is like very institutional yet um
21:53but I think that you do get more
21:55sophistication um in terms of the market
21:58participants. I think that you're also
21:59getting some bifocation where um a lot
22:01of the retail they they are not happy
22:03with like making like let's say 50 to
22:06200% in a year anymore. They really want
22:08their 10 20x and they know that they
22:10can't get it in all this so called
22:12midcap like a smart money coin. They all
22:14go down to the extreme left curve end
22:16which is all the meme coin stuff like
22:18they paying like 100k to like 2
22:21three million or like 50 million market
22:22cap. um when all the you know people
22:25with size that don't really bother to
22:27play this game we all just like
22:28congregate to like either you do the
22:30majors or you do the midcap good
22:32fundamentals so I think um you you get a
22:34lot more um I think that like consensus
22:37in terms of smart money and I think
22:39that's been very healthy and I would
22:40love for this to continue to play out I
22:42would love to outperform Bitcoin every
22:44year and that's how I will be able to
22:45like you know like raise more money and
22:48etc. Yeah, I I would say like one other
22:50point is I I think you have to also at
22:54least try to think about or model out
22:56who like who will be the marginal buyer
22:59of the things that I'm buying today.
23:01>> Um which again to Rajie's point like if
23:04you're looking at something and you're
23:05seeing oh the fundamentals are
23:07inflecting and um you know this thing
23:10seems good to me. I think there is like
23:13the the the speed at which that idea
23:16usually compresses is much quicker today
23:18than it was in the past where if you
23:20you're seeing something that you think
23:22hey the fundamentals are lining up I
23:24think there's general trends that this
23:25will benefit from over time like that I
23:29think you do have to have an idea of
23:30like hey let me own some of this thing
23:32and then do more work on it and size
23:33into this over time because it just
23:35happens way quicker than it used to. I I
23:37think one one last point I want to add
23:39is like
23:40for both uh crypto and probably uh
23:43equity this is probably the golden age
23:44of a longshot strategy. Yeah, I think
23:46that you usually historically when you
23:48do long shot you usually expect to make
23:50money on one side of your trade like
23:51your other side is just like you know
23:52you want to be delta neutral. Uh but you
23:55know you are usually going to lose money
23:56on like on a short side if you're like
23:58in a bull market but I think this year
24:01in both agree and crypto I think you can
24:02actually make money both on a long leg
24:04and a short leg because there are some
24:05project are really not making it and
24:07like getting close to shutting down and
24:09you're shorting those you actually going
24:10to keep getting like a 50% down every 6
24:13months and then yeah you can actually
24:14keep shorting obviously you need to
24:15close the trade and reshot it again. Um
24:18yeah, but or like you can even do a
24:20cross like industry kind of like like
24:22your long equity, short crypto or your
24:24long crypto, short equity and a lot of
24:26these like fundamental thesis really
24:28playing out and they they just keep
24:29trend all the way up like like nobody
24:31expected the treasury to like keep
24:33dropping like people like if you look at
24:35technical is so over. So you
24:37should like stop. Nope. It just keep
24:39going down, keep going down like the
24:41fundamental driver just so strong it
24:43kind of overpower all the positioning
24:45driven like a factors and it's kind of
24:48like for all this like a fundamental guy
24:50and we were surprised how well the
24:52fundamental driver has been playing out
24:54to drive the price action as well.
24:56>> Do you guys think that Bitcoin is still
24:58the correct benchmark amidst this? I
25:00mean you can just see so much sectoral
25:02dispersion. I always wonder like
25:04obviously it's it's been the benchmark
25:05in terms of alpha now performance but is
25:08that is that still correct?
25:11>> It is still the best benchmark we have
25:13uh for most of the LP because they still
25:15see this as the same space. You can
25:17obviously try to create other index like
25:18the CCI30 or the Bloomberg crypto index
25:21all the thing but I don't think they
25:23just gain the popularity and also like
25:24for uh allocator that do size this is
25:28still the best benchmark they have. So I
25:29think for better or worse, I think
25:31>> it's still the benchmark that most
25:33people care about the most. I mean,
25:34nobody's going to be happy when and also
25:36absolute returns. Both absolute returns
25:38and Bitcoin out performance. Yeah,
25:39that's the only two thing people care
25:40about in my view.
25:42>> What do you guys think?
25:43>> Yeah, I don't have a super strong view.
25:44Um, I think like benchmarking relative
25:47to peers is also like a pretty
25:48reasonable thing. Um, and you know,
25:51smart uh LPS I think do that in a pretty
25:54uh systematic fashion.
25:56>> Yeah, I think that's probably true. like
25:59it's hard to especially in the age of
26:01ETFs where it's hey if I'm an LP and I
26:03can go buy IBIT for example like why why
26:07was the point of allocating to um like a
26:11fund that doesn't outperform that so I
26:13think it's like a b a bare minimum kind
26:15of baseline I think um and then I don't
26:17know you can get into
26:20is crypto becoming like is it becoming
26:23less of its own thing and so you should
26:24look at some of the other kind of deep
26:27tech style assets that trade in public
26:29equities or technology indices or things
26:31like that. But yeah, that that's kind of
26:32like where I sit today.
26:34>> Um, last topic here as we wrap up, but
26:37speaking specific to the venture side of
26:39things, obviously most the concentration
26:41of most checks over the last, you know,
26:43six years or so have been written in the
26:45in the seed stage of things and, you
26:47know, that's had some success and it's
26:48also hasn't. Um and now you know the
26:52phase that the industry is at it's
26:55becoming quite clear
26:57what has product market fit what are the
26:59key use cases and I feel like the next
27:01question to solve there is okay do we do
27:03we go start to fund the later stages do
27:06we focus more on the growth side versus
27:07the seed stage side um and you know some
27:10of these existing protocols and
27:12companies that have shown product market
27:13fit do you how do you guys think about
27:16the the concentration between seed stage
27:18versus growth investing for the for the
27:20next couple years here. Um Arthur, you
27:23want to start?
27:25>> Um now I'm not a venture guy. So uh but
27:28I think the observation I have is there
27:30has been a extreme overlap
27:33>> among the latest stage deals uh both in
27:35equity and I think to a big extent
27:37crypto because there's just so little
27:39late stage deals happening and obviously
27:41you have a huge amount most people can
27:44get the allocation. So I think there's
27:46kind of a lot of overlap among all the
27:47late stage deals and I kind I think it's
27:49also usually they did quite well and you
27:51kind of need to get in. So yeah, I think
27:54that's just what it is right now. I
27:55think Cstage is
27:56>> a lot of opportunity right now because
27:58everyone got so burned and I think the
28:00founder quality is still still hard to
28:02come by especially now with the AI
28:03taking away a lot of the good quality
28:05founders. is is you're really
28:07underwriting a lot of risk. But imagine
28:09like you did like uh like and but the
28:12good thing about crypto I think it's
28:13still kind of similar with AI is like
28:15you can get the 50x uplift pretty fast.
28:19I think you can get it in one to two
28:20year if you done the C stage right. Um I
28:23think really it's only crypto and AI
28:25probably a bit of like a space you can
28:27get this kind of uplift. I don't know
28:28whether any other industry can get like
28:30a 50x um in like two years. Yeah. would
28:33would you want would would you want
28:34actually the universe to get larger and
28:36like to from the token side because
28:38presumably if there's more seed and
28:40preede stuff getting done that then
28:41ultimately lead to tokens like is that
28:43as like a liquid only fund is that like
28:46better for you because the universe
28:48expand%
28:50we would love to have a more universe
28:51like I think the ideal portfolio size
28:53for us is like 10 to 20 names but
28:55usually we are just restricted to five
28:56to 10 names which is extremely
28:58concentrated for like a hedge fund
29:00perspective. Yeah.
29:01>> Yeah. My perspective is that like at the
29:03early stages the outlier uh the number
29:06of outlier founders has not dramatically
29:09decreased. Um there are a lot fewer seed
29:13stage founders than there were 2 or 3
29:15years ago partially cuz I think there's
29:17been a lot of right sizing sort of at
29:19the seed stage relative to the capital
29:21that was being deployed in the past
29:23>> but I still think that like the number
29:24of in extremely high quality founders at
29:26early stages is still there. um it seems
29:29like a lot of the growth stage companies
29:32um uh have traction that's like quite
29:34durable versus like needing to raise
29:35purely off narratives which I think was
29:37like a little bit different in the past.
29:39So my hunch is like there's probably
29:41more opportunity for people to do seed
29:43stage stuff. My guess is that as a lot
29:45of these growth stage companies continue
29:46to compound, there will probably be um
29:49there will be uh additional market
29:51participants maybe who aren't even
29:52really in crypto and and maybe more um
29:55more of the growth generalist funds sort
29:56of as we've seen come in and sort of uh
29:59allocate to some of these companies.
30:00>> Yeah, for us like we
30:05are not like we think we're best at
30:07preede and seed stage and so we're not
30:08going to kind of scope creep on our side
30:10on like the private side. for us it's
30:13stick to preede and seed and then um
30:15publiclix and everything between is kind
30:17of like we don't think that's where we
30:19should be competing um and so I have
30:22less of a view there I think that there
30:24is a world where some of the things that
30:26have been tried in the past in crypto
30:28and have largely been written off as oh
30:30that sector doesn't work or that is kind
30:33of like a broken thing that that won't
30:35work um I'm maybe
30:38naively optimistic that actually you
30:40know there's there's a world in which
30:42that was not the right time to try some
30:43of these things 3, four, 5 years ago.
30:46Um, and so for us, yeah, preed and seed
30:48stage is where we're going to continue
30:50to sit. And I do think there are
30:51versions of companies that have probably
30:53been tried in the past that could work
30:55in the future given like the right
30:57founder. Again, I think as crypto
30:58becomes less um less of like uh I would
31:03say a lower status industry compared to
31:07AI or some of the other kind of
31:09technology sectors. I think there's one
31:11point I would like to ask like do you
31:12think the capital intensity of uh early
31:15stage in crypto have came down due to AI
31:18and you know because you can kind of
31:19like you need less engineer and do you
31:21think like you're going to see like a
31:22billion dollar AI like a billion dollar
31:25startup
31:26>> with a solo entrepreneur like in crypto
31:28>> I think it's yeah I think it's
31:30definitely happened like we've even seen
31:31portfolio companies who are like we can
31:33actually operate with less employees
31:35because
31:36>> team members are becoming much more AI
31:38native and and using these tools and
31:40they're saying the team members who are
31:41able to do this and then leverage it are
31:43actually like those are the people we
31:46keep around and everybody else who's
31:47like slow to adopt or who can't really
31:49wrap their head around it. It's like we
31:50actually don't like they're not net ad
31:52additive.
31:53>> So so for let's say for a team who
31:54raised like 5 to 10 million early stage
31:56where do they spend the money on these
31:57days compute or like you know marketing
31:59budget?
32:00>> I mean it depends on the project right.
32:03Um, I think historically a huge chunk of
32:05that has gone to headcount. A meaningful
32:07chunk of that headcount is probably
32:09being reallocated to compute or API
32:12credits and stuff like that. Um, for
32:14something that needs to do a lot of
32:16marketing, I think that's still there
32:17for um, for people that need to do a lot
32:19of enterprise sales, you still do need
32:21to hire people. I think uh, people are
32:23still typically buying from people even
32:25if you're using a bunch of tools in the
32:26process. But that 10 million I think
32:28gets you a lot further. Um, unless you
32:30are training automodels, in which case
32:32like it does not get you very far at
32:34all.
32:34>> Yeah.
32:34>> Yeah.
32:35>> Yeah.
32:35>> All right, that's all the time we had.
32:37Uh, thank you all for joining. That half
32:38an hour flew by. I think we're going to
32:40break for lunch now and then we'll be uh
32:41back at it in a little bit. So, yeah.
32:43Thanks. Thanks guys for joining. That
32:44was great.
32:45>> Thank you. Thanks.

About this transcript

This page contains the full transcript of Where Is The Crypto Alpha Hiding This Cycle? by Forward Guidance, generated from the public captions YouTube serves with the video. The transcript has 6,745 words across 944 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.