1
00:00:00,000 -->
00:00:06,000
Matias Antonio: I actually quit the World Bank Pension Fund to start my own crypto fund that didn't work
2
00:00:06,000 -->
00:00:11,400
out. If you can guess by the timing, it's the perfect peak at a very difficult moment.
3
00:00:11,480 -->
00:00:14,660
What if we could lower inflation to zero and give them the value that they captured?
4
00:00:14,980 -->
00:00:18,620
That's a structural reason for validators to stay.
5
00:00:18,780 -->
00:00:23,780
You can use Ethereum as an example. You know, doing a transfer was $200
6
00:00:23,780 -->
00:00:26,200
and I was crying every time I did it.
7
00:00:27,260 -->
00:00:31,440
John Wu: Economics is kind of an art as well as it is a science. Human beings are not rational
8
00:00:31,440 -->
00:00:32,200
optimizers.
9
00:00:32,460 -->
00:00:36,400
Matias Antonio: The very early blockchain, it really was a game theory problem.
10
00:00:36,720 -->
00:00:41,040
What does big mean? Nobody knows. So price go up, right? Then an economy actually formed.
11
00:00:51,460 -->
00:00:54,860
Kelvin Sparks: Hello and welcome to the Layer 1 podcast. My name is Kelvin Sparks.
12
00:00:55,020 -->
00:00:59,120
This is my co -host John Wu, the president of Ava Labs. Joining us today, we have Matias
13
00:00:59,120 -->
00:01:04,020
Antonio, CIO at the Avalanche Foundation, and Eric Liu, lead economist at the Avalanche
14
00:01:04,020 -->
00:01:07,720
Foundation. Welcome to the show. How are we all doing? Great. Good to see you, Calvin.
15
00:01:08,620 -->
00:01:12,480
Matias, Eric, nice to meet you properly. Yeah, very nice to see you.
16
00:01:12,480 -->
00:01:16,540
Good to see you, Calvin. Before we get into today's episode, Layer 1 is a podcast focused
17
00:01:16,540 -->
00:01:19,660
on the intersection of crypto and the real world, and it's brought to you in collaboration
18
00:01:19,660 -->
00:01:23,400
with Avalanche. Nothing we see on this podcast is investment or financial advice.
19
00:01:23,580 -->
00:01:27,280
Make sure to always do your own research. In today's episode, we'll be talking about the
20
00:01:27,280 -->
00:01:31,660
economics research roadmap. But before we jump into that, I want to talk about the
21
00:01:31,660 -->
00:01:35,560
Avalanche Summit. I'm really excited because that's coming up September 16th and 17th.
22
00:01:35,720 -->
00:01:40,340
Speakers like Guy from Athena, Brett Tejpaul, co -CEO of Coinbase Institutional, Luigi
23
00:01:40,340 -->
00:01:45,600
Donorio DeMeo from Ave, Sandy Call from Franklin Templeton, and so many others will be in
24
00:01:45,600 -->
00:01:49,760
the building. And it's just around the corner. So let's get into today's episode.
25
00:01:50,020 -->
00:01:55,040
That being said, when I was getting prepared for this one, And Matthias, you're with the
26
00:01:55,040 -->
00:01:58,460
IMF in the World Bank Treasury previously.
27
00:01:58,900 -->
00:02:04,040
Eric, you hold a PhD in finance and led a research blockchain risk analytics
28
00:02:04,040 -->
00:02:09,680
team. So, John, you were also in the hedge fund world and you ran your own fund.
29
00:02:09,900 -->
00:02:12,260
So a good place to start to be, what brought you all to Avalanche?
30
00:02:12,480 -->
00:02:17,560
John Wu: I mean, I'll start very quickly. I mean, Avalanche is a technology I thought was really
31
00:02:17,560 -->
00:02:22,540
going to be able to and is in the process of transforming traditional financial rails.
32
00:02:22,540 -->
00:02:26,680
You know, Avalanche, Avalanche Labs is working with the Avalanche Foundation.
33
00:02:27,100 -->
00:02:32,920
We have tokenized so many real world assets and the entire economic system can be more
34
00:02:32,920 -->
00:02:38,920
efficient and new ways of creating value for and providing access to
35
00:02:38,920 -->
00:02:43,900
individuals can all be possible in an open permissionless blockchain.
36
00:02:44,280 -->
00:02:49,940
Matias Antonio: And my story started actually in 2017 when I read the Bitcoin white paper
37
00:02:49,940 -->
00:02:55,840
and Luma in mine then ethereum blown twice even red cardano
38
00:02:55,840 -->
00:03:01,200
believe it or not uh and then by i actually quit the world bank pension fund to start my
39
00:03:01,200 -->
00:03:07,180
own uh crypto fund that didn't work out if you can guess by the timing it's the perfect
40
00:03:07,180 -->
00:03:12,840
peak at a very difficult moment so and then 2020 came and with the new
41
00:03:12,840 -->
00:03:18,740
uh you know there were new layer ones and so i started researching them i i took a look
42
00:03:18,740 -->
00:03:23,840
at Solana and Avalanche stood out. The part that really stood out for me as I was reading
43
00:03:23,840 -->
00:03:29,720
the paper was the probabilistic consensus specifically because having studied mathematics
44
00:03:29,720 -->
00:03:35,500
and with a background of economics, I could appreciate how a
45
00:03:35,500 -->
00:03:41,020
sampling process can speed up the solution of an equilibrium outcome that you're trying to
46
00:03:41,020 -->
00:03:45,240
compute and so on and so forth. And I appreciate that, but I never thought that it could
47
00:03:45,240 -->
00:03:50,820
be applied to a blockchain consensus mechanism that actually produces a finality with
48
00:03:50,820 -->
00:03:54,040
precision. And I thought that was quite interesting.
49
00:03:54,620 -->
00:03:58,380
And on the other side of things, I thought, oh, wow, they're actually focused on
50
00:03:58,380 -->
00:04:04,240
tokenization and real world assets. And that's kind of the world that I lived
51
00:04:04,240 -->
00:04:08,860
in. And so when the opportunity came, I said, yes,
52
00:04:10,020 -->
00:04:15,520
I'm in, please take me. And then having met the people inside the ecosystem,
53
00:04:15,520 -->
00:04:20,740
John, who unfortunately I don't get to see as often as I would like to, but I always enjoy
54
00:04:20,740 -->
00:04:26,200
talking with him. It was a no -brainer for me to stay and
55
00:04:26,200 -->
00:04:27,860
haven't looked back since.
56
00:04:28,460 -->
00:04:34,080
Eric Lu: Yeah, so I actually got interested in crypto very early on, actually when I was still in
57
00:04:34,080 -->
00:04:40,000
undergrad. But then, at that time, I feel like I have to dig
58
00:04:40,000 -->
00:04:45,820
deep into economics and finance. At that time, that's really where my passion is.
59
00:04:46,360 -->
00:04:52,320
And then after I got my PhD, I got into a consulting firm as a financial economist,
60
00:04:52,340 -->
00:04:58,300
and there I kind of get first professional exposure in a
61
00:04:58,300 -->
00:05:04,140
crypto -related case, where I essentially analyzed the order flow of a
62
00:05:04,140 -->
00:05:09,880
centralized crypto exchange. And I thought, this is the industry
63
00:05:09,880 -->
00:05:14,160
that I see is going to boom in the next decades.
64
00:05:14,600 -->
00:05:19,680
And that's where I want to spend a better part of my career at.
65
00:05:19,860 -->
00:05:25,500
So I moved to a blockchain analytics company where I was
66
00:05:25,500 -->
00:05:31,460
also the crypto economist, and building data and analytical products for
67
00:05:31,460 -->
00:05:36,200
institutional users mostly. And after that, the move to Avalanche Foundation is actually
68
00:05:36,200 -->
00:05:42,040
quite an easy decision for me because at that time I was thinking, I need to
69
00:05:42,040 -->
00:05:46,960
get to the front line of the industry and generate real impacts, right?
70
00:05:47,120 -->
00:05:51,240
And just joining the foundation allows me to do exactly that.
71
00:05:51,580 -->
00:05:53,580
And that's how I ended up here.
72
00:05:54,240 -->
00:05:59,480
John Wu: Eric, I read some of your recent stuff, the stuff that you and Matthias have put out.
73
00:05:59,660 -->
00:06:05,380
And I think, first of all, thank you for a refresher my undergrad economics studies.
74
00:06:06,100 -->
00:06:12,080
But certain points that really resonate, for instance, like in the US, we
75
00:06:12,080 -->
00:06:15,780
forget that GDP and GNI are very similar.
76
00:06:16,060 -->
00:06:20,120
The numbers are almost always on top of each other, but there are definitely countries in
77
00:06:20,120 -->
00:06:23,660
the world where GDP and GNI are completely different.
78
00:06:23,860 -->
00:06:29,860
And if you apply it to the world of blockchain ecosystems or layer ones,
79
00:06:29,860 -->
00:06:33,280
you can see that there is a significant difference.
80
00:06:33,700 -->
00:06:39,460
For the audience, maybe you or Matias, will you take a review of how you measure, capture,
81
00:06:39,840 -->
00:06:44,860
and all the frameworks that you've put together for that, and then we can
82
00:06:44,860 -->
00:06:48,460
dive into it so that after we level set with everyone?
83
00:06:48,880 -->
00:06:54,440
Matias Antonio: Yeah, absolutely. And this is actually one of the key areas of the foundation
84
00:06:54,440 -->
00:07:00,260
in research, right? as a nonprofit that is here for the ecosystem,
85
00:07:00,260 -->
00:07:05,880
we naturally have a longer -term view, and so naturally we gravitate towards that gap that
86
00:07:05,880 -->
00:07:11,580
needs to be filled. And as we were thinking about what is out there, what became clear to
87
00:07:11,580 -->
00:07:17,580
us, and this is the first bit, the measure bit, is that there isn't
88
00:07:17,580 -->
00:07:20,640
a way to actually measure what is the value that runs on the chains.
89
00:07:21,360 -->
00:07:27,280
There's the accounting approaches that have been made with the PE and so on, And you can
90
00:07:27,280 -->
00:07:32,160
see in Token Terminal some of these, let's call it national accounts that they have
91
00:07:32,160 -->
00:07:37,900
compiled. But in it, it's actually not capturing the bit that
92
00:07:37,900 -->
00:07:42,520
I wouldn't say matters, but it's a different angle that enables us to understand what is
93
00:07:42,520 -->
00:07:45,540
it the gap in terms of the capture, which I will speak to in a bit.
94
00:07:46,140 -->
00:07:52,100
and and that's where really gcp was born because economies let's say the u .s
95
00:07:52,100 -->
00:07:57,460
it doesn't measure itself by the number of dollar transactions and dollar volumes that go
96
00:07:57,460 -->
00:08:02,380
through the economy right it is a way of viewing it in the sense that it has an important
97
00:08:02,380 -->
00:08:07,880
it's an important piece of understanding but really the value is about the the revenue
98
00:08:07,880 -->
00:08:13,740
that the companies receive and that we observed as something that was missing and
99
00:08:13,740 -->
00:08:19,740
how this GCP was born the GCI is a similar story it's a
100
00:08:19,740 -->
00:08:24,420
different lens of measuring also the value that runs a chain that I'm sure Eric will dig
101
00:08:24,420 -->
00:08:29,880
much deeper on later on and that helps us have a benchmark in terms of what is there
102
00:08:31,000 -->
00:08:36,840
the second question is then what can we capture and how
103
00:08:36,840 -->
00:08:41,300
should we capture and what should we capture it There's many layers of the questions there
104
00:08:41,300 -->
00:08:44,060
because now we're starting to get into the tokenomics side of things.
105
00:08:44,620 -->
00:08:49,640
Because the fundamental problem there is that all layer ones, this is not unique to us,
106
00:08:49,760 -->
00:08:55,140
the value accrual process in terms of how these value that the chain is helping generate
107
00:08:55,140 -->
00:09:00,940
that wouldn't exist if the chain wouldn't be there is not going
108
00:09:00,940 -->
00:09:05,280
through the token. And that's a fundamental question that hasn't been answered.
109
00:09:05,280 -->
00:09:10,500
So in many ways, the FAT protocol thesis fails to answer why should that be the case.
110
00:09:11,200 -->
00:09:16,680
And so that's where the question of how do we capture comes from, right?
111
00:09:16,760 -->
00:09:20,420
How do we strengthen the value accrual mechanism?
112
00:09:20,960 -->
00:09:24,580
And the first piece is understanding the value. Then how can we capture it?
113
00:09:24,660 -->
00:09:28,320
And then there's a third bit, which is how do we distribute it? So what does this
114
00:09:28,320 -->
00:09:34,160
distribution mean? it means how do we the value that we have captured how
115
00:09:34,160 -->
00:09:39,220
do we make it flow such that it maximizes the value of the token
116
00:09:39,220 -->
00:09:45,040
explicitly why does that matter because well on one sense uh
117
00:09:45,040 -->
00:09:50,620
you know validators right now are depending on inflation in order to um
118
00:09:50,620 -->
00:09:55,380
state right what if we could lower inflation to zero and give them the value that they
119
00:09:55,380 -->
00:10:00,860
captured that's a structural reason for validators to stay and
120
00:10:00,860 -->
00:10:03,640
stay on the long term. That's a sustainable reason.
121
00:10:04,680 -->
00:10:07,940
And the other side is that it helps strengthen equilibrium outcomes.
122
00:10:08,120 -->
00:10:13,820
So now the reason of why they stay is no longer bound to the price of the token or the
123
00:10:13,820 -->
00:10:19,340
inflation rate of the token, but rather form a real cash flow that is being received due
124
00:10:19,340 -->
00:10:22,800
to the capturing mechanism that the chain has.
125
00:10:23,560 -->
00:10:29,140
And then on the other side it's about scaling right in a proof of stake chain
126
00:10:29,140 -->
00:10:34,860
the the the security of the chain is a direct function of how much
127
00:10:34,860 -->
00:10:40,740
value is actually at stake so if the value crop process is weak and the
128
00:10:40,740 -->
00:10:46,680
token doesn't strengthen that value doesn't grow and so when you go to
129
00:10:46,680 -->
00:10:52,300
larger companies who need more assurances in order to be onboarded,
130
00:10:52,420 -->
00:10:57,620
that's an important component. And so that's why this measure capture distribute
131
00:10:57,620 -->
00:11:03,260
is a nice framework that kind of captures how we're thinking about the research and really
132
00:11:03,260 -->
00:11:06,880
the why of why we're here and why we're doing it.
133
00:11:07,040 -->
00:11:08,320
That's very insightful.
134
00:11:09,100 -->
00:11:13,880
John Wu: If I may, I'm just going to summarize a few points that I think are important for our
135
00:11:13,880 -->
00:11:19,560
audience here. What Matthias says is you have to look in Matthias and
136
00:11:19,560 -->
00:11:25,000
Eric's work. GCP is equivalent of GDP and GCI
137
00:11:25,000 -->
00:11:29,100
is equivalent of GNI, traditionally speaking.
138
00:11:29,600 -->
00:11:35,600
And what's being measured right now is sometimes just to focus
139
00:11:35,600 -->
00:11:41,220
on effectively the GDP. And on top of that, in the traditional dashboards
140
00:11:41,220 -->
00:11:47,200
in the blockchain and crypto world. And then on top of that, if you think of capture, the
141
00:11:47,200 -->
00:11:53,180
issue with capture is you can almost think of it as a taxation for
142
00:11:53,180 -->
00:11:59,120
a country or for an area. And right now, especially on Avalanche, the
143
00:11:59,120 -->
00:12:05,100
taxation rates are very low. And because it's so low in theory, the
144
00:12:05,100 -->
00:12:11,080
applications and the people developing on top of Avalanche are able to get a very good
145
00:12:11,080 -->
00:12:16,980
deal, if you will. And lastly, however, that deal can be separated between a nominal
146
00:12:16,980 -->
00:12:22,640
with a real number and the inflation adjustment
147
00:12:22,640 -->
00:12:27,040
associated with the token can be optimized for all parties.
148
00:12:27,520 -->
00:12:31,520
If that's my summary, I'll let you and Eric move on from there.
149
00:12:31,880 -->
00:12:31,940
I
150
00:12:31,940 -->
00:12:36,680
Matias Antonio: may not use the word taxation, but yes, good summary.
151
00:12:37,780 -->
00:12:43,620
Eric Lu: Yeah, I'll get to that, actually. But yeah, thank you for the summary, actually, Jiang.
152
00:12:43,840 -->
00:12:49,760
And I would want to add, I think, two important perspectives on the two measures
153
00:12:49,760 -->
00:12:55,360
in particular. So obviously, we're not creating these two measures
154
00:12:55,360 -->
00:12:59,740
for the sake of having two measures, and they're nicely
155
00:12:59,740 -->
00:13:05,580
corresponding to these national accounting measures that we're kind of
156
00:13:05,580 -->
00:13:11,540
familiar with. But the GCP and GCI metrics that we
157
00:13:11,540 -->
00:13:17,400
build actually correspond really nicely to two different sources of
158
00:13:17,400 -->
00:13:21,400
value base of a blockchain ecosystem like Avalanche.
159
00:13:21,560 -->
00:13:25,660
So on the one hand, the GCP is really measuring the economic output.
160
00:13:25,780 -->
00:13:31,680
So you can think of it as the productive process, the production output.
161
00:13:31,680 -->
00:13:37,540
Whereas GCI, with the analog to GNI, is really kind of the
162
00:13:37,540 -->
00:13:43,460
income accrued to the residents or, broadly speaking, the ecosystem
163
00:13:43,460 -->
00:13:49,040
participants. And they all have very different implications when it comes to capture.
164
00:13:49,860 -->
00:13:55,360
So any capture we want to impose on the GCP, which
165
00:13:55,360 -->
00:14:00,640
ultimately is backed by all the revenue generated by the applications or L1s in the
166
00:14:00,640 -->
00:14:04,860
Avalanche ecosystem, they would have an implication on the productive process.
167
00:14:05,160 -->
00:14:10,720
So how can we sustainably capture part of that value without
168
00:14:10,720 -->
00:14:16,180
distorting that productive process so that the ecosystem can remain productive,
169
00:14:16,520 -->
00:14:21,540
competitive, and growing in the long term is something that we need to
170
00:14:21,540 -->
00:14:26,480
understand and investigate. So that's the GCP.
171
00:14:26,720 -->
00:14:30,180
And then for GCI, it's a similar story.
172
00:14:30,280 -->
00:14:32,740
So it's an income stream.
173
00:14:33,340 -->
00:14:39,200
Any capture we want to impose on that is going to have an implication
174
00:14:39,200 -->
00:14:44,960
on the adoption. So the vast majority of the GNI actually comes from, or
175
00:14:44,960 -->
00:14:50,480
GCI actually comes from, for example, the reserve assets of the RWAs or
176
00:14:50,480 -->
00:14:55,580
stablecoins in the ecosystem. And if we take a cut of the
177
00:14:55,580 -->
00:15:00,120
yield from the reserve assets, obviously that's going to have an implication on adoption.
178
00:15:00,660 -->
00:15:05,960
So again, we have to think about, what's the optimal way of capturing that without
179
00:15:05,960 -->
00:15:10,380
jeopardizing the long -term sustainability of that revenue stream as well.
180
00:15:10,440 -->
00:15:14,340
So that's one important perspective of these two measures.
181
00:15:14,480 -->
00:15:19,560
On the other hand, the second important perspective I think you already very nicely
182
00:15:19,560 -->
00:15:24,820
touched on is that these two measures are not just two simple measures.
183
00:15:25,300 -->
00:15:31,180
These really are based on an accounting system that gives us a near
184
00:15:31,180 -->
00:15:36,360
real -time and granular lens to look at the individual components of the whole economy.
185
00:15:36,360 -->
00:15:42,220
So we can ask questions like which verticals or applications or
186
00:15:42,220 -->
00:15:46,280
L1s is actually generating growth in terms of these two measures.
187
00:15:46,560 -->
00:15:51,660
And maybe ultimately in the future, we can even use these two measures to identify
188
00:15:51,660 -->
00:15:56,400
the growth opportunities or which part of the economy is working better than the others.
189
00:15:56,640 -->
00:16:01,980
Or where we should really put the support in the economy to sustain the
190
00:16:01,980 -->
00:16:07,940
growth of these two measures. And finally, doubling down on your point
191
00:16:07,940 -->
00:16:13,600
about these measures enable us to look at the real versus nominal
192
00:16:13,600 -->
00:16:19,500
term of the economy. I think that's kind of a good benefit of
193
00:16:19,500 -->
00:16:25,020
having these two measures that are very much rooted from the classical
194
00:16:25,020 -->
00:16:30,000
national accounting. is that we get to look past the
195
00:16:30,000 -->
00:16:33,240
price changes, which is like
196
00:16:33,240 -->
00:16:39,020
the price of the crypto assets obviously
197
00:16:39,020 -->
00:16:44,860
fluctuate a lot. By distinguishing between the nominal changes and the real changes,
198
00:16:45,040 -->
00:16:50,920
that really gives us a deeper understanding of how the ecosystem is evolving over
199
00:16:50,920 -->
00:16:56,620
time. And one concrete example I can give you is that obviously we know that we've
200
00:16:56,620 -->
00:17:01,200
seen the trading volume across the industry decline
201
00:17:01,200 -->
00:17:06,920
significantly since late last year. But if you look closely through the lens of
202
00:17:06,920 -->
00:17:12,340
real versus nominal terms, what you can see is that decline is actually driven
203
00:17:12,340 -->
00:17:18,060
more by the price changes of the crypto assets, which are the assets being traded in
204
00:17:18,060 -->
00:17:23,800
these crypto ecosystems. But if you keep the price constant and look at the real
205
00:17:23,800 -->
00:17:28,820
activities, is actually declining not nearly as much.
206
00:17:28,980 -->
00:17:34,440
So that provides an understanding of the real activities behind
207
00:17:34,440 -->
00:17:39,940
these drastic changes that we see all these data vendors are reporting,
208
00:17:40,180 -->
00:17:42,540
which are nominal terms.
209
00:17:42,740 -->
00:17:47,900
Kelvin Sparks: I was just curious on the topic of the ecosystem itself, what are the biggest value
210
00:17:47,900 -->
00:17:52,600
-creative opportunities you're seeing today, be it protocols, apps, centralized exchanges,
211
00:17:52,860 -->
00:17:55,820
DeFi, so on and so forth? We definitely love your guys' perspective.
212
00:17:56,700 -->
00:18:01,380
Eric Lu: Yeah, so I think it's... I kind of tainted at it
213
00:18:03,120 -->
00:18:08,700
through the GCP versus GCI lens. So I think that the high potential opportunities that we
214
00:18:08,700 -->
00:18:12,380
see map nicely to the two different metrics.
215
00:18:13,680 -->
00:18:18,320
And in turn, these two metrics give us a better understanding of the magnitude of these
216
00:18:18,320 -->
00:18:21,460
opportunities and a potential mechanism to capture them.
217
00:18:21,460 -->
00:18:27,180
As I mentioned earlier, so the GCP kind of really, most of it is coming from the
218
00:18:27,180 -->
00:18:28,880
revenue of the different applications.
219
00:18:30,400 -->
00:18:36,040
And, you know, obviously right now, you know, across most of the
220
00:18:36,040 -->
00:18:41,700
general purpose L1s, not just Avalanche, transaction
221
00:18:41,700 -->
00:18:47,620
fees is one of the main sources of capture revenue for
222
00:18:47,620 -->
00:18:53,600
the protocol. And one important challenge of
223
00:18:53,600 -->
00:18:59,260
transaction fee being the sole or the most important source of revenue is that it
224
00:18:59,260 -->
00:19:03,980
doesn't really give us enough exposure to the ecosystem growth.
225
00:19:03,980 -->
00:19:09,820
So if we measure the ecosystem value and growth through the lens of GCP or GCI,
226
00:19:10,020 -->
00:19:14,920
obviously, you wouldn't see transaction fee co -vary a lot with
227
00:19:14,920 -->
00:19:20,260
the value of ecosystem, essentially.
228
00:19:20,660 -->
00:19:26,580
So really, that brings us to the question that if we view
229
00:19:26,580 -->
00:19:32,260
the value of the ecosystem through the lens of the GCP, then how can we really get
230
00:19:32,260 -->
00:19:38,200
a meaningful share of the GCP that the ecosystem
231
00:19:38,200 -->
00:19:43,840
is generating? And one of the priorities of our research agenda is
232
00:19:43,840 -->
00:19:49,020
exactly looking at that. So what's the sustainable incentive compatible way
233
00:19:49,020 -->
00:19:54,620
to capture part of that, or to bring part of that
234
00:19:54,620 -->
00:20:00,260
GCP value back to the protocol and eventually flow that to the
235
00:20:00,260 -->
00:20:03,160
token holders.
236
00:20:03,480 -->
00:20:09,160
And it's kind of the same story for the GCI. And obviously,
237
00:20:09,440 -->
00:20:15,160
many other L1s or applications,
238
00:20:15,500 -->
00:20:21,240
they are already sharing some of the yield from the stablecoin
239
00:20:21,240 -->
00:20:24,200
balance on their chain or using their applications.
240
00:20:24,200 -->
00:20:29,880
and that's obviously one of the area or direction that we're actively exploring as well.
241
00:20:30,720 -->
00:20:35,200
John Wu: So it's a fair, again, to try to make it into a real -world comparison.
242
00:20:37,280 -->
00:20:42,860
Kind of what I'm hearing is, you know, traditionally GDP, GCP, you have NV equals
243
00:20:42,860 -->
00:20:48,520
PQ. And here, what we have is a lot of dormant capital because a lot of the real -world
244
00:20:48,520 -->
00:20:52,920
assets. So the money supply is really just sitting there. There's no velocity associated
245
00:20:52,920 -->
00:20:58,340
with it. and one of the things that you guys are probably thinking about is how do we take
246
00:20:58,340 -->
00:21:04,320
this large base of money call it m1 m2 whatever you want to call it and how
247
00:21:04,320 -->
00:21:08,220
do we get the velocity working is that one way to think about it because otherwise having
248
00:21:08,220 -->
00:21:09,980
tvl is just having dormant capital
249
00:21:09,980 -->
00:21:15,640
Matias Antonio: definitely velocity is is a question that um we're thinking through and it has to do
250
00:21:15,640 -->
00:21:21,360
with some of the challenges and compostability on defy or perhaps some uh that has to do
251
00:21:21,360 -->
00:21:27,200
with some primitives that don't exist that make that capital actually productive in the
252
00:21:27,200 -->
00:21:31,560
sense of trading like I can think of a cross currency basis which I'm sure you're very
253
00:21:31,560 -->
00:21:35,760
familiar with right and and the instruments just don't exist on chain to make it happen
254
00:21:35,760 -->
00:21:41,540
and so there there is a bit of a transition friction in the
255
00:21:41,540 -->
00:21:47,160
sense that there is things that exist on TradFi that don't exist here and and they're
256
00:21:47,160 -->
00:21:51,500
making slowly their way here and and in that construction of the primitives is where
257
00:21:51,500 -->
00:21:57,200
dormant capital kind of seems to exist that actually may be very productive
258
00:21:57,200 -->
00:22:02,080
and people are hedging and doing things on the background we just don't see it because the
259
00:22:02,080 -->
00:22:07,420
activity is not happening on chain and i think that that you know to your point absolutely
260
00:22:07,420 -->
00:22:11,920
it is something that we're thinking about it's just in that process it's it's a bit
261
00:22:11,920 -->
00:22:17,000
further away to so say in terms of the capacity to capture and so
262
00:22:17,000 -->
00:22:21,480
currently the way that we're thinking about it is that there's two sides in terms of what
263
00:22:21,480 -->
00:22:25,500
is the potential capture path that we have. One is the supply side and one is the demand
264
00:22:25,500 -->
00:22:29,340
side. The one that you're talking about currently is on the demand side in terms of the
265
00:22:29,340 -->
00:22:34,160
things that are happening, how can we, you know, the activity, let's say, that effectively
266
00:22:34,160 -->
00:22:37,240
is asking for more block space, how do we capture that?
267
00:22:37,580 -->
00:22:42,460
But there's other things like on the block space demand. For instance, one of the ACPs
268
00:22:42,460 -->
00:22:48,220
that we put forth had to do with increasing the minimum gas fee, right?
269
00:22:48,220 -->
00:22:54,120
Because what we are observing is that gas fee, having cheaper gas fee, is
270
00:22:54,120 -->
00:22:57,300
not necessarily the edge. The edge is trust.
271
00:22:57,560 -->
00:23:01,440
Why do people use the chain? It's because they trust it, not because it's expensive.
272
00:23:02,000 -->
00:23:06,920
And you can use Ethereum as an example. Not to say that there's no price point after which
273
00:23:06,920 -->
00:23:12,660
you actually have user migration. And that clearly happened during 2021 when Ethereum
274
00:23:12,660 -->
00:23:17,620
was proof of work and doing a transfer was $200.
275
00:23:18,220 -->
00:23:20,100
And I was crying every time I did it.
276
00:23:21,500 -->
00:23:26,780
But, you know, like, it clearly created migration. That's why Avalanche benefited from
277
00:23:26,780 -->
00:23:29,120
that, right? Solana benefited from that.
278
00:23:30,000 -->
00:23:35,380
BNB benefited from that. So, like, there's definitely a price at which that flips, right?
279
00:23:35,400 -->
00:23:40,480
But we're nowhere near that price. So there's an argument to be made that you could easily
280
00:23:40,480 -->
00:23:46,420
increase the gas fee and get to a situation where you're capturing more of the
281
00:23:46,420 -->
00:23:51,360
value on the transaction side, i .e. through the supply side, and
282
00:23:51,360 -->
00:23:53,580
is direct.
283
00:23:54,580 -->
00:23:59,980
And the interesting bit here that sometimes is easily missed is that it's dynamic.
284
00:24:01,280 -->
00:24:06,700
Validators vote for it. So that means that you can actively engage on,
285
00:24:06,780 -->
00:24:12,200
I don't want to say monetary policy, but some sort of algorithmic policy where the
286
00:24:12,200 -->
00:24:18,100
collective validators can decide what is the appropriate gas fee because they
287
00:24:18,100 -->
00:24:21,080
themselves benefit from those transactions being higher.
288
00:24:21,700 -->
00:24:26,720
So they will also benefit from the transactions being more expensive globally.
289
00:24:26,940 -->
00:24:31,260
So there will be an equilibrium outcome, which I think is going to be very interesting to
290
00:24:31,260 -->
00:24:36,100
see how it plays out and where it lands, that I think is going to be a function of many,
291
00:24:36,120 -->
00:24:39,360
many things that are out there. Yeah.
292
00:24:39,900 -->
00:24:44,260
Kelvin Sparks: On the topic of validator economics, have you guys been following the, I think it was the
293
00:24:44,260 -->
00:24:49,500
EIP 8363 discussion that's been going on?
294
00:24:49,580 -->
00:24:53,820
I mean, it seems like Justin Drake, who's huge in the community, put that forward.
295
00:24:54,240 -->
00:24:58,600
And again, another huge proponent of Ethereum just outright hated it.
296
00:24:58,760 -->
00:25:03,520
DeFi builders, so on and so forth. So curious what you guys are learning from the
297
00:25:03,520 -->
00:25:05,840
discussions going on there. It's an interesting one.
298
00:25:06,400 -->
00:25:12,380
Matias Antonio: And correct me if I'm wrong, but the goal here was to target a particular staking
299
00:25:12,380 -->
00:25:17,720
ratio for Ethereum such that if it goes above it, then the yield goes down and otherwise
300
00:25:17,720 -->
00:25:22,500
it goes up, which I think it's actually an interesting way of going about it.
301
00:25:22,840 -->
00:25:27,560
I think in general, the tokenomics or Ethereum are interesting in the sense that they're
302
00:25:27,560 -->
00:25:33,020
all based on long -term equilibrium outcomes in the sense that you have inflation is an
303
00:25:33,020 -->
00:25:37,980
outcome of the activity and so the rate of yields in some sense kind of
304
00:25:37,980 -->
00:25:41,320
modulates towards that.
305
00:25:41,720 -->
00:25:46,360
So the long -term supply is relatively stable if you think about it, which is an
306
00:25:46,360 -->
00:25:49,880
interesting solution, not saying right, wrong, it's just an interesting different one.
307
00:25:50,160 -->
00:25:56,120
And so in that context, having a staking yield target is interesting.
308
00:25:56,500 -->
00:26:02,060
I don't necessarily think it's a bad thing, but I do think it creates risk which currently
309
00:26:02,060 -->
00:26:08,020
are difficult to hedge and manage specifically for defy applications and staking like
310
00:26:08,020 -->
00:26:13,600
yield stakers because what then you have is a high degree of fluctuation potentially
311
00:26:13,600 -->
00:26:18,660
in terms of what's the yield that you're you're receiving and how how do you even manage
312
00:26:18,660 -->
00:26:22,800
that there's there's no there's no swap here that you can go from fixed to floating right
313
00:26:22,800 -->
00:26:28,700
or floating to face for that matter so it becomes very very difficult and
314
00:26:28,700 -->
00:26:34,300
and so i can see why it can be very jarring in terms of the different
315
00:26:34,300 -->
00:26:40,060
perspectives because ultimately it impacts people's pockets and when it impacts
316
00:26:40,060 -->
00:26:44,580
people's pockets people have a very strong emotional reaction and rightfully so right and
317
00:26:44,580 -->
00:26:48,900
and that's that's i think where where it's coming frame. Where do I land on it?
318
00:26:48,980 -->
00:26:54,140
I haven't made my mind up yet, so I can't really say.
319
00:26:54,460 -->
00:26:58,880
But I do find it interesting. I've been following somewhat from afar.
320
00:26:59,080 -->
00:27:03,460
Eric is significantly deeper in the trenches, and I'm sure he has his perspective, which
321
00:27:03,460 -->
00:27:04,820
he's going to share.
322
00:27:05,380 -->
00:27:09,480
John Wu: I would love to hear Eric's perspective. In the traditional world, you manage that because
323
00:27:09,480 -->
00:27:14,780
you have a target inflation rate. And with that target inflation rate, you adjust policy,
324
00:27:14,920 -->
00:27:18,560
assuming you're reaching that. Sometimes arbitrary target inflation rate.
325
00:27:18,680 -->
00:27:23,240
You're bouncing it off with labor and unemployment and all that stuff.
326
00:27:23,580 -->
00:27:29,540
But I'd love to hear how Eric would think about a solution to manage it
327
00:27:29,540 -->
00:27:32,400
in a blockchain crypto ecosystem.
328
00:27:32,760 -->
00:27:36,600
Eric Lu: I'm not sure if I can give you the solution right now, but it's probably even more
329
00:27:36,600 -->
00:27:40,280
questions. But I would add my perspective.
330
00:27:40,560 -->
00:27:46,300
So I think ultimately what that EIP is about
331
00:27:46,300 -->
00:27:52,280
and the whole discussion or controversy around it is really, in the end, a
332
00:27:52,280 -->
00:27:56,660
matter of how much validators should get paid and who should pay.
333
00:27:57,320 -->
00:28:02,920
And if we bring that to Avalanche, obviously we face the same
334
00:28:02,920 -->
00:28:08,260
question as well. And right now for Avalanche, we're effectively imposing
335
00:28:09,460 -->
00:28:14,780
the inflation tax on all token holders and use that to pay
336
00:28:14,780 -->
00:28:20,240
the entirety of the validator reward. And so, yeah, so they're entirely paid by the
337
00:28:20,240 -->
00:28:26,080
inflation tax. And more importantly, the price of validation
338
00:28:26,080 -->
00:28:30,320
is set by formula in the protocol, right?
339
00:28:30,440 -->
00:28:35,580
So it's kind of declining over time as more AVAX is being minted.
340
00:28:35,580 -->
00:28:39,820
But most importantly, it's not determined by any economic forces.
341
00:28:40,300 -->
00:28:46,260
So this actually relates to multiple streams and questions that we
342
00:28:46,260 -->
00:28:51,280
laid out in our research agenda. For example, how much we should pay for validation and
343
00:28:51,280 -->
00:28:54,120
how much validation does the network actually need?
344
00:28:54,540 -->
00:28:58,200
And ultimately, this may turn into a mechanism design problem.
345
00:28:58,200 -->
00:29:04,040
So is there a mechanism where we can, on the one hand, review the willingness to
346
00:29:04,040 -->
00:29:09,460
pay by the network users for validation services?
347
00:29:09,660 -->
00:29:15,400
And on the other hand, what is the willingness to validate or the
348
00:29:15,400 -->
00:29:19,880
reserve price for the validators to be willing to validate?
349
00:29:19,880 -->
00:29:25,640
So one is the supply side, one is the demand side. And then perhaps we need some kind
350
00:29:25,640 -->
00:29:31,020
of equilibrium mechanism that can clear that market and give us the actual price
351
00:29:31,020 -->
00:29:36,840
that's going to eventually give us the optimal level of validation as well as the
352
00:29:36,840 -->
00:29:39,580
price that gives us the optimal level of validation.
353
00:29:41,220 -->
00:29:47,180
And on top of that, validation is actually a public good,
354
00:29:47,320 -->
00:29:52,700
right? Because the validators themselves, they don't fully internalize the benefit.
355
00:29:52,860 -->
00:29:57,900
Because when you validate, there's an additional validator or an additional stake, it's
356
00:29:57,900 -->
00:30:03,340
kind of increasing the security of the entire network. So everybody benefits from it.
357
00:30:03,580 -->
00:30:09,300
And the basic economic theory would suggest that if we rely
358
00:30:09,300 -->
00:30:15,280
on a competitive market equilibrium, we're going to end up in a situation where the
359
00:30:15,280 -->
00:30:21,280
validation service is undersupplied. So it's less than what the network
360
00:30:21,280 -->
00:30:26,720
as a whole actually needs. So there is also a question of if we
361
00:30:26,720 -->
00:30:32,680
want to get to the optimal level of validation services for the entire
362
00:30:32,680 -->
00:30:38,180
network, what is the right level of subsidy
363
00:30:38,180 -->
00:30:43,880
we need to provide so that we incentivize a bit more
364
00:30:43,880 -->
00:30:47,180
validators to participate in this validation.
365
00:30:48,320 -->
00:30:52,080
Yeah, so as you can see, there's different layers.
366
00:30:52,340 -->
00:30:57,460
Obviously, I'm not giving you any solutions so far. It's more questions after the other.
367
00:30:57,820 -->
00:31:03,020
But ultimately, I think that's the question we need to resolve, not just for us, but
368
00:31:03,020 -->
00:31:05,340
probably all the POS chains out there.
369
00:31:05,640 -->
00:31:08,460
John Wu: Well, I mean, it's true.
370
00:31:09,020 -->
00:31:12,100
You don't have the answer yet, but you're working towards that.
371
00:31:12,580 -->
00:31:15,920
And what you're describing basically is economics as a discipline.
372
00:31:16,180 -->
00:31:19,660
This is why it's a BA, not a BS.
373
00:31:19,960 -->
00:31:24,260
It's an arts. There's art into this. It's not an exact science.
374
00:31:25,120 -->
00:31:29,360
But there are frameworks. And if you think of these frameworks, basically you're talking
375
00:31:29,360 -->
00:31:34,060
about trade -offs right now. have you thought of something that would be equivalent like
376
00:31:34,060 -->
00:31:39,920
the Laffer curve for taxes? Like there's got to be an optimal thing you can draw
377
00:31:39,920 -->
00:31:42,980
out on a graph or something and we can work towards that.
378
00:31:43,440 -->
00:31:46,640
Eric Lu: That's an interesting perspective. So I think
379
00:31:50,780 -->
00:31:53,620
if you look, we go back to the
380
00:31:56,320 -->
00:32:02,200
research agenda. So there are many kind of different optimalities
381
00:32:02,200 -->
00:32:07,580
that we're trying to look for in the research agenda. And we have already
382
00:32:07,580 -->
00:32:13,220
talked through a number of them. I think, you know, right now the top priority is probably
383
00:32:13,220 -->
00:32:19,200
like, you know, at the capture level. So, and I think this is
384
00:32:19,200 -->
00:32:24,080
a good point to circle back to one of your tax analog, actually.
385
00:32:24,080 -->
00:32:29,740
So when we're doing the capture, we tend to not think of it
386
00:32:29,740 -->
00:32:35,440
as a taxation problem, not because we don't want to take a cut, but really,
387
00:32:35,600 -->
00:32:41,440
we're not really a sovereign stake that we impose a compulsory tax
388
00:32:41,440 -->
00:32:46,920
on the different applications. But instead, I
389
00:32:46,920 -->
00:32:52,480
did mention earlier that we want to find the sustainable and incentive -compatible
390
00:32:52,480 -->
00:32:57,080
mechanism that can capture that part of that revenue.
391
00:32:57,080 -->
00:33:03,020
And what I mean by that is actually, I think I do believe there is a mechanism
392
00:33:03,020 -->
00:33:09,000
where the protocol and application mutually benefit from
393
00:33:09,000 -->
00:33:14,200
some kind of value capture from the application to the protocol.
394
00:33:14,200 -->
00:33:20,080
Because if you think about it, all these applications, they have to rely
395
00:33:20,080 -->
00:33:25,400
on the value in the network, which is directly
396
00:33:25,400 -->
00:33:29,280
determining the security of the network.
397
00:33:29,600 -->
00:33:35,440
And if they have a high exposure to the network value, then they must benefit
398
00:33:35,440 -->
00:33:41,420
from some value flowing to the protocol, which boosts up the
399
00:33:41,420 -->
00:33:44,600
value of the network. So I feel like there must be a sweet spot.
400
00:33:45,180 -->
00:33:51,140
And that kind of is the optimality or optimal level of capture and also the
401
00:33:51,140 -->
00:33:56,420
mechanism of capture that we're shooting for. So that's kind of the capture part.
402
00:33:57,400 -->
00:33:57,780
And
403
00:33:57,780 -->
00:34:01,900
John Wu: you don't think about it as a pure sovereign state. You think about the better analog,
404
00:34:02,180 -->
00:34:05,300
maybe it's a co -op. Yes. And it's cooperative.
405
00:34:05,660 -->
00:34:10,640
And how do you think about a cooperative and how everyone benefits from the whole as well
406
00:34:10,640 -->
00:34:12,060
as the individual? So
407
00:34:12,060 -->
00:34:18,040
Matias Antonio: like, because there's things in here, you know, to follow your train
408
00:34:18,040 -->
00:34:23,760
of thought, John, that does make this particular discipline of economics as it is
409
00:34:23,760 -->
00:34:28,980
applied to blockchain very unique to it. Because there is a strong game theory component
410
00:34:28,980 -->
00:34:34,120
that in a sovereign state, it doesn't exist in the same way.
411
00:34:34,320 -->
00:34:38,740
Because in the sovereign state, you have political economics, which is slightly different.
412
00:34:39,060 -->
00:34:43,720
It's game theory, but it's different. In this case, you have it at the network level and
413
00:34:43,720 -->
00:34:49,700
so on. So to your point, it's kind of like a co -op with a
414
00:34:49,700 -->
00:34:53,260
sovereign state. It's a mix of bag of things.
415
00:34:53,420 -->
00:34:58,420
And so as Eric was talking about the optimality problems and you mentioned the Laffer
416
00:34:58,420 -->
00:35:04,420
curve, there's many optimal problems and each one is going to have a different
417
00:35:04,420 -->
00:35:10,040
chart and framework. So the Laffer curve will be a downward view, you know, and other ones
418
00:35:10,040 -->
00:35:14,140
may look very different. And the framework itself may be very different.
419
00:35:14,180 -->
00:35:18,960
And maybe it's not even math driven. It's like a pure game theoretic problem.
420
00:35:19,360 -->
00:35:25,260
So it's really, really difficult. It's a non -trivial, non -solved issue anyway.
421
00:35:25,900 -->
00:35:26,280
Well
422
00:35:26,280 -->
00:35:31,340
Kelvin Sparks: said, well said. And I guess as somebody who's never taken an economics class in their
423
00:35:31,340 -->
00:35:35,900
life, much less anything finance related. I come purely from like engineering and science
424
00:35:35,900 -->
00:35:40,860
background. What about your guys' background, like
425
00:35:40,860 -->
00:35:45,020
uniquely equips you to be able to fix these problems? I mean, you spoke to it at length,
426
00:35:45,140 -->
00:35:48,320
but I kind of want you guys to brag a little bit here because I just love where the
427
00:35:48,320 -->
00:35:49,020
conversation is
428
00:35:49,020 -->
00:35:54,180
John Wu: right now. Well, I mean, I thought about all of these things as a practitioner when I was
429
00:35:54,180 -->
00:35:58,560
an investor. Obviously, I was a technology investor, so it's more bottom up.
430
00:35:59,100 -->
00:36:03,600
But even as a technology investor in individual
431
00:36:03,600 -->
00:36:09,460
companies, you still have to understand the overall macro environment and
432
00:36:09,460 -->
00:36:14,820
what kind of environment lends to certain types of companies to invest in, obviously.
433
00:36:15,160 -->
00:36:20,580
So I thought about these issues. I studied it in school from an economics perspective,
434
00:36:20,640 -->
00:36:26,360
nowhere near what Eric has done. But having had a background both from an academic
435
00:36:26,360 -->
00:36:30,500
as well as a practitioner helps me think through some of these issues.
436
00:36:30,700 -->
00:36:36,540
Matias Antonio: In my experience, I have a bit of a mixed bag of experience in the sense that I
437
00:36:36,540 -->
00:36:41,660
studied math, so it helps me follow and understand the technical side of the chain.
438
00:36:41,800 -->
00:36:47,480
Obviously, I'm not an engineer, so there's a lot of details that I do miss and I don't
439
00:36:47,480 -->
00:36:53,360
pretend to really fully understand. and I did study economics and was at the World
440
00:36:53,360 -->
00:36:59,320
Bank and the IMF, which I had the luck or bad luck of actually being
441
00:36:59,320 -->
00:37:04,020
very close to the European crisis. And at the time when I was at the World Bank, there was
442
00:37:04,020 -->
00:37:08,400
a really interesting Jamaican crisis where the bond market was frozen.
443
00:37:08,520 -->
00:37:14,400
And I was very lucky to see that. And both the World Bank and IMF got involved and
444
00:37:14,400 -->
00:37:19,420
I spoke to the stakeholders on the ground and got really interesting insights.
445
00:37:19,600 -->
00:37:25,600
And that kind of helps me understand, broadly speaking, what does macro look
446
00:37:25,600 -->
00:37:30,920
like, right, and how to measure it. And more importantly, what are the risks that within
447
00:37:30,920 -->
00:37:35,380
an economy can exist? And I see blockchains as
448
00:37:36,460 -->
00:37:41,040
economic objects, right? There's a lot of details to it, as I was speaking earlier, but
449
00:37:41,040 -->
00:37:44,680
it's ultimately an economic object. So a lot of that framework applies.
450
00:37:44,860 -->
00:37:48,960
And so you get to see what are the opportunities, what are the risks, and how to think
451
00:37:48,960 -->
00:37:53,740
about it. And the interesting bit here is that one of the areas
452
00:37:53,740 -->
00:37:59,480
where blockchain particularly excels at, and Avalanche is
453
00:37:59,480 -->
00:38:04,480
one that's very well positioned for this particular problem, is value transmission.
454
00:38:05,160 -->
00:38:10,100
And that's where RWA's make sense right like tokenize the world right that's that's the
455
00:38:10,100 -->
00:38:15,700
avalanche mantra so for to really understand that bit you need to understand capital
456
00:38:15,700 -->
00:38:21,480
markets and that's where my asset management background is and was and i understand right
457
00:38:21,480 -->
00:38:27,120
because i was at the at the world bank pension fund i had the luck of being able to cover
458
00:38:27,120 -->
00:38:32,520
almost all asset classes i looked at public equities private debt, market neutral hedge
459
00:38:32,520 -->
00:38:38,280
funds, and helped figure out what tactical positioning the fund should do.
460
00:38:38,440 -->
00:38:44,060
And then I was at the fixed income desk where I was a portfolio manager there, managing a
461
00:38:44,060 -->
00:38:50,040
fairly large amount of assets and making calls in terms of what the fixed income and
462
00:38:50,040 -->
00:38:55,320
positioning should be for the portfolio. It's Tittner, do we want to be longer the
463
00:38:55,320 -->
00:38:59,940
Japanese bonds, we do cross -currency base. There's a lot of positioning that went with
464
00:38:59,940 -->
00:39:05,000
it. And in that experience, I got to see the value
465
00:39:05,000 -->
00:39:10,700
aggregation of the entire asset management space and therefore the capital
466
00:39:10,700 -->
00:39:15,020
market space. And that helps me really understand that side
467
00:39:17,040 -->
00:39:23,020
of this particular economy that is key to help it grow, given the focus areas.
468
00:39:23,020 -->
00:39:23,920
Yeah,
469
00:39:24,400 -->
00:39:27,960
Eric Lu: so I was trained as a financial economist, but
470
00:39:27,960 -->
00:39:33,740
before I graduated, I don't want to write academic
471
00:39:33,740 -->
00:39:39,400
papers for the rest of my life. And I want to solve real -world problems.
472
00:39:39,680 -->
00:39:45,440
So for me, I think, on the one hand, I do have the rigor in
473
00:39:45,440 -->
00:39:51,380
the classical economic finance theories that I believe that
474
00:39:51,380 -->
00:39:57,340
it's still give us probably all the foundation
475
00:39:57,340 -->
00:40:01,800
that we would need to better understand what's going on in the crypto world.
476
00:40:02,620 -->
00:40:08,080
But on the other hand, having been working
477
00:40:08,080 -->
00:40:14,040
as an economist most of my professional career, I also know the practical
478
00:40:14,040 -->
00:40:18,520
burning issues that the industry is facing.
479
00:40:18,520 -->
00:40:23,960
So I think I'm in the position to essentially bridge
480
00:40:23,960 -->
00:40:27,820
the two sides. So I do know the
481
00:40:27,820 -->
00:40:32,660
classical, the useful,
482
00:40:32,660 -->
00:40:38,560
important theory from economics and finance literature,
483
00:40:38,560 -->
00:40:44,540
even though right now we still don't really have the asset
484
00:40:44,540 -->
00:40:47,600
pricing model to price crypto assets.
485
00:40:48,260 -->
00:40:54,240
with all its unique features. But I know these are the pieces we may
486
00:40:54,240 -->
00:41:00,180
need to put together so that we can use these tools in this literature to help us
487
00:41:00,180 -->
00:41:05,580
solve these real -world practical problems we have in the industry.
488
00:41:06,140 -->
00:41:09,380
Kelvin Sparks: That's awesome. Thank you for that. And are you guys going to the summit at all?
489
00:41:09,800 -->
00:41:12,380
I was curious. Completely out of nowhere.
490
00:41:15,220 -->
00:41:18,240
It's fine. We can just keep going. We just keep going. I'm just trying to find a way to
491
00:41:18,240 -->
00:41:22,660
slot it in. It's been a bit difficult given how technical the conversation is.
492
00:41:22,980 -->
00:41:25,260
But no, that's fine. We can move on, John. You go ahead.
493
00:41:27,400 -->
00:41:32,180
John Wu: So, I mean, it's interesting because I don't know if it's Eric or Matias mentioned
494
00:41:32,180 -->
00:41:37,340
earlier, or maybe even Kelvin because he's an engineer by training.
495
00:41:39,460 -->
00:41:43,680
Is this, well, maybe this is for Eric because he's a pure economist.
496
00:41:43,680 -->
00:41:49,400
Well, not a pure economist, but has this been exceedingly difficult
497
00:41:49,400 -->
00:41:55,400
because these token economics and these systems were created by scientists,
498
00:41:55,600 -->
00:42:01,540
not economists? And as I talked about before, economics is kind
499
00:42:01,540 -->
00:42:07,300
of an art as well as it is a science. And I'm reminded from my
500
00:42:07,300 -->
00:42:09,900
undergrad, one of the professors was Richard Thaler.
501
00:42:10,980 -->
00:42:15,700
Some would call him the father of behavioral economics. human beings are not rational
502
00:42:15,700 -->
00:42:20,660
optimizers. Now, if we're trying to put science behind this and put it all formulaic, does
503
00:42:20,660 -->
00:42:23,480
this make this whole exercise exceedingly difficult?
504
00:42:23,800 -->
00:42:28,780
Eric Lu: Yeah, so I think the answer is yes. It's definitely very challenging.
505
00:42:30,060 -->
00:42:36,000
But I think at the same time, because these blockchains, the industry is
506
00:42:36,000 -->
00:42:40,920
actually built at the very beginning by engineers and scientists.
507
00:42:40,920 -->
00:42:46,600
there are a lot of room for improvements from
508
00:42:46,600 -->
00:42:52,240
the economic perspective. And I think that's where my expertise
509
00:42:52,240 -->
00:42:58,120
will contribute the most. So you do see a lot of areas where we
510
00:42:58,120 -->
00:43:04,020
get to apply for basic economics or finance intuition where maybe at the
511
00:43:04,020 -->
00:43:09,840
very beginning of the design of the blockchain ecosystem or the quote -unquote
512
00:43:09,840 -->
00:43:14,500
tokenomic model are not really seriously considered.
513
00:43:14,780 -->
00:43:19,100
But that leaves the room for improvement, especially I think right now.
514
00:43:19,200 -->
00:43:24,400
So as the industry mature, people have been fighting for
515
00:43:24,400 -->
00:43:29,060
adoption using better and better technology, lower and lower transaction fees.
516
00:43:29,220 -->
00:43:35,100
And now we really have to get more into these incentives of these different
517
00:43:35,100 -->
00:43:40,200
ecosystem participants how we model them, how we
518
00:43:40,200 -->
00:43:46,000
factor in their incentives into the design of the tokenomics itself.
519
00:43:46,600 -->
00:43:51,780
And those should come in right now, if not sooner.
520
00:43:52,040 -->
00:43:54,840
That would be my take.
521
00:43:55,880 -->
00:44:01,660
Matias Antonio: You know, and so like, I don't think it's because necessarily, it is
522
00:44:01,660 -->
00:44:06,680
difficult, 100%. But I don't think it's necessarily because an engineer or a scientist
523
00:44:06,680 -->
00:44:12,440
were the ones that designed it. Because the truth is, at the very early
524
00:44:12,440 -->
00:44:16,300
blockchain, it really was a game theory problem.
525
00:44:16,600 -->
00:44:22,060
And that's what Bitcoin did. It's just a game theoretic solution for alignment of
526
00:44:22,060 -->
00:44:27,380
incentives such that a network that just needs internet access for it to just work in some
527
00:44:27,380 -->
00:44:33,180
sense, right? And if you fast forward a bit later, it wasn't even a
528
00:44:33,180 -->
00:44:36,500
thing because there was no economy, right?
529
00:44:37,040 -->
00:44:41,500
even when Ethereum came out in 2017 or so, it was the FAT protocol thesis.
530
00:44:41,640 -->
00:44:46,120
It's just going to be big. But what does big mean? Nobody knows. So price go up, right?
531
00:44:46,240 -->
00:44:52,240
So there was a lot of lack of understanding because there was nothing there to really
532
00:44:52,240 -->
00:44:57,920
analyze, understand, and have a real interesting solution for that needed to be solved,
533
00:44:57,960 -->
00:45:03,740
right? As 2020 and 2021 progressed, then an economy actually
534
00:45:03,740 -->
00:45:09,540
formed. and then it's like you had the uniswap the ave the the the benchy right the
535
00:45:09,540 -->
00:45:15,180
faro that that came in and and gave it shape it gave it an economic activity and the users
536
00:45:15,180 -->
00:45:20,300
that were coming in to actually do something on the chain so now now it becomes a real
537
00:45:20,300 -->
00:45:25,720
problem where uh the complexity of the situation but also the tools that come with it
538
00:45:25,720 -->
00:45:30,400
are available because now you understand more the problem it's just that it's not that it
539
00:45:30,400 -->
00:45:35,680
wasn't present before it's just that it wasn't known what could be the solutions for a
540
00:45:35,680 -->
00:45:41,580
problem that was not observable at the time. And so actually, you know, maybe the
541
00:45:41,580 -->
00:45:47,300
fact that it was engineers and scientists that did it provided a very wide solution
542
00:45:47,300 -->
00:45:51,420
space for, you know, helping solve this problem.
543
00:45:52,600 -->
00:45:57,380
Because it's not true for every chain, but there's chains that you can do actually a lot.
544
00:45:57,940 -->
00:46:01,580
And ours, thankfully, is one because it's very well designed.
545
00:46:01,580 -->
00:46:07,540
So, you know, I think it's a benefit, but I wouldn't necessarily say
546
00:46:07,540 -->
00:46:12,620
it's because the genesis was that one, is because the problem exists today.
547
00:46:12,620 -->
00:46:15,840
It's just a matter of something that was just not observable at the time.
548
00:46:16,280 -->
00:46:22,220
Kelvin Sparks: Well said. And let's say 12 months from now, we look back on this podcast.
549
00:46:22,320 -->
00:46:28,200
What's the one proof point that you'd point to to say that the research agenda was
550
00:46:28,200 -->
00:46:29,520
successful? yeah
551
00:46:29,520 -->
00:46:35,400
Matias Antonio: um honestly i have a few uh so like one of them
552
00:46:35,400 -->
00:46:41,120
is definitely that validators are actually getting paid something above
553
00:46:41,120 -->
00:46:47,100
inflation something on top of it right is there going to be inflation
554
00:46:47,100 -->
00:46:51,160
12 months is a very short time span for the fundamental changes that we're talking about
555
00:46:51,160 -->
00:46:55,280
so most likely the answer is yes but i just want it to be that there's something on top
556
00:46:55,280 -->
00:47:01,000
because if you're just earning inflation, your net return is actually negative
557
00:47:01,000 -->
00:47:05,860
because you have costs, right? So if you have now something that is on top of inflation,
558
00:47:06,020 -->
00:47:09,020
now your real return is above zero.
559
00:47:09,360 -->
00:47:13,500
And that's where I think it starts becoming much more sustainable.
560
00:47:14,000 -->
00:47:19,880
And the other side of it is that we have found a way to capture
561
00:47:19,880 -->
00:47:25,060
sustainably in a non -disruptive manner. And that's something that Eric keeps saying,
562
00:47:25,200 -->
00:47:30,380
which is super important because it has to be incentive -aligned because otherwise you're
563
00:47:30,380 -->
00:47:36,160
distorting and you're actually driving users away or driving economic activity
564
00:47:36,160 -->
00:47:40,380
away. And that's not the job here. How do you find that
565
00:47:42,820 -->
00:47:47,260
game theory solution for this incentive -aligned problem? And so if we're able to capture
566
00:47:47,260 -->
00:47:52,540
the value and have something that is significantly higher, like orders of magnitude higher
567
00:47:52,540 -->
00:47:57,540
than what is currently captured, that's the other thing that I would say, okay, success.
568
00:47:58,540 -->
00:48:02,800
So value accrual and sustainable value data economics.
569
00:48:03,380 -->
00:48:08,280
Eric Lu: Yes, very well said. I guess I'll just add something quick.
570
00:48:10,080 -->
00:48:16,060
I think from a practical perspective, everything that we're going to put out
571
00:48:16,880 -->
00:48:22,620
that either proposes any changes to the architecture,
572
00:48:22,880 -->
00:48:27,240
the tokenomic design of the ecosystem, we do it for a purpose.
573
00:48:27,400 -->
00:48:31,560
So maybe it's mostly around capture or distribute.
574
00:48:31,860 -->
00:48:37,640
They must have a server purpose. And they should always have a verifiable
575
00:48:37,640 -->
00:48:43,500
outcome. So after we implement those changes,
576
00:48:44,400 -->
00:48:49,960
for example, do we see a significant uptick
577
00:48:49,960 -->
00:48:55,500
in the non -inflation -funded validated rewards or protocol revenue,
578
00:48:55,780 -->
00:48:59,260
that's something we can verify after the change.
579
00:48:59,540 -->
00:49:05,320
And beyond that, are we reducing or replacing the
580
00:49:05,320 -->
00:49:11,220
inflation -funded staking rewards with some additional real
581
00:49:11,220 -->
00:49:16,900
source of revenue for the protocol? then that's another example of real verifiable
582
00:49:16,900 -->
00:49:22,660
outcome that we can look at. So if all those checks out, I would say, yeah,
583
00:49:22,860 -->
00:49:25,720
we achieve what we set out to achieve.
584
00:49:26,520 -->
00:49:31,100
John Wu: I do want to mention one thing that's very important here, and what we haven't talked
585
00:49:31,100 -->
00:49:37,100
about, is the reality of what we're talking about here is just related to
586
00:49:37,100 -->
00:49:42,360
the main Avalanche chain. The network of networks is what Avalanche is about.
587
00:49:43,000 -->
00:49:47,640
and there's a lot of hidden value in all the other L1s.
588
00:49:47,960 -->
00:49:52,840
So we talked about this as a co -op earlier. We're only talking about co -op on the
589
00:49:52,840 -->
00:49:56,580
Avalanche C -chip. It's almost like a network of many co -ops.
590
00:49:56,880 -->
00:50:01,960
And so the difficulty of just trying to find the optimal part of the curve for this one co
591
00:50:01,960 -->
00:50:07,580
-op, and then we have to balance with that, how do we connect it to the hundreds of co
592
00:50:07,580 -->
00:50:12,420
-ops that are tethered to Avalanche? That's going to be a whole other challenge.
593
00:50:13,280 -->
00:50:19,280
And, you know, let's work on getting one done, but we'd love to figure out how to
594
00:50:19,280 -->
00:50:25,280
get that trap capital or the dormant capital that's sitting in the other L1s and
595
00:50:25,280 -->
00:50:26,520
making it also productive.
596
00:50:27,500 -->
00:50:33,040
Matias Antonio: 100 john and that's that's actually top of mind as well we focus on the c -chain because
597
00:50:33,040 -->
00:50:38,960
that's that's the thing from the design space the closest thing at hand but what the
598
00:50:38,960 -->
00:50:44,880
the the goal after that is is how do we get the token to grow with the growth
599
00:50:44,880 -->
00:50:50,600
of the entire ecosystem not just one area And so we're
600
00:50:50,600 -->
00:50:56,540
hoping to also find solutions for, or be it by focusing on this area
601
00:50:56,540 -->
00:51:02,320
first, is that it gives us a sense as to what are potential scalable solutions
602
00:51:02,320 -->
00:51:07,820
that we can implement more broadly as well. That can give things to these other chains
603
00:51:07,820 -->
00:51:13,300
that they would need and also, you know, and therefore strengthen the value accrual
604
00:51:13,300 -->
00:51:17,540
mechanism, not just it relates to seeding, but it's the entire ecosystem as well.
605
00:51:17,540 -->
00:51:23,460
There are some technical challenges with that at the validator level that
606
00:51:23,460 -->
00:51:26,260
we may have to think about.
607
00:51:26,580 -->
00:51:32,300
But absolutely, that is one area of research that we are very much aware
608
00:51:32,300 -->
00:51:33,840
of and looking at.
609
00:51:33,840 -->
00:51:39,640
John Wu: Yeah, that would be huge because Avalanche, more than
610
00:51:39,640 -->
00:51:45,120
any other layer one, has far more
611
00:51:45,120 -->
00:51:50,180
dormant assets than understandable by the average
612
00:51:50,180 -->
00:51:55,080
explorer, if you will, because the explorers don't capture any of that.
613
00:51:55,080 -->
00:52:00,780
I mean, it's multiples. It's almost like if you're just a real estate
614
00:52:00,780 -->
00:52:04,760
company capturing just the cash flows associated with the apartment buildings or the
615
00:52:04,760 -->
00:52:10,080
shopping malls that are active, not counting for all the real estate
616
00:52:10,080 -->
00:52:15,180
Kelvin Sparks: you actually own. very very well said and uh yeah i appreciate all you guys taking the
617
00:52:15,180 -->
00:52:20,120
time out to chat today but unfortunately we actually ran a bit over so that is all the
618
00:52:20,120 -->
00:52:23,920
time that we have today matthias eric thank you for taking time out to chat with john and
619
00:52:23,920 -->
00:52:28,600
i and uh as a reminder our newsroom works tirelessly to get you accurate informed crypto
620
00:52:28,600 -->
00:52:32,760
news if you want to stay ahead read the block thanks see you next time
621
00:52:40,840 -->
00:52:41,920
You
