<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by DC on Medium]]></title>
        <description><![CDATA[Stories by DC on Medium]]></description>
        <link>https://medium.com/@analystDC?source=rss-91a15c91277d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*5OSqv5Due3D6P2U6tXsNUg.jpeg</url>
            <title>Stories by DC on Medium</title>
            <link>https://medium.com/@analystDC?source=rss-91a15c91277d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 23:59:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@analystDC/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Impact Measurement]]></title>
            <link>https://medium.com/gauntlet-networks/impact-measurement-9cee8fdcf47d?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/9cee8fdcf47d</guid>
            <category><![CDATA[defi]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Wed, 07 Sep 2022 13:50:23 GMT</pubDate>
            <atom:updated>2022-09-07T13:50:23.938Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part Four: Compound Case Study</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4t8GlkspNdWeOdNZh5Sg6g.png" /></figure><h3>Summary</h3><p>Gauntlet has executed 13 proposals during the most recent Compound engagement, the first of which was executed on 11/09/2021, the last of which was executed on 07/04/2022. Gauntlet increased Collateral Factors (CFs) in every proposal, except for one time decreasing the CFs of BAT and ZRX on 11/27/2021, which were later increased.</p><p>Below is the time series of collateral factors for each asset from 11/01/2021 (8 days before the engagement), until 08/15/2022.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4ngCTmrneH6HqK9OJyuGzA.png" /></figure><p>The resulting proposals resulted in net CF increases for the majority of assets, all while incurring no insolvencies despite significant market volatility, as outlined in our <a href="https://www.comp.xyz/t/compound-market-downturn-risk-review-may-2022/3251">Market Downturn Analysis</a> posts. In each of our forum post proposals we include screenshots of the projected increase in VaR and LaR (proxies for risk) as shown below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RWahr69ClRCDWS_48shiMg.png" /></figure><p>Since we do not expect our recommendations to increase protocol risk, these collateral factor changes proceeding without insolvency was a successful result for our product. But what value did Gauntlet add to the protocol as a result of this? Besides managing risk, we seek to confirm that our recommendations generated quantifiable increases in usage and revenue, which will be the main topic of this case study. Using the methodology detailed below, we found that Gauntlet’s impact totaled $96M in added borrows and $5.15M in added borrow interest over the course of our engagement with Compound.</p><h3>Expected Impact</h3><p>When collateral factors increase, we project users to take advantage of some of their additional borrowing power. Thus, we can think of expected impact as the increased usage that we estimated at the time the change was made. As changes go live, we can then calculate how much of this additional usage is realized, which depends on market conditions and user behavior. To make this more concrete, let’s review the formal definition of borrow usage, which we introduced previously <a href="https://www.comp.xyz/t/dynamic-risk-parameters/2223">here</a>,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/824/1*86wnHsMdDS04NQA_KlA8hA.png" /></figure><p>Essentially, for a given asset, the borrow usage is the average percentage of borrowing power used by each user of the asset, weighted by their share of total supply in that asset. The system borrow usage is the average of each symbol’s borrow usage, weighted by the protocol percentage supply. It can be a little hard to intuit however, so let’s take a look at an example.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QhDcZZ8gRbJxqOobLzYhSQ.png" /></figure><p>Imagine the total supply of ZRX in Compound is $60M, and it is composed of two users who supply $40M and $20M ZRX, respectively. The current CF of ZRX is 65%, and we want to know what the increase in borrow usage will be if we increase the CF to 70%.</p><p>Taking a look at the first user, we see he has supplied $40M ZRX and no other assets. As a result, this user has $40M * 65% = $26M borrowing power, and has decided to borrow $13M USDC, so this user’s initial borrow usage was $13M / $26M = 50%. With a CF of 70%, the user now has $40M * 70% = $28M borrowing power. In order to keep the borrow usage consistent, this user will borrow $28M * 50% = $14M USDC. As a result of the change, we would see a $1M / $13M = 7.7% increase in total borrows and borrow usage.</p><p>The second user has $20M ZRX supplied, but also $60M WBTC supplied. WBTC has a CF of 70%. As a result, this user has $20M * 65% + $40M * 70% = $55M borrowing power, and has decided to borrow $40M USDC, so this user’s initial borrow usage was $40M/$55M = 72.73%. With a ZRX CF of 70%, the user now has $20M * 70% + $40M * 70% = $56M borrowing power. Note that the increase in borrowing power is relatively smaller compared to the entire supply position, because ZRX only makes up 25% of the supply for this user. In order to keep the borrow usage consistent, this user will borrow $56M * 72.73% = $40.73M USDC. Proceeding similarly for this user, we would see a $0.73M / $40M = 1.8% increase in total borrows and borrow usage.</p><p>Essentially, we see that the increase in borrows for each user is the ratio of the old and new collateral factors (70% / 65% ) -1 = 7.7%, multiplied by the percentage of the user’s supply which is ZRX. So if the suppliers of a given symbol have even more supplies of other assets, we do not expect to see large borrow usage changes as a result of that symbol.</p><p>The assumption that borrowers will increase their borrows to take full advantage of their increased borrowing power as a result of our parameter recommendations is what we refer to as <strong>elasticity</strong>, which we will assess later. However, assuming that users are indeed elastic, we can obtain our projected borrow usage increases and an impact dollar value.</p><h3>Impact Metrics: Borrow vs Interest</h3><p>Referencing the Borrow Usage visual above again, if we were to increase the CF of ZRX, we see that there is a projected dollar increase in total borrows for each of the users that supply ZRX. Given that about 94% of all borrows in Compound are stablecoins, this usually results in an increase in stablecoin borrows. However, we may care more about the increase in borrow interest of these borrowed assets, since that’s extra income the protocol can decide to either allocate to suppliers (to incentivize more TVL) or to the reserves (to be able to cover insolvencies or other unexpected events). Therefore, we will include the projected increase in borrows in our numbers, but use projected increase in interest as the primary metric for Gauntlet’s impact.</p><p>The way to calculate the annual projected increase in borrows for a given CF change is to iterate over every user that supplies assets, calculate the projected dollar increase in borrows for that user, and multiply that increase in borrows by the borrow rate for each borrowed asset. So if a CF increase in ETH results in $25M projected increase in USDC borrows relative to the previous CF, and the borrow rate for USDC is 4%, this would result in a projected USDC-specific annual impact of $25M * 4% = $1M. Note that in reality an increase in USDC borrows would increase the USDC utilization, thus increasing the borrow interest and even more projected impact, but for simplicity we can assume the borrow interest is static.</p><p>Gauntlet’s projected annualized increase in borrow interest at the time of our CF changes was $24M total. However, projected annualized impact assumes that total borrow amount and borrow rates remain constant throughout the year, which of course will not happen. Even though our projected impact was $24M, we may actually see more or less depending on realized market conditions. Since we simulate a range of outcomes and do not attempt to predict the overall direction of markets, our pre-recommendation estimates assume market prices and interest rates remain unchanged throughout the impact period.</p><h3>Calculating Realized Impact</h3><p>To calculate our realized impact, we must:</p><ol><li>Account for the fact that realized borrows and borrow rates change over time</li><li>Verify our elasticity assumptions</li></ol><p>Incorporating the realized borrows and borrow rates is relatively straightforward to do, given we have daily borrower data and can calculate interest from the known borrow rate on that day. Verifying the elasticity assumption is more complex. To better illustrate, let’s look at the following example:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3Ta8KfvCEcbPDS6han0utA.png" /></figure><p>Imagine there is a user who on 4/20 supplies $100 of DAI, which has a CF of 60%, and borrows $30 of DAI. Assume DAI CF increases from 60% to 80% on 4/21, but the user’s TVL and borrows remains the same. Now we’re left with the question of, “what was our realized impact of this user on this date?” There are a couple ways we could think about this.</p><h4>Constant Usage View</h4><p>We could see this case as showing zero realized impact, since the user did not increase their borrows to take advantage of their increased borrowing power. Essentially, assuming that the user’s risk profile was established on 4/20, this user was not elastic. It is tough, however, to compute long term impact in this framework, since users’ risk profiles change, and it is not always clear which changes in borrow usage are due to updated risk profiles as opposed to elasticity. As we will later discuss, this assumption is appropriate for some users, but in general we use a more flexible elasticity assumption described below.</p><h4>Dynamic Usage View</h4><p>We could also say the user understood that they had increased borrowing power on 4/21 and changed their risk profile purposely such that their new comfortable borrow usage was 37.5% (due to market volatility or whatever other reasons). We can then posit that if the DAI CF hadn’t increased, the user actually would have <strong>decreased</strong> their borrows to maintain their new baseline 37.5% borrow usage risk profile.</p><p>This framework allows us to more cleanly calculate impact, since we can update the user’s risk profile with each data point as we get more info on their behavior. This avoids the issue of having to choose a time in the past to use as the benchmark for risk tolerance, which is a problem for volatile assets. Of course, we still have to make sure usage is stable in the long run to validate elasticity, but this is much more realistic than expecting it would be constant on a daily timescale for most users.</p><h3>Observing User Elasticity</h3><p>From the real data, it may or may not be possible to identify an appropriate constant risk profile for every user. In our analysis, we use both frameworks described depending on how we classify an account. The two types we use here are:</p><ol><li>Recursive: These are users who have exactly one symbol with a positive supply and borrow. For example, a user who has $1M DAI supplied and $500k DAI borrowed is a recursive account. Since recursive users restrict their activity to one symbol, their usage is less sensitive to relative price fluctuations. For these accounts, we found it appropriate to assume constant usage as they are observed to fit this behavior quite accurately.</li><li>Non-Recursive: These are all other users who do not exclusively borrow and supply only one symbol. Though accounts in this group may have some overlapping borrow and supply symbols (i.e. partially recursive users), their exposure is not insulated from market volatility like that of a purely recursive user. We thus consider it more appropriate to apply the dynamic usage framework for this entire group of accounts.</li></ol><p>Performing the calculations for every user and every date during the most recent engagement starting 11/9/2021 until 08/15/2022, we found that the realized interest impact given the above reasoning is <strong>$5.15M. </strong>Now let’s go forth and analyze the elasticity assumption.</p><p>The realized impact we observed is broken down by borrower type as follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rxQN4R4y2m63Tv0Jb8_dgw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D-x4HVFwbvzIRqGtBL3vUw.png" /></figure><p>In the above chart, the inner pie sections correspond to the user impact density. For example, in the Non Recursive section, the blue inner pie piece corresponds to the impact came from the largest impact Non Recursive user, the orange piece corresponds to the impact from the second largest impact user, etc., and the brown piece at the bottom aggregates the impact across the remainder of users. Let’s zoom in on the impact of the largest individual borrowers, during which we will determine if our elasticity assumption holds up.</p><h3>Recursive User Examples</h3><p>The total impact for recursive borrowers was <strong>$1.3M</strong>, broken down as follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WQSWRbBfRmwXQKwnZj-vBw.png" /></figure><p>The above pie chart shows that the top 9 accounts account for over 85% of the total recursive impact. Let’s analyze some of the top users descending order of impact to see if the impact is justified.</p><h4>Recursive User 1</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/986/1*yPzL6x1l0crLayBZJC1k_g.png" /><figcaption><strong>Address: 0x1676055fe954ee6fc388f9096210e5ebe0a9070c</strong></figcaption></figure><p>In the above charts we see supply, borrow, and usage time series for the highest impact recursive user, a large USDC account. On 11/27/2021, we increased CFs from 75% to 80%, at which point the user’s borrow usage decreased. However, a few days later, the user updated their position to take full advantage of the increased borrowing power, thus showing the user was elastic. Eventually the user decided to decrease their borrow usage in April 2022 with no collateral factor change. There were 2 more relevant proposals afterwards, when the user had lower TVL. After the second proposal it seemed the user did not update their position to keep their previous borrow usage, but after the third proposal the user increased their borrow usage even higher than it previously was, just to decrease it again later. As we know, it isn’t always clear to distinguish between the user updating their position to be elastic and return to a previous baseline borrow usage, or when a user has determined to move to a new baseline borrow usage. But looking at this user’s behavior, we should be able to at least claim impact through April 2022 based on the constant usage framework, which we do for this account.</p><h4>Recursive User 2</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*mG31RXos61-_U0RrjsHWgQ.png" /><figcaption><strong>Address: 0x7900c70a377f89df29d1d1939469ae3b74c5b740</strong></figcaption></figure><p>In these charts we show the same data for the next largest recursive account, a recursive DAI borrower. On 11/09/2021, we increased CFs for DAI from 75% to 80%, at which point the user’s borrow usage decreased. However, a few weeks later, the user updated their position to take full advantage of the increased borrowing power, thus showing the user was elastic. Though we could claim impact immediately after the proposal by positing that the user’s risk profile may have also changed, from the chart it seems fairly unlikely that is the case. As a result, for this user we will take the baseline borrow usage to be what it was the date prior to the first proposal, and only claim impact with the constant usage framework described above.</p><h3>Non-Recursive User Examples</h3><p>Moving on to non-recursive borrowers where our impact assuming elasticity was <strong>$3.85M</strong>, the pie chart shows that the top 9 accounts make up about 50% of the total impact here.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xTzshrYRB88aqxeotaVZ3w.png" /></figure><p>Let’s again look at some the largest users to see if the impact is justified.</p><h4>Non-Recursive User 1</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/828/1*B0AyHC4wagXS7XUI_K8jiw.png" /><figcaption><strong>Address: 0x8aceab8167c80cb8b3de7fa6228b889bb1130ee8</strong></figcaption></figure><p>The above charts shows the largest non-recursive user’s borrow usage over time. Dots are weighted by how much impact they drove on that particular day. The vertical green lines indicate the dates of relevant proposal changes for this user (where CFs of symbols this user supplied were increased). The red dots indicate user’s borrow usage on the date prior to the proposal change. The blue horizontal line is the weighted borrow usage across all dates. As we can see, the user tends to hover around a borrow usage in the mid-50% range throughout the course of their interaction with the protocol, sometimes letting their borrow usage increase to roughly 70%, and decrease to roughly 40%. There doesn’t seem to be any clear pattern of decreased borrow usage over time, which justifies this user as elastic. The results for other non-recursive users are very similar, with some more examples shown below. As a result, we will claim <strong>$3.85M</strong> as realized impact for non-recursive borrowers under the dynamic usage framework.</p><h4>Non-Recursive User 2</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/864/1*dd-g4tta0f0LqBr5_wTDPA.png" /><figcaption><strong>Address: 0x10d88638be3c26f3a47d861b8b5641508501035d</strong></figcaption></figure><h4>Non-Recursive User 3</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/936/1*8BRJ4Hjl8JEV-mBUeuTMzg.png" /><figcaption><strong>Address: 0x99fd1378ca799ed6772fe7bcdc9b30b389518962</strong></figcaption></figure><h3>Additional Impact on TVL and Future Interest</h3><p>Another important concept we haven’t touched on — because of how difficult it is to quantify — is how many new borrowers joined Compound as a result of our capital efficiency improvements, how many borrowers decided not to leave Compound as a result of our capital efficiency improvements, and how many existing borrowers decided to increase their borrows / decrease their borrows less as a result of our capital efficiency improvements. During the market crashes, and after the proposal which cut COMP rewards, many users did decide to leave the protocol. It’s very likely capital efficiency improvements kept certain users from leaving and seeking better yield elsewhere. As a basic example, let’s look at the TVL time series of SUSHI, over a chart of our CF recommendations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qLPxJREe0poHr0pXPl0M1A.png" /></figure><p>We see that the TVL for SUSHI (due to one SUSHI whale) increased from roughly $0 to $8M shortly after a set of proposals which increased SUSHI CF from 40% in November to 70% in mid April were executed. It’s probably safe to say that this SUSHI whale would not have entered the protocol if not for the drastic increase in capital efficiency. Yet we claim very little SUSHI impact because our calculations assume TVL would be the same regardless of our recommendations. There are similar examples with other symbols that saw TVL rise materially at times, such as MKR, LINK, and AAVE.</p><p>Finally, the impact we have claimed so far is only from the realized borrow interest that has already occurred due to our increased CFs. Assuming CFs remain the same, the protocol will continue to benefit from higher interest going forward, which we do not include in this case study. Since these more indirect forms of impact would likely be material, we see the calculation here as a conservative estimate of the most quantifiable value we added. As we refine our methods, we may be able to better measure other types of impact in future case studies.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9cee8fdcf47d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/impact-measurement-9cee8fdcf47d">Impact Measurement</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Impact Measurement]]></title>
            <link>https://medium.com/gauntlet-networks/impact-measurement-63f1b60d07ef?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/63f1b60d07ef</guid>
            <category><![CDATA[defi]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Mon, 29 Aug 2022 16:19:24 GMT</pubDate>
            <atom:updated>2022-09-07T14:27:04.319Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part Three: Generalizing Impact Measurement</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c1N9SuPXRVqANGH5aKuBMw.png" /></figure><p>In a <a href="https://medium.com/gauntlet-networks/impact-measurement-20d9d820fcba">previous article</a> on Impact Measurement, we introduced the case of recursive accounts as a simplification of the general problem of quantifying the impact of our recommendations. Since we observe empirically that recursive accounts tend to maintain their borrowing at a constant level relative to the amount available, it is possible to extend changes in available borrow capacity to an estimate for impact using relatively simple math. In the broader case of non-recursive accounts, we must add some more steps to account for the fact that utilization may change along with parameter updates. In this article, we will look more closely at the behavior of lending protocol users in response to changes, and show some results from the overall impact analysis for our Compound recommendations.</p><h4>Borrower Behavior</h4><p>When borrowing limits are increased, the usage of all borrowers across the protocol relative to the limit immediately drops due to the now-larger denominator. We would then expect users to borrow more over time as they adjust for the effects of the change. In the case of recursive borrowers, this happens very quickly and we can simplify by modeling these users as effectively having constant utilization percentage.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-ZNjywYrdE5KapraAwDohg.png" /></figure><p>In the general case, utilization is affected by market prices and users’ day-to-day transactions, so it may be difficult to pinpoint a clear reaction function like the one shown in the stylized example above. While users on average may adjust to a higher limit, the effect on any specific account is difficult to isolate. For this reason, our methodology for non-recursive impact focuses on determining whether utilization is broadly stable in response to parameter changes. If we can confirm the market is adapting to higher borrowing limits without persistently lower utilization, we know that changes are having the desired effect and can estimate the impact.</p><p>Looking at data from one of the larger accounts on Compound, we note that utilization has fluctuated between 35% and 70% during the course of several borrow limit increases marked by the vertical lines. Over the period observed, this account has also varied considerably in total size, as shown by the weight of the blue dots on the chart. Though the fraction of available borrow used has been volatile, the range it varies in has stayed consistent throughout all of this. As we will show in the next article, many users fit this pattern of significant short-term variability in utilization within a steady range.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8RVG92PCcOjUtPNEgZvOBw.png" /></figure><p>In a market that generally follows this behavior, we can validate that non-recursive users are indeed adapting to higher borrowing limits efficiently. For example, if this user borrows 55% of their limit on average, an increase in the limit would grow the user’s borrow position proportionally as they adjust to the change. Though we cannot confirm this broadly holds true by looking at just a single account, the data observed supports this type of reasoning. Our methodology uses these observations conservatively when assessing impact, as we seek to verify users fit our expectations before including their contributions in the total. Now that we have set the stage by looking at the theory behind impact measurement and the details of borrower behavior, we can begin to assemble the complete impact calculation for lending risk management.</p><h4>Non-Recursive Impact</h4><p>In our impact analysis of Compound, we have identified several non-recursive accounts that follow the patterns described above. By estimating their equilibrium usage levels, we have been able to attribute material increases to interest earned from these accounts to our parameter recommendations. As shown in the diagram, these users have contributed an additional $3.85M of interest to Compound suppliers and reserves since November 2021 due to higher borrowing limits:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iNm7M6U9pFW3zXv2mbSZow.png" /></figure><p>The next article on impact measurement will combine all the ideas discussed so far to provide a full view of our Compound analysis. By sharing a case study, we look to show that the theory introduced in this series is robust and applicable to complex real-world markets. Finally, with the results, we seek to demonstrate that our model’s optimizations drive quantifiable impact.</p><p><a href="https://medium.com/gauntlet-networks/impact-measurement-9cee8fdcf47d"><em>Continued in Part Four: Compound Case Study</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=63f1b60d07ef" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/impact-measurement-63f1b60d07ef">Impact Measurement</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Gauntlet’s View on The Merge]]></title>
            <link>https://medium.com/gauntlet-networks/gauntlets-view-on-the-merge-7c79e39e4b59?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/7c79e39e4b59</guid>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[ethereum]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Wed, 24 Aug 2022 21:11:51 GMT</pubDate>
            <atom:updated>2022-08-24T21:11:51.304Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5CY7UyN0SjrlZBi5hjOhAA.png" /></figure><p>Originally posted via <a href="https://coinmarketcap.com/alexandria/article/gauntlet-s-view-on-the-merge">CMC Research</a></p><h3>Overview</h3><p>The highly anticipated <a href="https://ethereum.org/en/upgrades/merge/">Merge</a> is scheduled for <a href="https://hackmd.io/@benjaminion/BkxQTpqpi9">September 15</a>, transitioning the Ethereum network from proof-of-work (PoW) to proof-of-stake (PoS). While the Ethereum community and developers are broadly supportive of this network upgrade, the upgrade significantly impacts the existing mining ecosystem by removing the <a href="https://whattomine.com/">most profitable coin to mine</a> from PoW, thereby reducing the utility of energy-intensive mining hardware. As a result, prominent ETH miner Chandler Guo has advocated creating a fork of Ethereum with its current PoW properties. Although Ethereum <a href="https://www.gemini.com/cryptopedia/ethereum-classic-etc-vs-eth">hard forks</a> have occurred, this particular fork has created some uncertainty in the Ethereum ecosystem, where more decentralized applications (DeFi, NFT, and GameFi) are running today. In this post, we share our thoughts on the impact of the potential hard fork. In short, we are supportive of the Merge and believe it will provide long-term value to all stakeholders of the Ethereum ecosystem.</p><h3>Support for PoS Reflected In Tokens</h3><p>Conceptually, the state of the Ethereum virtual machine on both the PoS chain and PoW chain will be identical after the Merge, so users who hold ETH prior to the fork will have equal amounts of ETH on both the PoS chain and PoW chain (where ETH is called ETHW). The relative values of ETH and ETHW give us an idea of the support for each token, and exchanges like Poloniex currently enable users to trade ETHW IOU, a proxy for the impending ETHW. At the time of writing (8/16), ETHW IOU is trading at 3.5% of ETH’s price, implying that the PoS chain has wider support than the PoW chain.</p><p>Further, as most issuers of fiat-backed stablecoins (<a href="https://www.circle.com/blog/usdc-and-ethereums-upcoming-merge">USDC</a>, <a href="https://tether.to/en/usdt-supports-eth-proof-of-stake-transition/">USDT</a>) and custodial wrapped coins (WBTC) have expressed their support for the PoS chain, those tokens will have zero value on the PoW chain post-merge. Also, cross-chain bridges are not likely to support the PoW chain and thus owners of bridged tokens on L2 or EVM chains will only be able to withdraw them back to the PoS chain post-merge.</p><h3>Expected User Behavior</h3><p>Assuming the Merge is successful (which is likely given PoS has been activated on <a href="https://hackmd.io/@benjaminion/HkPEOKGA5#Goerli-Merge-review">the Goerli testnet</a>), users who believe ETHW will have traction and accrue non-trivial value will attempt to “farm” ETHW by accumulating ETH in advance of the Merge; their behaviors may introduce some transient risks to the ecosystem. The users’ potential actions include:</p><ul><li>Pre-Merge — maximizing ETH holdings by one or more of the following:</li><li>Simply holding ETH</li><li>Withdrawing ETH from centralized exchanges to the mainnet</li><li>Swapping tokens to ETH</li><li>Borrowing ETH or withdrawing ETH collateral from lending protocols</li><li>Withdrawing ETH-paired LP tokens from DEXs</li><li>Withdrawing bridged ETH back to Ethereum mainnet</li><li>Post-Merge — immediately transferring and selling ETHW to centralized exchanges, and unwinding their pre-merge actions involving ETH</li></ul><p>Note that these strategies around the Merge may not be “free money” due to risk from potential front-running bots, token price volatility spikes, interest rate spikes in lending protocols, and more broadly, opportunity costs. Proper execution of these strategies requires technical sophistication.</p><p>An interesting note is that DAOs (e.g., <a href="https://gov.frax.finance/t/fip-101-make-frax-stablecoins-only-redeemable-on-pos-eth/1753">FRAX</a>) are also planning to sell any significant value on the PoW chain and transfer the proceeds to the treasury.</p><h3>Risks To Lending Protocols</h3><p>Oracles are a critical component of the DeFi ecosystem as they enable protocols to receive data from outside the blockchain. Chainlink, one of the primary oracle networks serving the price feeds for major lending protocols, is<a href="https://docs.chain.link/docs/ethereum-proof-of-stake-merge/"> committed to supporting the PoS chain</a>. Lending protocols on the PoS chain should operate as expected if the oracle functions correctly. In contrast, lending protocols on the PoW chain may not function properly due to lacking oracle support.</p><p>We expect heightened asset volatility during the Merge, which can potentially trigger liquidations for users with insufficient levels of collateralization. Furthermore, there could be a liquidity crunch on the PoW chain post-merge if the lending protocol is only partially functioning, as we expect that users may squeeze the remaining value from what’s left on the PoW chain by withdrawing or borrowing the available ETHW from the lending protocols.</p><h3>stETH’s Activity Pre-Merge</h3><p>stETH is a staking derivative representing a claim to the staked ETH managed by <a href="https://lido.fi/">Lido</a>. stETH holders can redeem stETH for ETH after the <a href="https://ethereum.org/en/upgrades/merge/">Shanghai upgrade</a>, which will happen at least 6~12 months after the Merge. Factoring in the time value of money, stETH should trade at a discount to ETH. As mentioned above, the potential PoW fork provides ETH holders with an equivalent amount of ETHW and thus may attract some stETH holders to sell stETH for ETH to get access to ETHW after the Merge. This should further push down the stETH / ETH price. Thus, a valuation framework for stETH is:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dONbgWL6oTSdkZEU7BcGuQ.png" /></figure><p>The potential for the PoW fork introduces the term</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vdPtp7sk7KFhr9SD7pxkiA.png" /></figure><p>into the price deviation, and thus the expected ETHW price is the main driver of the incremental price deviation of stETH. However, one potential offset is the fact that a successful network upgrade also reduces the uncertainty of the stETH withdrawal timeline, thereby affecting the time value of holding stETH. stETH holders should evaluate all these factors when considering a swap for ETH.</p><p>As rational users may borrow ETH or withdraw ETH collateral from lending protocols during the Merge, the resulting ETH interest rate spikes might also affect the profitability of stETH / ETH recursive borrowing on lending protocols, but the spikes should only have a short-term impact on users.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_cQdVpcCr2OvDJ7FVgWqnQ.png" /><figcaption>Curve stETH-ETH pool (Reference: <a href="https://dune.com/queries/374112/712351">https://dune.com/queries/374112/712351</a>)</figcaption></figure><p>Finally, Curve’s stETH-ETH pool is the largest trading venue for stETH. stETH’s price could come under pressure (depending on the expected value of ETHW) as stETH-ETH liquidity providers withdraw their liquidity in preparation for farming ETHW post-merge. Users with stETH as collateral on the lending market could be at risk of liquidation. That said, long-term believers in Ethereum could take this opportunity to acquire discounted stETH.</p><h3>Conclusion</h3><p>Gauntlet is supportive of the Merge and firmly believes this is a step toward a sustainable, energy-efficient, and scalable Ethereum ecosystem. We recognize that this creates some uncertainties in the short term, so we will actively work with different communities, closely monitor the market condition, and propose additional preventive measures to different protocols if needed.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7c79e39e4b59" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/gauntlets-view-on-the-merge-7c79e39e4b59">Gauntlet’s View on The Merge</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Impact Measurement]]></title>
            <link>https://medium.com/gauntlet-networks/impact-measurement-20d9d820fcba?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/20d9d820fcba</guid>
            <category><![CDATA[defi]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Mon, 15 Aug 2022 15:07:44 GMT</pubDate>
            <atom:updated>2022-08-29T17:48:32.402Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part Two: The case of recursive borrowers</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*As1dKP6JlZgGwVD-Uj3yQg.png" /></figure><p>To follow up on our introduction to <a href="https://medium.com/gauntlet-networks/impact-measurement-78315cafc9c7">Impact Measurement</a>, this article will look in more detail at the practice of recursive borrowing and how it affects our calculations. To start, let’s define recursive borrowing as a user supplying assets that were borrowed from a lending protocol back to the same protocol. Since it would be logical that interest paid to suppliers is lower than the interest cost of borrowing, it may be a bit counterintuitive why users would want to borrow recursively in the first place. However, sometimes opportunities do arise from details in protocol reward mechanics and sophisticated users can use automated strategies to locate and take advantage of them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vQ9G2zuXNvK0KVkCw5H_XA.png" /></figure><p>Due to recursive borrowing being effectively a form of arbitrage, it is favorable for users of this strategy to borrow close to the allowed limit in order to maximize the rewards earned. This drives recursive accounts to react quickly when borrowing limits are changed — a key observation for our impact analysis.</p><p><strong>Recursive Borrower Behavior</strong></p><p>To calculate our impact on recursive accounts, we must first verify that they behave as expected when risk parameters are adjusted. If borrowing limits are in the primary constraint, recursive strategies, we would expect these accounts to top-up their exposure whenever limits are increased, and reduce exposure when limits are lowered.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hCoSXnFvvFLOWhd7P75LNw.png" /></figure><p>Looking at a typical account that we have identified as a recursive borrower on Compound, we note that as a share of total available borrows their utilization remained very consistent during our parameter adjustment in late-2021 (<a href="https://compound.finance/governance/proposals/66">Proposal 66</a>) and afterward. While utilization from recursive users has dropped off more recently, this is probably more associated with changes in Compound reward mechanisms that have made their strategies less profitable (<a href="https://compound.finance/governance/proposals/92">Proposal 92</a>). During the period between these two proposals, our methodology holds that the impact of our recommendation on these accounts is directly proportional to the increase in borrowing power available, as the users fully recalibrated to the new limits. By comparing to the base case where limits and utilization were unchanged from the start, we can estimate the total impact on recursive accounts using relatively simple algebra. For example, if available borrows increased by 5% after a parameter change, we attribute 5% of recursive user activity after the change to our impact. As we will discuss in a later follow-up, the general (non-recursive) case of impact measurement requires more nuance, but we make this simplification for the specific period that recursive accounts kept their utilization almost constant.</p><p><strong>Recursive Borrower Impact</strong></p><p>Using what we know about borrower behavior, we can treat recursive users separately in the impact calculation to obtain a fairly clear view of their contribution to key protocol metrics. Since our experience with these users so far is mainly from the period following risk-on recommendations, our analysis quantifies the impact as an increase in total borrow usage, total interest paid, and total protocol reserves over the relevant period. The example view shows some results from our analysis of Compound between November 2021 and June 2022, when we believe the majority of our impact on recursive users occurred there. With the effects of our parameter recommendations, we estimate recursive users generated an additional $1.3M of borrow interest for the protocol, broken down by major accounts as shown below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sC6ggYPw_p3_JoOKG8tlUg.png" /></figure><p>As some opportunities have been exhausted and market volatility has made the recursive strategies riskier, we recognize that our impact on non-recursive borrowers may be a more dominant part of the calculation in the future. Due to their relatively straightforward behavior, however, we see recursive borrowers as a simple case to introduce the concept of impact measurement without the added complexity of analyzing shifts in utilization. In later articles we will generalize the ideas discussed here to the broad non-recursive user base and share more data on the total impact we found across all account types.</p><p><a href="https://medium.com/gauntlet-networks/impact-measurement-63f1b60d07ef"><em>Continued in Part Three: Generalizing Impact Measurement</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=20d9d820fcba" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/impact-measurement-20d9d820fcba">Impact Measurement</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Ref Finance Onboards Gauntlet for Budget and Incentive Optimization]]></title>
            <link>https://medium.com/gauntlet-networks/ref-finance-onboards-gauntlet-for-budget-and-incentive-optimization-bc1b35a28c67?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/bc1b35a28c67</guid>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[crypto]]></category>
            <category><![CDATA[incentives]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Wed, 10 Aug 2022 16:02:13 GMT</pubDate>
            <atom:updated>2022-08-10T16:02:13.281Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yaI0y692zzfuA3PR3GPsYA.jpeg" /></figure><p><a href="https://gauntlet.network/">Gauntlet</a> is excited to announce our partnership with <a href="https://www.ref.finance/">Ref Finance</a>. Ref Finance is a leading decentralized exchange (DEX) on NEAR focused on offering a fast and user-friendly trading experience. The Ref Finance team recognizes the importance of ensuring its LP rewards are budgeted and allocated efficiently to maximize volume and revenue back to the protocol and token holders. As part of this partnership, Gauntlet’s Incentive Optimization (IO) platform will ingest the Ref Finance protocol data to manage rewards quantitatively.</p><p><strong>Modeling Incentive Spend</strong></p><p>Incentives within DeFi are an essential component to bootstrapping liquidity and growth. Successful incentive programs create a virtuous cycle of attracting users, users generating fees, and fees supporting future growth efforts. It is important to understand the key drivers of efficiency, such as which liquidity pools should receive more incentives than others and how incentives should change over time in response to market conditions. Additionally, the total level of incentive spend needs to be set to ensure alignment with overall protocol budgets. Accounting for factors like partner pools and other strategic initiatives is required. The Gauntlet IO platform incorporates all these specifics and a broad range of market variables when determining optimal rewards to maximize spend efficiency.</p><p><strong>Improving Spend Efficiency</strong></p><p>Ref Finance and many similar protocols face the difficult challenge of predicting the effects of parameter changes to maintain efficient spending. Our platform simulates realistic market behavior to predict the impact of changes across all pools. To provide dynamic incentive optimization for Ref Finance, Gauntlet will use its incentive models to process Ref Finance protocol data and regularly recommend emission parameters for native, partner, and strategic pools.</p><p>The Ref community can find members of the Gauntlet team in the governance forum to deliver, communicate, and clarify parameter changes to facilitate understanding of the model outputs.</p><p>If you’re interested in learning more about Gauntlet’s IO platform, you can explore the platform and get in touch with our team <a href="https://gauntlet.network/platform/">here.</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc1b35a28c67" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/ref-finance-onboards-gauntlet-for-budget-and-incentive-optimization-bc1b35a28c67">Ref Finance Onboards Gauntlet for Budget and Incentive Optimization</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Impact Measurement]]></title>
            <link>https://medium.com/gauntlet-networks/impact-measurement-78315cafc9c7?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/78315cafc9c7</guid>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[crypto]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[risk-management]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Thu, 04 Aug 2022 16:25:59 GMT</pubDate>
            <atom:updated>2022-08-26T17:32:45.879Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part One: How we think about measuring impact</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Y_PIyVtsWjd6CG5lczOrBw.jpeg" /></figure><p>At Gauntlet, we aim to provide clients with the most optimal recommendations for their protocol based on simulations of a wide range of possible conditions and outcomes. As part of this process, it is important for us to also verify that the recommendations we are providing are having the desired effect. In this article and later follow-ups, we will share some of our work on Impact Measurement, where we seek to create a quantifiable track record to evaluate the performance of our products.</p><p><strong>Risk Management Goals</strong></p><p>To introduce how we think about risk management, the goals of our recommendations are centered around two variables: Risk and Capital Efficiency. Since improving along one of these dimensions is typically accompanied by sacrificing some of the other, our recommendations usually involve an inherent trade-off. To see why this is the case, let’s consider the basic functions of a lending protocol and how these relationships emerge.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vQrwpoOpNJUlic_7KHEe0A.png" /></figure><p>The purpose of a lending protocol is to match suppliers who are willing to lend a certain asset with borrowers who wish to take the other side of the trade and provide sufficient collateral. The collateral guarantees that the borrower will repay the loan since it can be sold quickly at a discount should the margin of safety fall below a threshold.</p><p>By setting this collateral factor and discount prudently, protocols can ensure loans are resolved in an orderly manner without incurring excessive losses, but setting them too conservatively also impairs the functionality to users. Since high collateral requirements directly limit how much users can borrow, it reduces the capital efficiency of the protocol by hurting utilization and revenue. Our philosophy in designing the risk management product is to <strong>achieve maximum capital efficiency given an acceptable level of market risk. </strong>In the context of measuring impact, this translates into two distinct regimes in which our recommendations have different aims. During a risk-on regime (e.g. when market conditions are favorable), we are focusing more on the capital efficiency side because risk is well within the tolerable range. Conversely, in the risk-off regime we are sacrificing some capital efficiency to reduce protocol risk back to an acceptable level.</p><p><strong>Risk-On Impact: Increase Capital Efficiency</strong></p><p>To measure our impact in a risk-on cycle, we are looking to see whether borrowers increase their utilization as we expect, while making sure insolvencies remain at or near zero. Since this depends on borrower behavior, we separate borrowers in our analysis between those who react quickly to parameter changes and those who move more gradually. Due to the sometimes complex ways in which protocols incentivize lenders and borrowers, it is occasionally profitable for users to borrow and supply the same assets within the same protocol. Because this practice of recursive borrowing seeks to benefit from arbitrage, it is reasonable for these participants to use as much capacity as they can to maximize earned income. This means that recursive borrowers are usually a large part of the group of borrowers reacting quickly to parameter changes. Since non-recursive users have external reasons for borrowing that may not depend as heavily on parameters, they may not react as quickly to changes in their account limits.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g5UbuEB80KQria7YmP10xQ.png" /></figure><p>In our model for Impact Measurement, we carefully track user behavior and calculate our impact for the two groups separately. The predictable actions of recursive borrowers means that we can ascribe an increase in usage more easily to a specific change, while non-recursive users require more careful analysis to justifiably attribute the effects. In a follow-up article, we will share data from impact calculations on both groups of users and explain our thinking behind some of the details.</p><p><strong>Risk-Off Impact: Reduce Risk</strong></p><p>During a risk-off regime, we are seeking to reduce protocol exposure to potential losses due to volatile market conditions. Since excessive risk is unacceptable regardless of whether or not an issue occurs, we cannot rely on realized metrics to determine our degree of risk reduction. While aiming to maintain actual losses at or near zero, we also look at the expected impact of a severe market crash on the protocol, which we seek to reduce to below a level acceptable by the protocol’s community. On the other hand, setting protocol parameters more conservatively also affects capital efficiency so we must watch utilization to evaluate how much it is being reduced by.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZSfGRLCqFQbQUY8fSywYbQ.png" /></figure><p>While we aim to tighten risk parameters gradually to allow users to reduce utilization in an orderly way, we are aware of customer concerns of parameter changes causing uncertainty for users with near-threshold accounts. To allay these worries, we have developed a strategy for communicating the potential edge cases with the community in a timely way to reduce the risk of market disruption. In a further follow-up, we will discuss in more detail how we approach risk-off regimes and some examples of how it works in practice.</p><p><a href="https://medium.com/gauntlet-networks/impact-measurement-20d9d820fcba"><em>Continued in Part Two: The case of recursive borrowers</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78315cafc9c7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/impact-measurement-78315cafc9c7">Impact Measurement</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Risk Parameter Tuning with CMA-ES]]></title>
            <link>https://medium.com/gauntlet-networks/risk-parameter-tuning-with-cma-es-63f6f3ee7596?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/63f6f3ee7596</guid>
            <category><![CDATA[simulation]]></category>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Tue, 26 Jul 2022 14:47:46 GMT</pubDate>
            <atom:updated>2022-07-26T14:47:46.472Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c6yIvSFmANKrCQLXjokcTg.jpeg" /></figure><p><strong>Introduction</strong></p><p>Since the rapid growth of decentralized finance applications hit the mainstream about 2 years ago, some of the largest protocols in the space have relied on Gauntlet to provide valuable quantitative insights on how to best manage protocol parameters. In this article, we will use the example of lending protocol risk management to provide more detail on the methodology behind our analysis and recommendations. To start let’s briefly recap how a lending protocol works, and how the parameters in question affect market functioning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*V-HnAkhmEqNrqI4tSQsRFQ.png" /></figure><p>Without the loss of generality, we can consider the example of a lending pool where users can borrow a generic USD token (USDx) using ETH as collateral, as shown in the diagram above. The lenders at the top right provide their assets to the protocol as available liquidity, which the borrowers on the left can access by taking a loan against their collateral. Part of the proceeds from repaid loans go towards paying a yield to lenders, providing an incentive to continue supplying their assets to the pool. If a loan is not repaid or the value of the collateral drops to an unacceptable level, the protocol allows liquidators to take over the distressed loans and underlying collateral. To incentivize liquidators to unwind defaulted loans, the protocol offers slightly more collateral than is needed to cover the obligation, leaving a profit margin or bonus for the liquidator.</p><p>The primary risk to lending protocols arises from this process of liquidating offsides positions: what if the collateral is unable to cover the full amount of a problem loan? Such an event is called an insolvency and could trigger losses to lenders if it occurs at a large scale. To manage the risk of insolvencies, lending protocols set aside reserves to cover small losses and tune risk parameters to try to ensure liquidations can occur without major issues.</p><p>This dynamic adjustment of parameters leads to a trade-off between the protocol’s risk level and capital efficiency. To minimize the risk of insolvency, a lending protocol would ideally want to liquidate problem loans quickly long before collateral value becomes a pressing issue, but setting collateral thresholds more conservatively reduces the amount of funding that borrowers are able to access. Similarly, paying liquidators a higher bonus may incentivize them to unwind positions more promptly in times of market stress, but comes at a direct cost to users which may cause them to seek out more competitive rates elsewhere. Achieving balance between protocol risk and user needs is thus a complex optimization problem, and needs a robust framework to quantify the objectives of the protocol and set parameters in order to achieve them. In the next two sections, we will explain how Gauntlet approaches lending protocol management, and in the process introduce some of the quantitative tools we use throughout our product suite.</p><p><strong>Gauntlet Approach to Optimization</strong></p><p>At the core of our methodology is an agent-based simulation of the lending protocol for a range of risk parameters, where we as faithfully as possible replicate observed market conditions and user interactions with a test protocol under realistic scenarios. To quantify the concepts of risk and efficiency, we estimate the percentage of available lending capacity utilized (or Borrow Usage) and expected liquidations and insolvencies in a severe market shock for each set of trial parameters. These outputs allow us to search the range of parameters for values that best meet the protocol’s desired balance, as shown in the diagram below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*16IErvAfdG2ffTlHpmoqxQ.png" /></figure><p>Since lending protocol management involves the trade-offs discussed earlier, we score these trial runs with an objective function that weighs the risks and benefits. At a high level, this can be seen as setting a reward for capital efficiency and a penalty for risk in order to compare simulation outcomes on a quantifiable ranking scale. Using community feedback to adjust the reward and penalty calculations (shown in the formula as R and P), we can select a measure that balances the trade-off to the desired level of risk tolerance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E4c7drfZrrfG_ZJQ8BqTvA.png" /></figure><p>A major challenge in running this type of optimization for a complex protocol is that while we can rank simulation results on an objective scale, it is difficult to know which way to adjust parameters to improve the simulation outcome. For a simple optimization with only two or three parameters it would be practical to sample enough settings to create a full map of performance, but for most real applications this would be very inefficient. If we also knew how the objective function would change with an adjustment in parameters (or in other words, the derivative), we could more quickly converge to the optimal solution by using methods of gradient descent. However, this would be impossible for risk parameter tuning without making extensive assumptions about market behavior. To be able to provide parameter recommendations in a reasonable number of simulation runs, we use a covariance matrix adaptation evolution strategy (CMA-ES), which adapts some elements of gradient descent methods to problems where the derivative is not directly observable.</p><p><strong>Risk Parameter Tuning</strong></p><p>The key insight of CMA-ES is that if we sample a cloud of parameter values, we can infer something about the derivative by looking at where the top performing points land. Thus, by sampling a Gaussian distribution of points and ranking them, we can iterate the shape and location of the next sample set to reflect our best guess of where the optimal value is. A stylized depiction of this method of parameter tuning is shown in the diagram below, where the strategy converges to the maximum in about four iterations:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FkJqIb-T1xzSzeXYGUjnWw.jpeg" /></figure><p>In this example, the top performing points in each sample set are marked in red, the lowest performing points in blue and the points in between in gray. With each step, we adjust the location or shape of the sample distribution to move it in the apparent optimal direction, indirectly applying the concept of gradient descent to this more abstracted problem. Since the strategy does not make assumptions about the nature of the problem besides the ability to rank outcomes, it is a robust and powerful approach for many optimization problems where other methods are not practical. When Gauntlet makes parameter recommendations, we rely on CMA-ES methods to tune our recommendations to protocol objectives. In future articles, we look forward to further elaborating on how we use these tools, both in the area of lending risk management and the variety of other problems we seek to help solve.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=63f6f3ee7596" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/risk-parameter-tuning-with-cma-es-63f6f3ee7596">Risk Parameter Tuning with CMA-ES</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Incentive Spend: Real World Efficiency]]></title>
            <link>https://medium.com/gauntlet-networks/incentive-spend-real-world-efficiency-60739d89a9a7?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/60739d89a9a7</guid>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Tue, 05 Jul 2022 16:24:04 GMT</pubDate>
            <atom:updated>2022-07-06T18:50:09.335Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part Three: An example of Spend Efficiency in practice</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zoh5NoPrHYO3Hddqygdf9g.jpeg" /></figure><p>Over the last two years or so, trading at decentralized exchanges has gone from a very niche concept to competing on a level with the largest traditional exchanges in many markets. With the rapid growth of volumes and revenues at decentralized exchanges, participants have developed new metrics to quantify their economics and understand how to run them more efficiently. Unlike a traditional exchange where users trade against each other in an order book, users at decentralized exchanges trade against Automated Market Makers (AMMs), which are smart contracts that execute their trades based on a preset formula. As explained in more detail in a <a href="https://medium.com/gauntlet-networks/incentive-spend-elasticity-and-optimization-18aa7f608d3d">past article</a>, the core business model of an AMM is to attract assets from Liquidity Providers (LPs) to make them available for user trading. Though newer venues have developed ways to attract LP assets by offering price input, most AMMs pay their LPs an incentive for providing liquidity. As end customers and arbitrageurs trade against the venue to adjust their positioning or take advantage of small price deviations, they generate fee revenue to recoup the venue costs. To frame the discussion with some concrete numbers, let’s look at data from a real decentralized exchange in operation today. Since this is meant only as an illustrative example, we will stick to normalized figures and omit any specific dollar amounts and identifiable venue details.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*82QLISy2Owk1S-VKYaAJmw.png" /></figure><p>The chart above shows the average liquidity provided over the trailing 30 days at this decentralized exchange, divided by the average daily incentive paid over the same period. A ratio of 1000, for example, would mean that over the past 30 days the venue was paying its LPs an average of 0.1% per day on their assets for making them available for trading. While this metric is a product of many individual pools that can all be adjusted to improve spend efficiency, having an aggregate view is useful in judging the overall trend and getting a sense of the economics of the exchange as a whole. In this case, the venue has attracted more liquidity per unit of spending over time, which suggests an improvement in spend efficiency at this level of the value chain.</p><p>However, attracting liquidity alone is not enough to run an efficient exchange. For fee-paying customers to trade with an AMM, the assets they actually want to trade need to be available. Similar to how a retail store would order inventory of items based on turnover, a decentralized exchange needs to ensure that the liquidity being attracted is utilized at a reasonable rate. While there are tradeoffs to utilization being too high, such as possible slippage and price deviations during high volume periods, declining utilization is a drag on efficiency as it means the venue is paying for more assets that are sitting idle.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Tt4Xq_w3i49Mg9b4byOkvQ.png" /></figure><p>In this chart, we see the average volume over the trailing 30 days at the example decentralized exchange, divided by the average liquidity available over the same time period. A reading of 0.1 on this metric would mean that over the last 30 days an average of 10% of the available liquidity was transacted daily. Excluding some occasional spikes for brief periods of high activity, utilization at this venue has mostly stayed in the range of 10–15% turn over on a daily basis. Though how much utilization is reasonable is ultimately a subjective question, having a quantifiable view of it can help exchanges understand how the mix of assets they are attracting fits with consumer needs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Uj8BXCOcClj-B3qFmpweag.png" /></figure><p>To get a full view of the end-to-end Spend Efficiency of the exchange, we can combine the two metrics into a ratio of volume per unit of spending. This final chart from the sample venue shows the <strong>average daily volume divided by the average daily spend</strong> on incentives for the trailing 30 days. Since at most AMMs revenue is directly proportional to volume, this measure can tell us to what extent an exchange is recouping its incentive costs. For example, an exchange that charges a 1% fee on trades would operate at breakeven at a ratio of 100 on the chart, or a ratio of 200 for a 0.5% fee, and so on. This is important information for DAOs to know as paying more incentive than earned revenue dilutes the value of the entire operation. To be sustainable in the longer term, venues need to be aware of their Spend Efficiency and manage their operations to avoid wasteful spending.</p><p>With this problem in mind, Gauntlet launched its Incentive Optimization product this year, aiming to provide quantitative insights and recommendations on how to maximize Spend Efficiency. Through detailed off-chain modeling, Gauntlet can provide granular data on the expected impact of changes and help communities decide between them with the benefit of a rigorous forecasting framework. While incentive optimization has to date been mostly a trial-and-error process, the rapid growth of AMM activity has made errors more costly and the rewards of a forecasting approach more compelling. As decentralized exchanges start to plan for the future more strategically, we look to provide the tools they need to reduce excess spending and redirect incentives to where they generate the most desirable growth.</p><p>—</p><p>If you’re interested in learning more about Gauntlet’s Incentive Optimization product, contact sales@gauntlet.network</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=60739d89a9a7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/incentive-spend-real-world-efficiency-60739d89a9a7">Incentive Spend: Real World Efficiency</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Incentive Spend: Elasticity and Optimization]]></title>
            <link>https://medium.com/gauntlet-networks/incentive-spend-elasticity-and-optimization-18aa7f608d3d?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/18aa7f608d3d</guid>
            <category><![CDATA[crypto]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[tokenomics]]></category>
            <category><![CDATA[defi]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Wed, 18 May 2022 14:56:10 GMT</pubDate>
            <atom:updated>2022-09-07T14:40:03.310Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part Two: The Components of Spend Efficiency</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JcUUviofkQTmcfqE3vIQqg.jpeg" /></figure><p><strong>Recap of AMM Economics</strong></p><p>As new types of trading venues have become popular in Decentralized Finance, much attention has been focused on the mechanics of trading and how it differs from market structures of the past. However, DeFi innovations such as Automated Market Makers (AMMs) also represent an evolution in venue economics. In this article, we will look at some of the unique business decisions faced by DeFi venues and metrics that can drive their success in a rapidly developing market.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KcpkHZ0am8AqXKhZ" /></figure><p>To start, let’s pick up the model of an AMM from a <a href="https://medium.com/gauntlet-networks/incentive-spend-efficiency-and-effectiveness-ce795446af53">previous article</a> and evaluate the three major groups of participants in terms of their business relationship with the venue. Starting from the left, end users pay a fee for the service of being able to trade, acting as the main source of revenue for an AMM. In order to meet customer trading needs, the venue needs assets, which are supplied by Liquidity Providers (LPs) in exchange for a fee. Incentive payments to LPs are the first major cost center of an AMM operation, as they must continuously be made to maintain a sufficient base of assets and attract customers.</p><p>The other major cost center from an AMM’s perspective is arbitrage. Since LPs usually have limited or no price input, the venue must rely on profit-motivated arbitrage traders to keep prices in line with other markets. Though arbitrage traders also pay fees, their profits in excess of fees come at the expense of the AMM’s assets, thus representing another cost center. For an AMM to be sustainably profitable, its fees earned from end users must exceed the costs of incentive payments and losses to arbitrage in the long-term.</p><p>While there are also interesting ways to think about managing arbitrage costs, this article will focus mainly on the tools AMMs can use to better understand their spending on LP incentives. Since payments to LPs are a cost that DAOs have complete control over, it can be a competitive advantage to adjust payments accurately and spend only how much is needed to secure a sufficient asset base. In practice, however, it is difficult to know how much this is and how it changes with market conditions. To dig deeper into this problem, let’s look at a theoretical model of a market reacting to LP payments and outline a possible approach.</p><p><strong>Optimizing Incentive Payments</strong></p><p>One of the biggest challenges with optimizing AMM incentive payments is that market response can be highly nonlinear, with regimes of high or low sensitivity to marginal spending. To see how nonlinear effects may arise, we can consider the relationship between incentive spending and the total value assets contributed by LPs for trading, a stylized example of which is shown in the chart.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Y7GWr5-_a50gZ5yl" /></figure><p>While real venues may have more complex behavior, we can see why in theory an AMMs reaction to incentive payments could look something like this. When incentive payments are relatively low, changes have little effect on total assets, as prospective LPs have options that offer more competitive returns and stay away regardless of small variations. This type of market condition is called inelastic, since it does not respond to marginal changes in inputs. At higher levels of incentive spend, the AMM becomes competitive with other sources of yield and the market condition becomes elastic. In the elastic region, small changes in incentive payments do affect decision making for prospective LPs, and assets available for trading can change significantly from a relatively minor adjustment. Finally, as incentive payments become much larger than competitive yields available elsewhere, the market again enters an inelastic region. In this third scenario, the universe of potential LPs is saturated and further increases in incentive do not attract much additional growth in assets.</p><p>In practice, the elastic and inelastic behavior of LP assets is only one piece of the puzzle. To operate its full business model, an AMM must use LP assets to attract customer trading and earn fees on this volume, which can introduce additional nonlinear factors. Taken together, these effects result in high marginal returns on incentive spending within a certain range, with the effects of spending diminishing the further an AMM is from competitive pricing. Though the curve may be difficult to measure and change constantly with market conditions, we can imagine overall marginal return on incentive spend — the spend efficiency — looking something like the stylized chart version below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kmRpqZaQWorrXIO3" /></figure><p>Supposing a DAO can accurately track spend efficiency in real-time, it can bring significant benefits to their ability to maximize their venue performance. While distributing incentives based on subjective judgment or simple rules of thumb is enough to maintain basic market functioning, it is hard to coordinate a longer term strategy with no view on where spending stands relative to the competitive range. By providing a direct measure of whether the venue is over or underspending, a view like the example chart could help AMMs target spending more productively.</p><p>Driven by this emerging customer need, Gauntlet has built its incentive optimization platform with the goal of giving users quantitative data on both end-to-end spend efficiency and the elasticity of their market across a range of metrics. With this information, decision makers can take a more strategic approach to deploying their funds in a way that maximizes value to LPs, investors, and end users.</p><p>With the general growth of decentralized exchanges, the notional amount at stake in optimization choices has risen dramatically as well. In recent months, over $100 billion of volume has regularly traded on DeFi venues in aggregate. Since this trading was enabled by LP incentives and served as the primary source of income for the venues involved, the value of even a small improvement in efficiency can be very material when applied at scale. In the coming years, AMMs will develop new ways of competing for customer flow and spend efficiency will ultimately determine which venues are able to sustainably generate value from the opportunity. As DeFi market structures continue to mature, market participants will need tools to evaluate the new economic decisions they face in this stage of trading venue evolution.</p><p>—</p><p>If you’re interested in learning more about Gauntlet’s incentive optimization platform, contact sales@gauntlet.network.</p><p><a href="https://medium.com/gauntlet-networks/incentive-spend-real-world-efficiency-60739d89a9a7"><em>Continued in Part Three: An example of Spend Efficiency in practice</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=18aa7f608d3d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/incentive-spend-elasticity-and-optimization-18aa7f608d3d">Incentive Spend: Elasticity and Optimization</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Incentive Spend: Efficiency and Effectiveness]]></title>
            <link>https://medium.com/gauntlet-networks/incentive-spend-efficiency-and-effectiveness-ce795446af53?source=rss-91a15c91277d------2</link>
            <guid isPermaLink="false">https://medium.com/p/ce795446af53</guid>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <dc:creator><![CDATA[DC]]></dc:creator>
            <pubDate>Wed, 11 May 2022 20:00:57 GMT</pubDate>
            <atom:updated>2023-04-05T19:33:18.402Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part One: The historical context behind incentive spend</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JOUOiW13wspOu-bTjz4-Jw.jpeg" /></figure><p>Over the last 500 years of financial history, trading venues have seen tremendous evolution in speed, access, and complexity. However, their purpose has remained the same since the earliest organized exchanges in medieval Europe: to match buyers and sellers of financial assets quickly and at a mutually agreed upon price.</p><p>As electronic trading made venues compete in a global marketplace, some operators started designing features to attract more market share to their trading systems. Since venues compete on speed of execution and quality of price matching, these features have mainly focused on either incentivizing faster trading or improving execution prices achievable by traders. As trading continues to evolve with the emergence of Decentralized Finance (DeFi), new types of venues such as decentralized exchanges have come to grapple with some similar optimization questions.</p><p>To unpack how trading venues use incentives to make their systems more attractive to customers, let’s first look at how order matching works in theory and what customers look for in choosing a venue. When an order to buy or sell an asset at a certain price is submitted, one of two things can happen: it can be matched instantly with an available counterparty willing to take the other side of the trade, or, if no such counterparty exists, it can be made available until someone willing to trade at that price comes along.</p><p>In the first case, the order is called a “taker” order, as it instantly accepts an available price and removes liquidity from the venue. The latter case is a “maker” order, which does the opposite: supplying liquidity to the venue for taker orders to trade against. At venues with a central order book, taker orders represent buying at the offer price or selling at the bid price, while maker orders represent submitting a new bid or offer to the market that is not instantly executed.</p><p>For a trader submitting a taker order, it is advantageous to have a large number of maker orders available, as this ensures ample competition to provide a fair price and allows the trader to execute large trades immediately. On the other hand, a trader submitting a maker order would prefer to have as many taker orders as possible at the venue, as this increases the chances of their order being executed quickly and reduces the need to compete with other maker orders on price.</p><p>Finally, it is important to also consider the incentives of the venue itself. While venues can make money in a variety of ways, most charge a small fee on every trade as a primary source of revenue. By fine-tuning the fees charged to each side of the trade, venues can optimize trading conditions on their system to attract customers and maximize earned fees. For the rest of this article, let’s look at how US equity venues have tried to optimize their fee structures over the last few decades, and how DeFi exchanges (on which tokens, not equities, are traded) have dealt with similar questions more recently.</p><p><strong>Incentive Optimization at US Equity Venues</strong></p><p>The first US equity venue to notably calibrate its fees in an attempt to attract customer volume was the Island Electronic Communications Network (Island ECN) launched in 1996. Driven by what they saw as unfavorable trading practices of market makers on the Nasdaq exchange, affiliates of Datek Securities launched Island ECN as a competing trading venue for most Nasdaq-listed equities. Since Datek was primarily an online retail brokerage looking to start its own trading venue to route client orders, they needed a way to encourage institutional traders to participate in the new system.</p><p>Their solution was to pay members of Island ECN a small rebate if their orders were matched on the maker side of a trade. To subsidize this rebate and generate sufficient revenue for the venue, Island ECN charged a modest fee to orders matched on the taker side. By breaking with the then standard practice of charging equal fees on either side of the trade, Island ECN quickly attracted maker volume and ensured that plentiful liquidity was available for its retail clients. By 1999, the venture was <a href="https://www.wired.com/1999/07/island-2/">hailed</a> as a huge success and a series of rival trading venues with similar fee models sprung up, bringing the practice into the mainstream. As of early 2021, <a href="https://www.cboe.com/insights/posts/a-deep-dive-into-u-s-equities-trading-venues/">industry research</a> estimates that venues using this type of fee model — called the maker/taker model — make up about 50% of US equity exchange trading.</p><p>As specialized fee models became more widespread, some venues experimented with reversing the roles in the original maker/taker model, charging a fee to provide passive maker orders to the venue and paying a rebate to aggressive taker orders. Known as the inverted fee model, this scheme was first tried by the Nasdaq Boston Exchange in 2009 and now makes up about 5% of US equity exchange volume. To visualize how the maker/taker and inverted fee structures differ, we can imagine a hypothetical market in which the available bid and offer prices are the same across all venues, excluding the impact of fees.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2iCrhQRd7S0OTVjB" /></figure><p>Assuming trading happens only at these prices, we can then see the net result for makers or takers after fees in either type of venue, as shown in the diagram. Note that the inverted fee venue effectively creates a narrower spread between the bid and offer price, as some of the advantage makers have from trading on the less aggressive side is eroded by the fee structure. Conversely, the maker/taker venue has an effectively wider spread as taker orders must pay a fee in addition to trading on the more aggressive side of the bid or offer.</p><p>In reality, however, available quotes and market behavior are not the same across different types of trading venues. The higher incentive to provide maker quotes in maker/taker venues produces more competition on the maker side, which offsets some of the disadvantages for the taker side by leading to narrower spreads. In inverted venues, the disadvantage to the maker side created by fees is offset by incentivizing more aggressive trading, which leads to wider spreads and faster execution for orders on the maker side.</p><p>Since trading fees at US equity venues have been competitive for over 20 years, the nuances of trading at major exchanges are well-known and small inefficiencies are mostly arbitraged by established players. The costs passed through to end users in this mature market are fairly uniform once market impact and the opportunity cost of execution speed are taken into account, as suggested by a recent <a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3625801">study</a> on a full year of trading data from a large US hedge fund. Due to the highly competitive nature of the equity exchange business and the widespread presence of sophisticated arbitrage traders, microstructure tradeoffs in US equity markets occur outside the view of most participants, who can nonetheless benefit from the quick execution and favorable pricing they facilitate.</p><p><strong>New Venue Structures in DeFi</strong></p><p>Up until this point, we have looked primarily at traditional equity venues with a central order book, but these same principles apply to all order book venues, such as commodity futures exchanges and centralized crypto exchanges. Most of these venues use some version of the maker/taker model or charge flat fees to each leg of the trade, though recently some (like eToroX) have experimented with an inverted fee model as well.</p><p>But what about new types of venues where there is no central order book? In the last few years, an increasing volume of crypto trading has occurred through automated market makers (AMMs) where prices are set by a formula rather than a traditional order book structure. To understand how fee optimization occurs at these new venues, let’s first look at the theoretical concept of a simple AMM and how it matches orders.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*_4IjAljZthg91aXT" /></figure><p>At the most basic level, an AMM is a smart contract that creates a market between Asset X and Asset Y by holding both assets and trading them according to agreed rules. To supply the needed inventory for trading, assets are deposited into the AMM by market participants called Liquidity Providers (LPs), who retain fractional ownership of the pooled assets. Though some AMMs have given LPs the option to provide price input, let’s first consider the simpler case where their role is merely passive. In this case, the AMM is programmed to maintain quantities of those assets based on a predetermined bonding curve. If a trader swaps a certain quantity of Asset X with the AMM, they will receive a quantity of Asset Y as given by the slope of the bonding curve around the current level of both assets in the AMM.</p><p>For any convex bonding curve, this creates a situation where the price of Asset Y denominated in Asset X increases as more Asset Y is removed, leading to an equilibrium level at which prices match those available at other venues. As the price fluctuates around the equilibrium point, arbitrage traders are incentivized to remove small amounts of the temporarily underpriced asset from the AMM and sell it at other venues, pocketing a small profit and correcting the AMM price to keep it in line with the broader market.</p><p>This trading dynamic splits AMM markets into three categories of participants: LPs, Arbitrageurs, and Traders. LPs passively provide assets for others to trade with at any price and maintain fractional ownership of the combined asset pool. Arbitrageurs trade small quantities with the AMM frequently to maintain price alignment with outside venues. And finally, traders who transact with the AMM for the same reasons they would transact with any other venue. While traders who transact with the AMM strictly as a trading venue can be seen as an analogy to takers in an order book market, price making in an AMM market is done by both LPs and arbitrageurs.</p><p>Since arbitrage needs a profit incentive, the profit of the arbitrageurs comes at a cost to LPs, called an “impermanent” or “divergence loss”. LPs also need a profit incentive to provide their assets to an AMM for trading, so they must be compensated for their losses and opportunity cost from fees charged to arbitrageurs and other traders. The way most AMMs go about this is charging a fee on every transaction made by traders and arbitrageurs, and passing on some of the revenue to LPs as a yield or incentive. Though early generations of AMMs used a flat fee set mostly by subjective judgment, recent developments have been made in calibrating fees for market conditions and the business needs of the venue, creating some familiar trade-offs.</p><p><strong>Incentive Optimization in DeFi</strong></p><p>The end goal of AMM optimization is the same as for any other type of venue: to attract user volume and generate higher earnings from the trading the venue facilitates. One possible approach is to pay higher yields to LPs for depositing their assets at the AMM, with the goal of accumulating a large pool of assets that can support heavy trading volume. Making these yields sustainable, however, requires charging a high fee on transacting with the AMM and limits the activity of arbitrageurs, as they become unable to profit off dislocations in prices that are smaller than the fee rate.</p><p>This friction leads to noisier prices and lower arbitrage volume, both of which are undesirable for the overall success of the venue. Fees that are too low, on the other hand, cannot support a high yield for LPs to make up for loss due to price divergence, and may struggle to attract assets from competitors. As AMM markets have started to mature and enter the mainstream, <a href="https://forum.sushi.com/t/optimizing-incentives-an-update-from-gauntlet/10131">progress</a> has been made on quantifying these trade-offs in an actionable way and tailoring AMM markets to the needs of their users.</p><p>As a more advanced generation of AMMs enters the market and brings a number of changes to the trading environment, fee structures will need to adjust for the new features being offered to users. Dynamic fees that rise and fall based on market inputs will allow more fine tuning of incentives to provide consistently balanced trading conditions. Tiered fee pools, like those used for Uniswap v3, allow traders and LPs to choose between AMMs at different fee levels depending on their individual risk tolerance and trading needs. Finally, concentrated liquidity features, which are also being used for Uniswap v3, give LPs the ability to limit the range of price over which their assets can be used for trading. This last innovation is especially interesting in the context of optimization as allowing LPs to have some price input could conceivably reduce the compensation they require for providing their assets for trading, opening the door for more experimentation with fee and incentive structures. As trading venues continue to evolve, there is a lot of work to be done in creating venues that most efficiently permit users to use all these features. By being aware of the historical trade-offs in market design, and the relation of DeFi markets to this context, we can approach these design decisions aware of their impact on both trading environments and the overall competitiveness of new market venues.</p><p>—</p><p>If you’re interested in learning more about Gauntlet’s incentive optimization platform, contact sales@gauntlet.network.</p><p><a href="https://medium.com/gauntlet-networks/incentive-spend-elasticity-and-optimization-18aa7f608d3d"><em>Continued in Part Two: The Components of Spend Efficiency</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ce795446af53" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gauntlet-networks/incentive-spend-efficiency-and-effectiveness-ce795446af53">Incentive Spend: Efficiency and Effectiveness</a> was originally published in <a href="https://medium.com/gauntlet-networks">Gauntlet</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>