On June 4th, we hosted a live AMA Spotlight with Yang Li, Cofounder at Bitlink Capital.
This is the third AMA that we're doing in our series focused on market making. Bitlink Capital is a proprietary trading firm conducting high-frequency trading across a variety of cryptocurrencies, markets, and multi-strategies. They are one of the biggest market makers on dYdX Layer 2 perpetuals.
Yang discusses:
His backgrounds and the story behind Bitlink Capital
Hedged vs. non hedged market maker strategies
Market color on dYdX liquidity, volume, order flow, and bid-ask spreads dynamics
Bitlink’s strategy to manage inventory, risk, collateral, and credit
Impermanent loss dynamics on AMMs
The future of market-making in DeFi
The video interview is available here.
A revised transcript is available below:
Vijay (dYdX): Hi everyone. My name is Vijay Chetty and I'm the Head of Business Development at dYdX. I'm excited to be joined by Yang Li, Cofounder at Bitlink Capital. This is the third AMA in our series focused on market-making and trading on a decentralized exchange. Bitlink is a proprietary trading firm conducting high-frequency trading across a variety of cryptocurrencies, markets and multi-strategies. They're one of the largest market makers on dYdX Layer 2 perpetuals as well as a number of other exchanges. And so in the next hour we're extremely excited to discuss community questions around Bitlink and market-making on dYdX more generally. So remember to please submit any additional questions in the chat on Discord in the AMA channel or on the YouTube live chat and we're happy to address those at the end.
Yang, so thanks for joining today. So to kick things off, could you introduce yourself a bit and the founding story of Bitlink and how you guys got to where you are today?
Yang (Bitlink): Vijay thanks for having me. I'm really excited to talk to you today. This is actually the first time I do this AMA on live so I'm a little bit nervous. A little bit background about myself. I studied financial engineering at Duke University. After graduation, I went to an Investment Bank to do the entry level as an investment banker. Back then it was 2015, I started to realize there was a huge opportunity in crypto. So I started to do arbitrage across different exchanges. It went pretty well. So I thought, why not find some other people, other folks who are also doing the same thing then I found my partners. So from 2015, we pretty much incorporated a company. And in 2016 we became the first market maker on Huobi. Of course, back then that was a centralized exchange.
We then expanded to other exchanges worldwide, mostly exchanges as you already know like OkEx, Binance, Bitstamp, Liquid, Bitfinex, you name it. And in 2019, we started to work with you guys, with dYdX. We got such a great time. We started to make a lot of markets on DeFi. And from last year around I think September, Bitlink got really deeply involved into a lot of DeFi projects. And I'm really excited to have this tonight talking about market making on decentralized exchanges with you. So that is the general background about me and the company.
Vijay (dYdX): That's great. So it sounds like you guys have been quite active in the crypto space more broadly for a long time. Is there anything that you'd say in terms of like the challenges for market maker very early on in crypto and connecting into the early exchanges like Huobi and OkEx compared to now and sort of what you're seeing in the exchange ecosystem more broadly? Are there some shifts in that?
Yang (Bitlink): This is a really interesting question because back then when it was 2016, when we first got connected with Huobi, the infrastructure of the exchanges are not complete at all. Like the latency and everything was beyond absurd. If I remember correctly, the latency to get a quote from the centralized exchanges is about two seconds. But now I think about it, it's probably a less than 10 milliseconds. Think about it, there's like 20 times difference.
And when it comes to the comparison of the decentralized exchanges at this moment, I think right now from the API perspective for market maker, there's not much difference. We pretty much connect to every exchanges, be it centralized or decentralized, we are totally comfortable with them.
Vijay (dYdX): That's good to hear. So stepping back a bit, what role would you say that market makers play in making crypto markets efficient? Because very early on it was really just kind of a retail focus market and now obviously there's a much wider range of players. So what would you say is kind of the key role of market makers?
Yang (Bitlink): I think in essence the role of market maker hasn't changed that much. The argument we always have is we try to make the market more efficient, letting people to buy things or sell things more fluently and efficiently. The difference I would say changes from 2017 is mostly in the product I would say. Remember in my own story, I started to do arbitrage manually, by hand in 2017. I tried to buy low and sell high between different exchanges. There's not much difference compared to what we are doing right now, except we are doing that much faster and probably smarter as well. But remember at that time, most products that I have traded were actually spot, like BTC-USD, BTC-USDT those kinds of things. But now with the derivatives products launching, especially after the BitMEX who was I would say the founder of this, the inventor of this. And now a lot of exchanges are trading derivatives, especially perpetuals like dYdX. So now market maker has a huge responsibility in order to bring the market, let's say either premium or discount back to the market price for people to trade fluently and efficiently into this market. But in essence, I would say the responsibility and the job that we are doing hasn't changed that much. Does that make sense to you?
Vijay (dYdX): Definitely agree in terms of the impact of the perpetual product structure on the greater market. It made these more complex instruments more easily kind of accessible and understandable to a broader audience and has been a great on-ramp for institutions as well. I think in fact, we just recently passed the anniversary of the perpetual itself.
So could you speak a bit to kind of the market making strategies that you guys deploy and generally, presumably you're staying market neutral but are there kind of specific high-level strategies that you guys are using?
Yang (Bitlink): I think this is a very deep question. Before I answer, explain our strategy, I help the audience to understand a little bit more about the traditional way to market make. So traditionally, I would say people when they try to make a market on a certain exchange, they do that in a hedging style. Meaning whenever they buy something on one exchange, they sell it on another exchange. Whenever they sell something, they buy it from other exchanges. It's very easy to understand. You buy something at one place and you immediately sell it in another place. And when you sell something at one place, you immediately buy it from another place. You minimize your exposure to the market volatility. So this is a typical way a lot of market makers are working.
But it has a lot of disadvantage apart from it's so easy to understand. You may involve a lot of inventory costs. You may have to add your managerial costs. You may lose money during the volatile time. Our method is very different from those exchanges. And this is what we are doing on dYdX. And I think this is one of the reason it makes us as one of the biggest market makers over here.
So our style, our market-making style is called non-hedging style. We use the data we gather from the exchanges worldwide to create a model that finds a crypto price of the token that we want to make. And we are always trying to buy low and sell high based on the model suggestion. Say we price the model as a hundred dollars and then we are always waiting to buy it at 99 and sell it at 101. When the market actually moves alongside it as our model predicts and the data refreshes from the exchanges, our pricing model changes as well. So we either change our order along the way, whenever our model suggests us to do, so we cancel the previous order and put a new order over here. By doing this, we do not have to trade, to put inventories on other exchanges, to manage all the withdrawal and deposit procedure. Instead, we can minimize the hedging cost as well so that we can offer a very tiny web on dYdX for people to trade. Meaning people can buy things at a cheaper price and sell things at a relatively higher price compared with a hedging style market making. So it actually benefits not only us, but also people who trade on dYdX.
Vijay (dYdX): That's very helpful. And what would you say is the impact of that strategy on uptime and the uptime that you are able to maintain in terms of using an un-hedged market making strategy versus hedged?
Yang (Bitlink): Right, right on top. Because as I explained, when you try to make market on exchanges by using hedging styles. I think for the sake of understanding, we will use spot exchange as an example like the previous layer one dYdX. When you try to make market for example on ETH-USDC and you use a hedging style, from time to time, your balance, your position will be skewed based on two different exchanges. You may be long on one exchange because of the nature of buying over here, you have to sell over here. So you are short on the other side. From time to time, you run out off your balance. So you have to rebalance between these two exchanges. That is the time you have to withdraw and redeposit to the other exchanges. Like you do that two way. The denomination currency and for example, BTC or Ether. So during this time, you either have to stop your strategy or you have to put actual money in order to guarantee you can always trade alongside this procedure.
So it either decrease let's say the capital efficiency or the uptime. However, Bitlink’s style, because it is non-hedging style, we do not withdraw or deposit money at all at an exchange. So that by doing so, we can always quote on the exchange no matter what, except we are at our full capacity. For example, we run out off our balance to buy or sell. Only in this kind of circumstance, we will stop, otherwise we can always trade. So this is I think one of the major advantages over hedging style our market making.
Vijay (dYdX): You guys are able to keep quotes in the book and almost all market environments and conditions.
Yang (Bitlink): Especially during the volatile time like March 12th last year and May 19th this year. Because the market is extremely volatile, a lot of market makers who are using hedging style market making think about it because the price always get a sequence of moving. It always happens for example, on the most liquid exchange and then to the second most liquid exchange and go on and go on. So it may have a slight latency between exchanges. Think about it, when people are doing hedging style market-making, they always use the most liquid exchange. For example, let's just admit it, most people use Binance. And in this case so you use Binance as your hedging exchange, you buy something on for example, dYdX, you have to sell something exactly the same amounts on Binance.
So when Binance price has already dropped from previous let's say 100 to 80. And then you get a hit, you just bought something at a 99 and you want to sell it on Binance. Previously probably you can sell it at 99 as well, but now Binance is already 80. So how are you going to do that? You either wait for it to go up a hundred, which doesn't make any sense at all because you never know whether it will go up or no. So you have to immediately sell over here. And you immediately take the 20 loss over here during the volatile time. That's why a lot of market makers, probably audience can notice, during the volatile time, a lot of markets like on most changes, you will see close to an empty audit book, that is the reason why.
A lot of market makers stop to quote over there. Because if they quote, they lose money. They don't know, they don't understand, they have no ways of predicting which way the market will go. That's why they stopped quoting. But for Bitlink it’s a little bit different. We use the data, we fetch from the exchanges from let's say the data we gather. And by going through our model, we are still able to quote during this volatile time. And we are confident enough even during this really volatile time, we are able to provide liquidity and to make money out of that.
Vijay (dYdX): That's very helpful background in terms of kind of the flows and the trading especially as tools like Humming bot help to kind of make this type of trading more accessible to a broader audience, both hedge market making, unhedged, and arbitrage. Yeah, it's definitely helpful I think some market education overall. So switching gears a bit to product focus. So what would you guys say is kind of the split of your focus on spot versus more of like derivative structures, whether it's perpetuals, termed features, options or kind of structured products?
Yang (Bitlink): We don't have exact data about the split between the trading volume on different products. Because as you know, we are trading probably if not hundreds, but at least dozens of pairs. But in general, I would say that derivatives, especially perps is larger, it is probably the largest trading product on the market so far. It's very easy to understand that because of the leverage that you can use with the same amount of money, you can trade it with a much, much higher amount compared with a spot exchange, compared with a spot product. And the deliverable future for example on OkEx is also very huge, but I think the popularity has gone to the perps at this moment. Because the perp is so easy to understand, it is very instinctive. So from Bitlink’s perspective, we trade more on the perps probably following, I'm not so sure either spot or deliverable future.
Vijay (dYdX): So what would you say a day in the life at Bitlink looks like and what do you guys look out for in terms of news cycles and things like that?
Yang (Bitlink): The day at Bitlink. Wow, this is, as people always say, there's no same day in crypto, pretty much. It's pretty the same for us, depending on what kind of role that we have at Bitlink. For example, for me, I'm more of a business guy. So I tend to have a lot of meetings with our clients to discuss, okay what was going on yesterday? How is our performance on different exchanges? And probably have a meeting with my traders as well to understand what is a general situation that we are facing today or problems. So this is more of my typical day.
We don't have the exact time range that we work, it is not nine to five, nine to six. In Chinese internet companies, they always say 9, 9, 6, which means from 9:00 AM to 9:00 PM, six days a week. But for us, it's more like 0, 0, 7, we work from 0 to 0 and seven days a week. That is if we go our way. Right now it's almost 11:20 PM Beijing time. And we are having this conversation. The very challenging, exciting but also tiring day for market makers or most people who are in the market of crypto.
Vijay (dYdX): Completely agree and understand what you mean by that. So what would you say are some of the biggest kind of macro and geopolitical considerations that Bitlink pays attention to for informing kind of overall positioning. And then related to that, do you guys ever run net long or net short at times?
Yang (Bitlink): So there are two questions. Let's start with the first one, the geopolitical consideration. Yes, we do. Because as I did my introduction to you, I mentioned that we are actually trading across different exchanges worldwide. We trade a lot, not only in the crypto to crypto exchanges, but also some fiat exchanges such as Bithumb or other Korean exchanges that I mentioned and also the Japanese one. In these countries, the fiat exchanges is more regulated, and also American sorry, like Coinbase, Kraken, Bittrex. So it is often the case that fiat crypto exchanges are more regulated than the pure crypto to crypto exchanges. So in those exchanges, we have to pay a lot of attention to the regulators and what kind of new policies they issue. It affects our positioning and our strategy. Because oftentimes the regulators probably does not understand that much about the way we trade. If we do the arbitrage between exchanges, we would probably make a lot of money on one exchange but lose a lot on other exchanges or our profit is very tiny. But if they ask us to pay tax on the profit side, then it doesn't make sense at all. We have to change our model in order to avoid this. So this is for the geopolitical consideration in different countries.
And your second question, your second question is our position. It has to go back to the model I just mentioned, our non-hedging style market making. In t traditional way of hedging style market making, market makers tend not to have any positions because they always buy on one exchange and sell it on another one. So you always stay delta neutral. But for us, it's a little bit different. Since we don't hedge on other exchanges, everything happens at one single exchange. When a client wants to buy, we have to sell that to them in order to maintain the liquidity on the exchange, if we have any obligation.
So in this case, if the momentum of the market is buying on a certain exchange, then probably we're in a short position. We always use the inventory, the price, we always adjust our price based on the inventories we have. Meaning even if I'm in a long position or short position, I am able to still quote which means I do hold positions and it is the case on dYdX and other markets that we make.
Vijay (dYdX): That makes sense. So almost a part of the strategy itself involves often having a net long or short based on what the model is.
Yang (Bitlink): We tend to make our price, or our position to be neutral. But it's not always the case. If the client wants to buy, we have to sell, we have to sell anyways, because this is our obligation. Especially for a lot of new exchanges, when there is not enough liquidity, probably you are the only one who provides liquidity over there so you have to sell. But in order not to lose money or to maintain the profitability, we always adjust the price based on the inventory. We have, for example, giving you a very brief example. If the market momentum is buying, even if the model suggests that we should only buy at 99 and sell at 101, but if we are in a very, very extreme short position, then we are willing to bite higher than this 99. For example, 101, 102 in order to make our position back to neutral. So this is, let's say probably more related to inventory management.
Vijay (dYdX): Switching gears a bit. How do you guys stay on top of all the DeFi activity and determine which projects to get involved with? Obviously, there's kind of a very hectic pace of innovation going on in DeFi these days. How do you guys think about that, and are there kind of specific metrics or statistics that you use to determine the good projects or ones to get involved with?
Yang (Bitlink): As people know, the crypto industry is booming, but also has a lot of chaos let's say. So it's really hard to determine which project is really good. So you have to pay a lot of efforts to get to know these projects. For example, with dYdX. We knew you guys probably almost two years ago, before all this DeFi booming. We tend to be really cautious at the beginning. For example, not putting too much money into the exchanges or asking for some protection based on the inventories that we put and probably get a reference from friends or other projects. We tend to believe that great people lead to great people. This is the typical way that we work with the projects.
Vijay (dYdX): Definitely agree with that as well that people are often the most important ingredient in the community. Switching gears a bit to more of a market color and market context, can you describe at a high level, the liquidity and order flow dynamics that you've seen on dYdX and how they compare across he large caps for small caps to the centralized exchanges that you guys have been active on as well? And of course, DeFi is still early in its development, but just in terms of kind of what interesting trends that you might be seeing in terms of the comparison.
Yang (Bitlink): So from the DeFi perspective, I think dYdX is absolutely the biggest perpetual trading exchanges. No doubt, no doubt at all. I don't see any competitors at this moment. So in terms of the order flow, volume, and liquidity, all rank as number one in DeFI industry, especially DeFi perpetuals on Ethereum, there's no doubt about it. But as you said, the DeFi industry is actually at a very nascent stage, it's very early and very young. And the centralized exchanges have existed for probably as long as 10 years if you consider some local, like some exchanges established in 2010 or 2011.
So with like 10, 11 years of evolvement and development, I think the overall trading liquidity, volume and the spread on centralized exchanges are still much better than the DeFi projects, it's just a fact. But I think with the, especially the effort that done by Sushi and Uniswap and dYdX, the trend is very clear to see at this moment. DeFi is catching up at a really frenzy speed. It’s getting more and more attention, not only from the crypto industry, but also from people outside of the industry. So I think the overall trend of the DeFi and the liquidity and everything will catch up very, very soon, probably faster than we can imagine.
Vijay (dYdX): And maybe room for both to coexist as well long-term, kind of serving different needs. So what about from a funding rate standpoint? So speaking to how the funding rate for the dYdX perps compares to kind of the funding rate for perps in the broader market. So like FTX perps and Binance perps and kind of any interesting kind of trends or opportunities that you guys see.
Yang (Bitlink): So I think the biggest difference is probably the timing interval of the perpetuals on dYdX and more of the traditional perpetuals probably designed by BitMEX. So on BitMEX, the funding rate is calculated every eight hours, but on dYdX, it's calculated on an hourly basis. That is the difference. The second different is about Oracle. So on the traditional centralized exchanges, they use the mark price from more centralized exchanges, such as Coinbase or Kraken, depending on what kind of pairs and the denomination currencies. It's always from those relatively big centralized exchanges.
But I think on dYdX, since it's a DeFi, project it is more from the Oracles, from Chainlink if I remember correctly. So probably the ultimate source is the same but the direct source is different. They fetch the data directly from the exchanges while you guys get the data from another DeFi project. This is I think the two differences for the product itself.
And if you talk about the trading experience for the client, then it's another story. On centralized exchange, all you have to do is log in, log out and trade over there, money stays over there. Here, you have to manage your own private key. You have to understand exactly what you are doing. You have to understand what is gas and et cetera.
Vijay (dYdX): So speaking a little bit about risk management and technology now. Obviously, the rapid price increases and flash crashes are very common in crypto and depending on the market environment that we're in. So how do you ensure that you're able to provide liquidity in all market conditions and more specifically, what is your playbook when a market crash is happening? What are the steps that you guys are going through?
Yang (Bitlink): I briefly mentioned the strategy we use, non-hedging style and the hedging style. So if you follow that logic, you will notice the speed is actually the essence. You have to be able to fetch the data, fetch the price really, really quickly and use your algorithm to calculate the correct price really, really quickly as well. And then execute your order in an extremely short period of time too. So you need to be very, very fast. That is the essence. So we try to get a lot of involvement with the exchanges, especially the exchanges with a stronger weight for our pricing model in order to get a co-location, for example, a more stable API, and an API with a higher limit per quote, per request. By doing so during the market crash or the volatile time, because the request is too much for a lot of exchanges, they sometimes get jammed or congested or get us, let's say a longer delay than expected.
We try to avoid all these issues by getting a co-location or by having a higher limit per request. By doing so, we make sure we are always at, if not the fastest one, at least the one of the few fastest ones on the market to get to the correct data, and then feed that data to our model. And this way, when we do that, we have to make sure our algorithm works really fast in order to give us the correct coding price that we should put on the exchanges and then execute the order really, really quickly. By doing so, we can have some advantages in coding the market to provide liquidity over there. And often case since we are one of the fastest players on the market, we tend to understand or get data probably faster by one or 10 to milliseconds than other people.
In some way, it looks like we can predict the market. We can predict the market in that very, very short period of time. If we notice, for example, the price has already dropped to 80, like from 100. And a lot of people who are still using limit order on dYdX, they haven't canceled their order or they are still waiting to buy at a certain price, for example, 90 or 95, then we will take that order, we will sell it to them because our models suggest the fair market value should be 80. If someone is waiting to buy at 90 or 95, we are going to take it. So not only we just make market on exchanges. During the volatile time, we tend to take orders as well because we have to take orders in order to give other people the correct price for them to buy or sell.
This is the way we work during the market crash. So often in this time, because the market is crashing and a lot of people tend to be slower to cancel their order on the buy side. We always sell our positions probably during this time. So we probably will have a short position overall. We sell all of our stuff at 95, 90 and then re-quote at 80. But this is not always the case because you sometimes see the V shape rebounding during the market crash. And in this time, same logic because we detect the price actually should be higher, it shouldn't be 80 anymore. It should be 100. And then we buy all the way up to 100. And in this case, I think we are in a very long position so we can either be extremely long or extremely short during the volatile time.
Vijay (dYdX): So what would you say is the impact of an approach like this on price discovery? Because it sounds like it really helps to the extent there is a kind of a fair market value. It can help drive towards that.
Yang (Bitlink): Yeah, exactly. This is the way that as I said, the job of market makers hasn’t changed that much. The essence is to find the crypto price for the market and for the people who want to trade against it. So during this market crash or bump, what we are doing, what we were doing and we are doing right now is to make the price, make a fair market price to the client who wants to trade at the current market situation. So this is the definition I would say of price finding.
Vijay (dYdX): Now you had mentioned co-location and other capabilities that probably map a little more to more of a traditional finance or centralized crypto model. How do you think that these ensuring that you're getting the most current data as quickly as possible and responding to it? How do you address that in a DeFi context where maybe there's not always an opportunity for something like co-location. How do you guys think about that? Or maybe it's a little bit early and it's still something that is a TBD.
Yang (Bitlink): No, no, we are actually considering all of this. I think probably this is a charm for the DeFi projects because as you said, if we are both on Ethereum, there's pretty much no co-location thing because everything happens in the same transaction. In order to probably have a co-location, you have to work very closely with the miner, probably take your transaction as a priority, always bundle that into the block. So you somehow get a sense of co-location, if that makes sense to you. But this is a very costly and probably not an efficient way to do that. DeFi as you said, is still in the very nascent and early stage. So a lot of things are changing alongside. We get Solana like we get a sequence so you cannot front run anybody else. You get another project that can allow sort off co-location. We don't know. We haven't got a fixed pattern for DeFi projects. We have to evolve and develop along with the DeFi projects.
Vijay (dYdX): Exactly. So what would you say are some of the biggest risks that keep you guys up at night? Obviously, the bots are running 24/7. So what keeps you up at night?
Yang (Bitlink): Probably most people can think of when market is really volatile we may get liquidated or like wiped out by the market. Yeah, probably, but we've been in this industry for five, six years. We experienced a lot of extreme cases and we are still surviving because of the strong model and the risk management system that we have internally. This is won't keep me awake during the night. At the very beginning, at Bitlink, I couldn't sleep during very volatile time, but now I can.
The biggest challenge actually we have right now is the counterparty risk that we have with the exchanges. A lot of centralized exchanges are very centralized as you understand. So we don't know what will happen to them. And if something happened to them, will our asset be safe. We are not so sure about it. So this is something that we are really concerned about. And I think this is the reason, this is the exact reason why DeFi has such a promising future because you are actually controlling your assets, your money at your hand. So I think this is my answer to what keeps me at the night.
Vijay (dYdX): And then it's more about ensuring that you are interacting with proven smart contracts that have a long track record.
Yang (Bitlink): Right, right. This is really great. Especially for a lot of retailer users, because even for us, it's really hard for us to distinguish which project is really solid or it doesn't have any like bugs or loops in its code. It's really hard for a lot of non-coders I would say. Not everyone is a hacker or really good at computing, coding. So the auditing or the some sort of guarantees probably from third party or from, I don't know, DAO or whatever is very important for people to participate in the DeFi protocol. As you said, the nature that we worry about is the safety of our assets. Be it centralized exchanges or decentralized exchanges. Even if it's a decentralized exchange, if the intent is to do something, to steal your asset, they are still able to do that by leaving some bugs into their code. So always choose the projects that you can trust.
Vijay (dYdX): So switching gears a bit and talking about what's ahead and kind of the future of the industry. So with the rise of AMNs, traditional market makers were becoming a bit more kind of disintermediated by liquidity pools. But now we've had a return with Uniswap v3 to order book principles and the constructs more generally. And so how do you see the role of market makers in kind of the evolving DeFi space going forward long-term?
Yang (Bitlink): This is a really good question because UniSwap, the AMM model, greatly contributed to the booming of the industry, there's no doubt about it. But I think the order book style may prevail in the future. This is my guess.
Think about an AMM model, think about an AMM model. It is very popular from September 2020 by SushiSwap. They did a really amazing job to promote the industry. And you put a certain amount of token on one side and the other side like equal and then you just leave it over there. And remember, that's probably the beginning of this bull market. So from Ethereum jumping from $200 to right now, like almost $3,000. It's a 15 times increase in not even one year. So during this process, the money you put over there, your only impact is the impermanent loss. So if you hold that capital, you hold your assets from that time to now probably you make $100. But now you're only making $80, let's not talk about the mining, I'm just talking about the asset itself.
You are exposed to $20 income in loss. But it's okay, you are actually just making less, you are not losing money. It's only a matter of whether I'm making $100 or $80. People are very insensitive about the money that they didn't make. They care about the money they lose, this is the nature of people. They care about the money they lose, but they don't care about the money they didn't make.
So this is the bull market. But there is not always going to be a bull market. At a certain period of time, there is always going to be a bearish market. During that time, not only will users be exposed to the impermanent loss because of the price movement, but also the asset itself, because it's dropping from a very high-level price to a relatively lower price. During this process, you are impacted by two losses. One, the asset itself. Second, the impermanent loss. So the TVL, meaning total value locked in certain exchanges, will drop dramatically like a cliff. It's going to be like a cliff. This is what will happen for a bearish market. Then people will suffer during this time because, they won't feel the money they didn't make, they will feel the money they lose and lose badly. So this is the future, this is probably what will happen for the AMM model. And the Uniswap actually already realized that. Previously, their model, their AMM model is from zero to positive infinity. You can trade along this infinity range. But now, you can only put your money into a selective range during the process. Think about it, if your things the range is really, really tiny, even to a thin line, then it's a limit order. It is a limit order.
So actually Uniswap is thinking about limit order at this moment. And they are migrating from the AMM to limit order. Because the current fee structure that you have to quote on Ethereum is still relatively very high, a lot of high-frequency market makers haven't traded, haven't tried to use limit orders on Uniswap. But I think with the Layer 2 and probably the POS of Ethereum, when the fee drops dramatically, probably one out of 10 thousands, then market makers are able to trade over there by using a limit order. It pretty much works as an order books style, even though for a lot of people, probably they still argue this is the AMM model, but in reality, this is a limit order, this is an order book.
Vijay (dYdX): That was are really good explanation in terms of highlighting how the two compare to each other. It's a good point that will become increasingly important to understand from the ecosystem more broadly. So it sounds like in terms of the role of market makers, that regardless of the trading or exchange model is, it sounds like you are seeing a role for market makers more broadly in DeFi.
Yang (Bitlink): Right, right. And if you talk about DeFi, I think now people are trading, as you said. We are just trading or copying the mature products from the centralized exchanges. Like you are trading perpetuals, you are trading spot, probably you're also trading options. A lot of protocols are doing that. But I think what is the future for DeFi apart from all the current products we have? It's about interest rate, it's about actually fixed income. This is very different. This is going to be the innovative product on DeFi other than the centralized exchanges. Because think about it, with so many lending protocols, Compound, or lending protocols on other chains. A lot of people deposit a lot of money to these lending protocols. And a lot of people are also borrowing from these protocols.
So think about it. It's actually the debt. You can structure the debt, you can securitize the debt and then because it already exists on DeFi, all you have to do is tokenize the debt and sell it on the market, probably on the dYdX, sell it at a certain price, making it as a fix, and then bundle this fixed debt together to create some new derivatives like interest rate swap, all these sorts of things. Like DeFi is building the foundation for the future derivatives of the interest rate swap, of the fixed income. So I think, I don't know when it will happen, but it will definitely happen. And DeFi already, like the current infrastructure in the system is already nurturing this innovation. This is what I think will evolve in the future.
Vijay (dYdX): That's super exciting to hear, and definitely agree with that. I think we're kind of entering the phase of the market where it's now kind of opening up new products and new use cases compared even to the traditional or centralized role so it's really exciting. So last question I had before transitioning over to user submitted questions is a follow-up to market makers taking larger roles in the DeFi ecosystem in general. How do you think that will continue to evolve? What do you see as the role of market makers for participation in governance systems and the governance processes and DAOs. And do you see market makers taking a bigger role?
Yang (Bitlink): Of course, of course. Actually, the most project I would say on DeFi are finance related. And if it's a finance-related projects, liquidity is always the priority. So market makers are the guys who provide the liquidity. So it put us into let's say the central role for the DeFi project. And that's why I think a lot of market makers who previously anonymous are now getting to the public. You got Wintermure, you got Three Arrows, you also got a Bitlink. So not only do we provide liquidity to exchanges, the institutions I mentioned, including Bitlink as well, we have some investment arm so that we can invest in the DeFi project as well. So by doing that, we get a deeper connection with the DeFi projects.
So by doing so, we gain a deeper connection with the projects and they get a deeper trust with us as well. And we can make the whole ecosystem better by providing liquidity at the very beginning. And also by giving opinions or suggestions on the finance-related questions, or the program the projects we will cover. So I think in the future, or at present let's say, a lot of market makers will probably being more exposed to the market. People will know us more. Often we will be the KOL. There are already some of my fellows becoming KOLs on Twitter. And I think this trend will continue.
Vijay (dYdX): I think another interesting positive side effect of that is education and sort of financial education more broadly. Obviously, market makers and trading firms have a very good understanding of the products and trading itself. And so I think we're starting to see kind of the impact of more education and more content coming out there. And then with tools like Hummingbot users can start to become more sophisticated themselves. So yeah, it's great to see, especially all the explanations that you provided earlier as well.
So let's maybe switch gears a bit and just check to see if there are any questions that we had submitted. So we had a question come in to explain impermanent loss and your point around the loss aversion and sort of how that's flipped in a bear market.
Yang (Bitlink): So during the bearish market, during the bearish market, for example you put $100 to the AMM and $100 worth of token over there. And now we are entering the bearish market. So during this process, the previously 100 worth of token, now only for example, only worth $10. And think about it, if you always hold your position from the very beginning, not put it to AMM at the very beginning, to the end when the price drop to $10, how much money will you have at this moment? You will have $100, it's always there, and $10 worth of tokens. So overall, your net position will drop from $200 to $110. And if you put that into the AMM, then it's very different. When the market price is dropping from a $100 to $10 during this process, during this process, you are always buying the token from 100 to 99. It is continuous, but I just discreetize that. So from 100 to 99, 98, 97, all the way down to $10. So in this way, this is not the correct way to calculate that, but you can think it, you are buying it on average from $100 to 10. You are buying it on average at let's say 110 divided by 255. But actually, you are buying it at even a lower price so let's say. During this process, you hold a lot of positions on the token itself and you spend a lot of your USDT. So if you calculate at the end of this, when the price drops to 10, your net position will be less than 110 compared to what you hold from the very beginning.
So by doing so, think about it previously, you got $200. If you hold that, you only get 110. But now if you put it out into AMM, you probably only get 80. Because I just put a random number, I don't know the exact number that it should be, but it's definitely less than 110. So you are not only losing this $90 of the actual price of the price movement, but also a certain amount of the impermanent loss happened along the AMM process. So you are losing more money. That's why I mentioned it's double impacting your net assets.
Vijay (dYdX): And maybe just to wrap things up here with a couple of minutes that we have left, could you speak a little bit to what you see as the challenges of getting more China and greater Asia users into DeFi, both from a DeFi product standpoint as well as from trading on a decentralized exchange standpoint? So what do you see as the key things that still need to happen or what's missing?
Yang (Bitlink): Right. So I think the first one is about language. Because a lot of our projects only support English or Western languages I would say at least from the beginning. A lot of Asian customers are not that familiar with English or let's say proficient in English. So it is the very first step, the very first obstacle that Asian users are facing if they want to trade on DeFi. The second part is more about the tools that we have to use for DeFi. I think a lot of people use MetaMask in order to trade on all different protocols. But the hurdle to understand, let's say the hurdle to know exactly how to use MetaMask, to set the gas fee, to understand all this DeFi knowledge, is still not that easy to understand for a lot of regular retailer users.
So a lot of education still needs to be done. I think this is the third part, the obstacle in order to attract more customers to the DeFi industry. The third one is about customer support. I don't know whether this is a challenge or not, but I guess this is more of an experience or what we are used to. There are a lot of successful internet companies in Asia, especially China. So they tend to focus a lot on the customer experience and the support for their clients. You can see that from a lot of centralized exchanges, Binance, Huobi, OKEx. Customers get used to being served by the exchanges with the questions that they have. However, in DeFi, it's very different, it's very different. People have questions only probably on Telegram or Discord.
There's no customer support team to support you with any question you have. They are not going to call you. If you ever trade on DeFi you know, I will not DM you first. I will not DM first. This is always the case right. So people are not getting the service that they had before on the centralized exchanges. So it may also need time to change alongside. I think right now I can think of these three factors that prevent more massive customers to join.
Vijay (dYdX): Got it, super helpful. Yeah, I think we'll probably slowly see those addressed over time as well. But always interesting to hear that insight. And I think there's a lot more that can be built as well. Great. So I think we're at the hour here, but Yang really appreciate you taking the time to come on especially at 11:00 PM local time. But it's always great to kind of hear your insights and appreciate the explanations that you provided and yeah.
Yang (Bitlink): Yeah, thanks for having me. I enjoyed this talk.
About dYdX
dYdX is the developer of a leading decentralized exchange on a mission to build open, secure, and powerful financial products. dYdX runs on audited smart contracts on Ethereum, which eliminates the need to trust a central exchange while trading. We combine the security and transparency of a decentralized exchange, with the speed and usability of a centralized exchange.