The GMX V2 version was officially launched on August 4, 2023. This article reviews the development and issues of GMX V1, contrasts it with the changes in V2, and analyzes the potential impacts.
I. GMX V1: An Effective Model for Derivative DEX Protocol
GMX V1 was launched at the end of 2021, adopting the GLP model. It provided a concise and effective trading model, introducing the narrative concept of “real returns”. It held a significant position in the derivatives DEX protocol. Many projects forked GMX V1’s model.
The GMX V1 protocol captured a considerable amount of fees. Since 2023, the income from the GMX V1 protocol has been 98.1 million USD, ranking eighth among all projects and first in the derivatives DEX sector.
However, GMX V1 also had limitations:
1. Imbalance of Open Interest (OI) exposed LP providers to significant risks.
GMX V1 fees consisted of opening/closing fees and borrowing expenses, without any funding rate. The borrowing fees ensured there were holding costs, preventing liquidity from being indefinitely occupied. Dominant parties paid more, but since both long and short sides were charged, no arbitrage opportunity existed. OI could not be quickly rebalanced through arbitrage.
If this imbalance isn’t addressed, in extreme cases, the GLP pool could suffer massive losses, with LP providers incurring these losses, leading to protocol collapse.
2. Limited Tradable Assets
GMX V1 allowed trading of only five assets: BTC, ETH, UNI, LINK, and AVAX. In contrast, DYDX and Synthetix offered dozens of trading assets. Gains introduced foreign exchange assets, and the new platform HMX included commodities and US stocks.
3. Higher Costs for Small and Medium Traders
Both the opening and closing fees of GMX V1 were 0.1%, relatively high. With intense competition in the derivatives DEX sector, many protocols charged fees below 0.05%.
II. GMX V2: Ensuring Protocol Security and Balance
The essence of GMX V2 is to ensure the protocol’s security and balance. It modifies the fee mechanism to maintain a balance between long and short positions, aiming to reduce the likelihood of systemic risks during intense market fluctuations. With the establishment of isolated pools, it adds high-risk trading assets while controlling the overall risk. Through collaboration with Chainlink, we offer more timely and efficient oracle services, reducing the likelihood of price attacks. The project team has also considered the relationships between traders, liquidity providers, GMX token holders, and the ongoing development of the project, ultimately making adjustments and balancing the distribution of protocol revenues.
2. Adjustment in the Fee Model: Introduction of Funding Rate and Price Impact Fee.
The fee model of GMX V2 has been significantly adjusted. The core focus is on how to balance the long and short positions and enhance the efficiency of capital utilization. The specific changes in the fee model are as follows:
• Reduction in the Opening/Closing Fee.
The previous fee of 0.1% has been reduced to 0.05% or 0.07%. The exact rate depends on whether the opening of a position contributes to the balance between long and short. A lower fee is charged if it’s beneficial.
• Introduction of Funding Rate, where the dominant side pays the funding rate to the weaker side.
The funding rate will be adjusted in segments. When the dominant side’s position/full position is between 0.5 and 0.7, the funding rate remains relatively low. When it reaches 0.7, it escalates to a higher level, enlarging the arbitrage opportunity and enticing arbitrage funds to enter, thereby restoring the balance.
• Maintaining Borrowing Costs to prevent infinite occupation of liquidity.
• Addition of Price Impact Fee. The larger the position and the more detrimental to the balance between long and short, the higher the fee.
The Price Impact Fee emulates the dynamic process of price changes in the order book trading market. That is, the larger the position, the greater its impact on price. This design elevates the cost of price manipulation, diminishes the likelihood of price manipulation attacks, and prevents sudden price crashes or spikes, all while maintaining balanced positions and optimal liquidity.
The chart below displays the Price Impact Fee rate faced by different position sizes in a simulated scenario. It’s evident that larger positions incur a higher fee. The X-axis represents position size (in million USD) while the Y-axis signifies the rate (in bps).
Moreover, the more a position opens against the balance of long and short, the higher the fee. The following table shows the fees charged under different long-short balance scenarios in a simulated environment. The first column represents the opening size, and the top row indicates the initial pool’s imbalance position size.
To briefly compare the fees of some major derivatives DEX protocols:
DYDX: maker 0.02%, taker 0.05%. The larger the transaction volume, the bigger the discount.
Kwenta: maker 0.02%, taker ranges from 0.06% to 0.1%.
Gains Network: 0.08% opening/closing fee + 0.04% spread + Price Impact Fee.
It is evident that the cost of GMX V2 remains relatively high. However, it has decreased from a previously high level to a medium level, with the opening/closing costs dropping by nearly 50%. For small and medium-sized traders, the charges for V2 are more accommodating.
3. Liquidity Provision: Addition of the isolated pool mode, addition of synthetic assets.
The liquidity pools of GMX V2 are called GM pools, and each pool operates independently. The official website provides details on the capital, capital rate, and capital utilization rate of each pool.
The advantage of the isolated pools is that different token markets can have varying underlying support and parameters. This facilitates individualized risk control and a high degree of flexibility, expanding the trading assets while maintaining manageable risks. For liquidity providers, they can also choose their risk exposure based on their risk preference/return expectations. The issue with isolated pools, however, lies in the fragmentation of liquidity. Some pools might not be able to attract sufficient liquidity.
Currently, GMX V2 has categorized three different types of markets:
• Blue Chips: BTC and ETH. These two tokens have a low likelihood of price manipulation, allowing for a lower pricing impact fee. This makes it more competitive than CEX. Both utilize native tokens for support.
• Mid-cap Assets: Assets valued between 1 billion and 100 billion USD, with significant liquidity and trading volume on CEX. However, they’re susceptible to sharp price fluctuations due to external factors like regulatory news. For these assets, a higher price impact fee is set, and liquidity is kept below other external markets to increase attack costs. LINK/UNI/AVAX/ARB/SOL fall under this category, using native tokens for support.
• Mid-cap Synthetic Assets: Instead of using native tokens, ETH serves as the primary liquidity support. DOGE and LTC are examples. The challenge with such assets is that if the associated token sees a significant short-term increase, the ETH in the pool might be insufficient to cover all returns. For instance, if the pool consists of 1,000 ETH and 1 million USDC, with DOGE’s maximum long position limited to 300 ETH, a tenfold increase in DOGE’s price (with only a twofold increase in ETH) would result in profits exceeding the pool’s ETH value. To avoid this, the ADL (Automatic Deleveraging) feature is introduced. If the pending profit exceeds a set market threshold, profitable positions might be partially or entirely closed. This ensures the market’s solvency and full payment of all profits at close. However, for traders, automatic deleveraging might lead to the loss of advantageous positions, missing out on subsequent profits.
According to a report by Chaos Labs, they recommend that during the initial phase of V2, the unclosed contract limits for BTC and ETH be set at 256 million USD each, and AVAX/LINK at 4 million USD each, with other tokens capped at 1 million USD. Adjustments can be made based on actual operational outcomes. As of now, the total TVL of GM pools stands at approximately 20 million USD, which is still far from the set limits.
3. Enhancing User Experience: Addition of coin-margined contracts, faster execution speeds, and reduced slippage.
In GMX V1, traders could only establish USD-margined contracts. No matter which asset traders used to open positions, it was converted into USD value based on the price at the time of opening. Profits were calculated as the USD value at closing minus the USD value at opening.
In GMX V2, coin-margined contracts have been introduced. Traders can deposit the relevant trading assets as collateral without converting it to USD. This meets more of the traders’ needs and offers a richer portfolio approach.
Furthermore, GMX V2’s oracle system will price each block, ensuring orders are executed at the latest possible prices, resulting in faster execution speeds and lower slippage.
4. Allocation Model
To sustain the project’s long-term development, the revenue protocol of GMX V2 has also been adjusted. 8.2% will be allocated to the protocol treasury, which can be used for project operations and related matters.
GMX V1: 30% allocated to GMX stakers, 70% to GLP providers.
GMX V2: 27% allocated to GMX stakers, 63% to GLP providers, 8.2% to the protocol treasury, and 1.2% to Chainlink. This allocation has been approved by a community vote.
III. GMX V2 Operational Status
GMX V2 has been in operation for about 2 weeks, with a TVL (Total Value Locked) of approximately 20 million USD, a daily trading volume of 23 million USD, a daily protocol income of 15,000 USD, outstanding contracts worth 10.38 million USD, and daily active users ranging between 300–500. As a start, this performance is decent without trading incentives.
Some V1 users have already migrated to V2. V2’s trading volume and daily active users are roughly 40%-50% of V1’s volume. Comparisons of trading volume, protocol revenue, and user data between V1 and V2 are illustrated in the following chart:
Currently, GMX V2 traders are in a net loss position, with a cumulative net loss of 40,000 USD.
Looking at the yield, GMX V1’s recent yield remains sluggish, with GMX staking yield at 1.44% this week, GLP (arbitrum) at 3.18%, and GLP (Avalanche) at 8.09%. By comparison, GMX V2’s yield is higher, as detailed in the list below:
After the launch of GMX V2, the market’s enthusiasm was not high, and the capital reaction was lukewarm. The primary reason is that the recent market volatility has dropped to historically low levels, the overall trading volume has shrunk, and there is intense competition within the sector, leading to lackluster growth in protocol revenue.
GMX V1 is a successful model in the derivatives DEX arena with many followers. The delivery of GMX V2 essentially meets market expectations, demonstrating the GMX team’s strong protocol design capabilities. Mechanistically, V2 has enhanced the balance of liquidity pools, expanded the types of tradable assets, and offered various collateral positions. For liquidity providers and traders, this means a richer investment approach, better risk balance, and lower fees.
However, in its initial phase, due to the adoption of independent pools, there’s an issue of fragmented liquidity, and some assets might suffer from insufficient liquidity. Moreover, the GMX project team has not actively engaged in market marketing actions or trading incentive measures, showing no significant impact on adding new users or trading volume in the short term.
Essentially, GMX V2 places more emphasis on protocol infrastructure, protocol security, and balance. In the current bear market, focusing on the foundational architecture, ensuring protocol safety, and using accumulated data for better risk parameter design might be more beneficial for the project’s future growth in a bull market. By then, it can offer a higher open contract capacity, a richer trading market, and in tandem with market enthusiasm, introduce more marketing initiatives to attract more new users.
LD Capital is a leading crypto fund who is active in primary and secondary markets, whose sub-funds include dedicated eco fund, FoF, hedge fund and Meta Fund.
LD Capital has a professional global team with deep industrial resources, and focus on develivering superior post-investment services to enhance project value growth, and specializes in long-term value and ecosystem investment.
LD Capital has successively discovered and invested more than 300 companies in Infra/Protocol/Dapp/Privacy/Metaverse/Layer2/DeFi/DAO/GameFi fields since 2016.