 We want tokens that basically provide useful services. So when we think about blockchains and their native cryptocurrencies, we want this cryptocurrency price to go up if the blockchain is providing useful services. So in other way, we don't want to have cryptocurrency which are essentially generating value or getting values without any underlying fundamental reason. We would like to see cryptocurrencies which are generating value because their underlying blockchain is providing useful services to the users and to the economy. So I'm Augustino Gapponi, an associate professor at Columbia University and the founding director of the Center for Digital Finances and Technologies. To pass the NOWI test, a security needs to derive utility from the effort of others. That's one of the requirements of the NOWI test and that's one of the reasons why, for example, traditional stocks pass the NOWI test because the shareholders extract value from the effort of the managers of the firm. And in the case of crypto, this is what is less clear, like at least if we think about proof of stake, proof of work bitcoins or proof of work cryptocurrencies, it's not so clear that the value that is being generated is being extracted from the effort of others. For this reason, I believe that we cannot really claim that crypto is a security in the traditional sense of the term and what the equivalent of the assistive would look like in my view remains still uncertain. Front running is one of the main risks that is faced in any public blockchain which supports smart contract functionalities and decentralized financial services. Every time you submit transactions through the public blockchain, these transactions while pending, while waiting to be executed are visible to everybody. So it's different from the traditional stock exchanges, where like transactions are only visible at the moment where they're executed. Here in the case of public blockchain, so you can really see the transaction pending and waiting for being executed. So that means that front runners or attackers could screen the pool where all transactions are waiting to be executed and execute a front running attack. For example, if you want to swap one token, say a Bitcoin token for an Ethereum token, then I can be the front runner. I can see that you intend to do the swap transaction and they can place a higher swap transaction ahead of you so that they pump up the price and they would execute at the price that you should have executed. So when it's your turn, you will basically buy one token for the other at the higher price and then I can background you at that point where you have transacted I can sell. So this is basically the front running risk. And this is a risk which basically is leading to a large amount of losses for all those users or arbitrageers who are being front run by these malicious actors who are trying to scan the pool and take advantage of these opportunities. One of the remedies against this front running risk has been exactly the private pool. So private pool means that instead of making the pool publicly available and visible to everybody, it becomes private. You submit your transaction directly to the validators who are going to execute this transaction. So nobody else except for these validators will be able to see the transactions. And one of the rules of the protocol is that these validators who are monitoring this private pool are not supposed to front run you. In this way you get rid of the front running risk because now those screeners, those malicious actors are no longer able to see your transactions. And the question is what is the role of the fee here? The fee is that in the case where you have private pools what could happen is that arbitrageers or like front runners who are looking for transactions to front run, they can still observe this transaction in the public pool and then they can use the private pool just to execute this front running because a front runner could be front run by another front runner. So if basically the front runners go to the private pool they could use this private pool as a way of competing and this can raise the fee even more than what we are observing in the public pool. So that's like why this private pool solution is not yet like a solution which mitigates these negative externalities imposed by the competition between front runners. There is another study that I'm currently working on. We looked at the role of the fees as a mechanism to provide information. So basically what happens is that high fee trade flows typically corresponds to informed trading. That's what we find in our studies that if you are more informed you are going to submit a higher fee trade because you want to trade on this information and you want to do it quickly before you are like exploited before somebody else takes advantage of that same opportunity. So what we find is that high fee trade flows on blockchains are typically informed and also they are more responsive to public shocks. So if there is a shock observed for example based on public news about Bitcoin or another cryptocurrency or like any other like news about the crypto market typically trade flows with high fees are basically more sensitive to this public shocks. So the transparency is a key concern precisely because it generates front running risk. The idea is that because the transactions that are pending on the pool and waiting to be executed are visible to everybody. So that's full transparency. That's exactly the reason why we have front running because I can basically take advantage and try to execute the same transaction as you and raise the price up. And the reason why I can do that is because this transaction are publicly visible to anybody. So every user of the blockchain will be able to see these transactions and they take advantage of this opportunity. So while transparent is good because you can monitor you basically can see everything. You can base your trading strategy on a lot of information. It also has this negative consequence coming from the fact that transactions on the public member pool can be front run. Here we have transparency in blockchain both on the pending transactions and on the executed trade transactions. So even when a trade is executed it's stored on a block and these block are visible to everybody. So if we think about agent welfare in the context of blockchain we can think about those users who could be front run those validators who are executing a service for the user by validating transactions we can think about the front runners those who are exploiting this front running opportunity. So if we think about the welfare of those users or arbitrageers who could be front run then their welfare will go up if there is a private pool because the negative personality coming from front running risk will be mitigated. If we think about validators then if we think about validators who are adopting the private pool then their welfare is also presumably supposed to go up simply because there is higher competition on fees by the front runners and therefore the validators will extract more value. If we think about those front runners these will be the agents who will lose in terms of welfare if a private pool is introduced because no user is submitting truth to the public pool then there will be nothing to exploit. So in this case the welfare will never really go down. One way with in our study we think that this misalignment of incentive can be solved is basically incentivizing every validator to be on the private pool because if every validator is on the private pool then it means that any user will never submit the transaction to the public pool because they will always find validators who will be able to execute this transaction. So the only reason some users may decide not to submit to the private pool is if they don't see a lot of validators monitoring these pools because then they will be subject to execution risk. I can't submit but maybe the validator who gets to append the block is not monitoring the private pool. So that means that my transaction will have to wait more and more blog to get executed. Now once you understand that it's basically clear or it's a result that we show in our study that it's always optimal, like socially optimal if all validators are monitoring the private pool so that front-running risk is completely eliminated. The reason is that why you need a subsidy to these validators who adopt this private pool is that it's not necessarily optimal for validators to all adopt the private pool. The reason is that if they all go in the private pool then there will be nothing to front-run. So all these fees that are being generated through the competition of the front-runners will surely not be that high because the front-runners will not be competing any more with each other because they don't have anything to front-run. So that means that validators will not be better off by all going through the private pool. Now the question is we know that it's socially optimal if they all go through the private pool. So how can we incentivize them to go there to the private pool so that we achieve the socially optimal outcome? And the way to do that would be for those users who could be front-run to pay some fees to these validators. And this fee will essentially be the transfer that will incentivize all these validators to join the private pools. So in a way what this means is that front-runnerable users or front-runnerable arbitrageers, those who are subject to front-running risk are essentially paying for the usage of these private pools. So decentralized exchange is a new type of financial innovation that is essentially built on public blockchains and more specifically your public blockchains which supports mass control functionality. One of the problems of decentralized exchange is that the pricing function that is used to determine the price at which you exchange one crypto token for another crypto token is basically specified by some function which mathematically is known to be convex, which basically means that the marginal price of exchanging one token for the other goes up as the number of tokens that you are exchanging goes up. What happens, what we show also in our study is that the convexity is a very important driver of liquidity provision and of social welfare. The reason is that if you have a very convex curve it means that there is a large cost for trading. So if you trade with a very high convexity it means that you pay a lot for trading. So if you are a user who is deciding whether or not to trade, you might decide not to trade if you see that the curve is very convex because it's too costly for you to trade. On the other hand, if you have a very low convexity then many users will want to trade. But as a liquidity provider you also need to counterbalance this fact that you are gaining from fees paid by the users who are trading with the cost that you might be arbitraged by arbitrageers who are trying to profit from deviation between the price on the decentralized exchange and the price on the centralized exchange. So that means that if it's very cheap to trade both users and those arbitrageers will have an incentive to the non-sac. So while you are gaining from the fees of the users you are losing from the cost of these arbitrageers taking wealth away from you as a liquidity provider. So you don't want the convexity which is too low. You also don't want the convexity which is too high because as I mentioned before if the convexity is very large then trading is very costly. So nobody will trade. The arbitrageer will not exploit you which is good but you will also not get any users who are trading therefore the fees that you are gaining are basically zero and you don't want basically to run a service if you are not profiting from it. And what we show is that the sweet spot in the middle. So you want to have some degree of convexity which is neither too low nor too high just enough to incentivize. Now what happens is that this optimal curvature of the pricing curve from the point of view of the liquidity providers is too high compared with what should be socially optimal. So if you ask like a regulator or a social planner to decide for the optimal convexity which benefits all the participants of this ecosystem including the liquidity providers, users, arbitrageers, validators, then the answer is that you want some convexity but this convexity should be a higher than the convexity that is optimal for the liquidity provider. And there when we think about convexity we should really think about the cost of trading. So higher convexity means higher cost of trading. So that means that the optimal cost of trading from the social planner perspective should be lower than what is the optimal cost of trading from the liquidity provider's perspective. Both trading volume and volatility are basically very important ingredients because if we think about liquidity provision then if you have a very volatile token pair so if you're like exchanging tokens which are very volatile it basically means that there is a lot of arbitrage opportunities to be exploited. So more volatility means more opportunities for the arbitrageers which basically means more wealth is being transferred from the liquidity providers to the arbitrageers which means that they're incentive to provide liquidity from the liquidity provider's perspective goes down with the volatility of the token pairs. On the other hand, if you have a higher trading volume which many users are willing to trade because for private reasons they need to trade and get access to specific tokens. For example, some users may have to access the Ethereum blockchain. They want to swap and get the Ethereum token because they want to use these tokens to execute the app on the blockchain. So then in that case, it's good for the liquidity providers because these users are going to pay fees because trading volumes means trading fees because the fee is proportional to the volume that is being exchanged. Higer volatility means less liquidity provision or reduces the incentive to provide liquidity. Higer trading volume increases the incentive to provide liquidity. We want tokens that basically provide useful services. So when we think about blockchains and their native cryptocurrencies, we want this cryptocurrency price to go up if the blockchain is providing useful services. So in other way, we don't want like to have cryptocurrency which are essentially generating value or getting values without any underlying fundamental reason. We would like to see cryptocurrencies which are generating value because that underlying blockchain is providing useful services to the users and to the economy.