Introduction

In this report, we provide an in-depth analysis of GMX v2 and its associated token, $GMX. With a primary goal of forecasting the token’s future price directionality, we will examine the factors that may influence its performance after the stealth launch of the protocol’s most recent upgrade.

Overview

With v1, GMX has demonstrated extraordinary performance, surging from $20 in January 2022 to an all-time high of $90. Amidst the downfall of prominent DeFi giants, GMX has carved out its own narrative, showcasing an appetite for leverage in both bullish and bearish market conditions.

During its early days, GMX experienced rapid success and managed to capture a significant portion of the decentralized perpetual futures market, reaching up to 90% market share. However, as the protocol attracted more attention, its weaknesses started to become more apparent. 

Why GMX v2?

As a result of the protocol’s success, more eyeballs looked at the project and its weaknesses were highlighted. At this point, the team recognized the importance of addressing these issues to maintain its competitive edge and sustain its position as a leading player. Consequently, the protocol underwent significant upgrades and improvements in its latest version, GMX v2.

The limitations of v1 stemmed from the underlying mechanism design, where $GLP represented a basket of assets that would attempt to self-rebalance while being the active counterparty to traders. This introduced a series of limitations:

• Implicit Directional Bias, since $GLP token’s price is influenced by the performance of its underlying assets. 
  • $GLP increases in price when the underlying assets are in an uptrend. 
  • $GLP decreases in price when the underlying assets are in a downtrend. 
  • This leads to price volatility and uncertainty for liquidity providers.

• Counterparty Risk: $GLP holders act as the counterparty to traders. 
  • When traders make profits, it comes at the expense of $GLP holders, and vice versa. 
  • As a liquidity provider (and $GLP holder), you are effectively betting on traders losing money over time.

We can observe how there is an unpredictable Risk-Reward Balance based on whether you are a liquidity provider or a trader. While earning fees can be attractive, it also involves potential losses if traders consistently outperform the market.

While the GLP model had its merits and unique features, the transition to GMX v2 intends to offer a more capital-efficient and flexible way to get exposure to decentralized perpetuals trading.

From a user’s perspective, this translates into key improvements in the following areas:

GMX v2

Unlike in GMX v1, where traders swapped against a basket of assets that would self-rebalance in the form of $GLP, in GMX v2 liquidity is provided through individual GM pools, also known as GMX Market pools. These pools enable users to engage in leverage trading, borrowing, and swaps.

Note that some markets will be labeled as SWAP-ONLY or SPOT-ONLY markets.

Each GM pool will consist of 3 components:

• Index Price Feed: The price feed that determines the opening and closing of long and short tokens within the pool.
• Long Token: This token backs long positions in the market.
• Short Token: This token backs short positions in the market.

Let’s consider an example market, ETH/USD (ETH-USDC):

• Index Price Feed: ETH/USD
• Long Token: $ETH tokens back long positions
• Short Token: $USDC tokens back short positions

As a result, liquidity must be provided in GM tokens. This means that in order to become a liquidity provider, users must buy GM tokens. This purchase may have a positive or negative price impact depending on the pool’s token balance. Because of that, it is recommended to use the “Pair” option and buy GM tokens with equal USD amounts of long and short tokens.

Note that since liquidity providers are dealing with GM tokens with their respective prices, they must be aware of the following properties in order to properly assess their exposure:

• GM token price depends on long and short token prices and traders’ open positions’ net PnL.
• Fees from leverage trading and swaps increase the GM token price.
• Some tokens may have a spread, affecting the buying/selling price.
• Stablecoin tokens have a spread based on the Chainlink price of 1 USD

Notably, there is no guarantee that long and short positions will always be balanced, and sudden price spikes may cause temporary imbalances. However, the shift to isolated markets represents a significant improvement:

• With isolated markets, liquidity providers can choose what assets they want to provide liquidity for.
• The implementation of funding fees, along with the price impact function, prevents price manipulation while balancing the skewness of longs and shorts. 
• Chainlink’s low-latency oracles have been integrated to increase efficiency and mitigate front-running. 
• The integration with Chainlink significantly improves the execution of Stop Loss and Take Profit orders, making them more likely to be executed at the set price.

It is also worth noting that GM (liquidity) tokens automatically compound, unlike GLP.

Isolated Markets

Isolated markets replace multi-asset trading through the GLP pool. This offers benefits such as:

• Diversified Asset Pools reduce directional bias and enhance predictability. 
• Better hedging, since with isolated pools, Liquidity providers (LPs) now have increased control over asset exposure, because they can choose what assets they want to provide liquidity for.
• Permissionless asset listings provide more market opportunities for traders. Instead of the original 5 assets ($BTC, $ETH, $UNI, $LINK, and $AVAX), v2 was launched with 9 tradeable tokens: $ETH, $BTC, $ARB / $AVAX, $LINK, $DOGE, $XRP, $LTC, $UNI, and $SOL.
  • Blue chips: $BTC, $ETH…
  • Medium market cap: $LINK, $UNI, $ARB, $AVAX…
  • Synths: $SOL, $DOGE, $LTC, $XRP…

Price Impact & Low-Latency Oracles

In v2, GMX continues to operate on an Oracle-based model, relying on external oracles to fetch prices from centralized exchanges. This ensures zero-slippage trading, executing trades at the market price determined by low-latency Chainlink oracles. The major difference is that, through this partnership with Chainlink, this new solution provides a pull-based that is much more likely to fulfill limit orders at the exact specified price.

• Substantially reduces latency with per-block updates for faster, more timely information.
• The probabilities of Stop and Take-Profit orders being timely executed increase significantly. 
• Mitigates frontrunning risks by maintaining price privacy until transactions are settled.
• Enhances gas efficiency by streamlining the validation process of Oracle updates.

Unlike standard Chainlink price feeds, the pull-based model of low-latency oracles generates Oracle reports on a per-block basis. This adds an extra degree of confidence since users can now retrieve these reports and ensure that the off-chain data objectively represents the on-chain execution price.

This partnership entails the following conditions:

• GMX will cede 1.2% of protocol fees to Chainlink, approximately $2-3M annually.
• Costs are covered by the GMX treasury, and not passed on to liquidity providers.

Funding Fees

By introducing funding fees, GMX addresses the criticism of slippage-free trading in GMX v1.

• Funding fees have now been introduced to maintain a balance between long and short positions, protecting liquidity providers’ funds.
• Funding prevents losses for liquidity providers in scenarios where either long or short positions dominate the market.

While this offers better risk management for liquidity providers (especially when they lose against traders), this carries on additional costs for traders. Regardless, the reason behind introducing funding fees is to redistribute risk more equitably. 

In practice, trading on the non-dominant Open Interest (OI) side significantly reduces fees.

Enhanced Trading Experience

The trading experience has improved significantly thanks to the support of new order types. With v2, GMX will support the following orders:

• Market Order: Buy or sell tokens immediately at the current market price for faster execution.
• Limit Order: Buy or sell tokens at a specific price or better, ensuring execution at the desired price.
• Stop-loss Order: Set a price to limit potential losses on a position, triggering a sell order if the price reaches or goes below the specified level.
• Take Profit Order: Set a price to secure profits on a position, triggering a sell order when the price reaches or goes beyond the specified level.

Reduced Trading Fees 

The fee structure is designed to incentivize liquidity providers and ensure a fair distribution of fees among users in the GMX ecosystem. What’s most relevant for v2 is that the spot and perpetual trading fees have been halved from 0.1% to 0.05%, making trading more cost-effective for users.

The fee reduction compensates for the introduction of price impact and funding fees, reducing risk for liquidity providers. Traders can view the fee details on the interface when making a trade and be aware of the potential costs associated with their transactions.

• Open / Close Fees: The fee for opening or closing a position is either 0.05% or 0.07% of the position size.
  • The actual fee percentage depends on whether the trade increases the balance of longs and shorts in the pool (0.05% fee) or not (0.07% fee).
  • This fee also applies to partially decreasing the position size or increasing the size of an existing position.
• Swap Fees:
  • For normal swaps, the fee is either 0.05% or 0.07% of the swap amount.
  • Similar to open/close fees, the actual fee percentage depends on whether the trade increases the balance of tokens in the pool (0.05% fee) or not (0.07% fee).
  • For stablecoin swaps, the fee is a fixed 0.01% of the swap amount.
• Price Impact: There may be a positive or negative price impact for both increasing/decreasing positions and swaps.
  • A positive price impact improves the long/short balance or tokens in the pool, resulting in a more favorable entry/exit price for the trade.
  • A negative price impact indicates a less favorable entry/exit price.
• Funding Fees: While a position is open, there may be positive or negative funding fees.
  • If there are more longs than shorts, longs pay a funding fee to shorts.
  • If there are more shorts than longs, shorts pay a funding fee to longs.
  • The funding fee rate is subject to change over time based on the balance of longs and shorts in the pool.
• Borrowing Fees: Borrowing fees are imposed to avoid a scenario where liquidity is fully reserved by users opening equal long and short positions for a small cost.
  • If there are more longs than shorts, longs pay the borrowing fee.
  • If there are more shorts than longs, shorts pay the borrowing fee.
  • The borrowing fee rate is subject to change over time based on the pool utilization percentage.
• Execution Fee: There are two transactions involved in opening/closing a position: the user’s request and the execution by keepers.
  • The cost of the execution by keepers, referred to as the “Max Execution Fee,” is displayed in the confirmation box.
  • Any excess execution fee is sent back to the user’s account after the order is executed.

Even though some users have already expressed concerns about how this might affect GMX’s profitability, others believe lower fees will attract more users, boosting long-term protocol success.

Fee Distribution

• 10% of protocol fees will be allocated to the GMX Treasury.
• Among GMX stakers and GM liquidity providers, the current 70:30 ratio will be maintained.
• The effective allocation for the protocolFee parameter will be as follows:
  • 10% distributed to the GMX Treasury.
  • 63% distributed to liquidity providers in each specific liquidity pool.
  • 27% distributed to a pool that goes to GMX stakers across all chains.
• Chainlink oracles will be funded by a sub-allocation from the 10% GMX Treasury allocation.

Below you can find an example of the fee distribution applied against a $200,000 open position:

Composability

$GLP continues to be one of the most widely adopted “money legos” in all of DeFi. The composability of GMX v1 has enabled the development of multiple protocols that would not have been possible otherwise (vaults, auto-compounders, delta-neutral or hedging strategies…). 

Even though the yield that one can get on $GLP has decreased in expectation of v2, it has consistently delivered >20% returns. This adoption, especially on the Arbitrum ecosystem, is one of the reasons why $GLP will not be phased out. Instead, both v1 and v2 will run in parallel, and each will have its own integrations built on top. 

In v2, funding fees will replace $GLP as the “money lego” of choice. In the past, dYdX was the primary on-chain venue offering sustained Open Interest levels. With dYdX now moving to Cosmos, GMX (along with Kwenta) can offer an alternative to centralized exchanges to facilitate funding rate arbitrage opportunities. 

At the same time, the ability for protocols like Lyra to hedge their exposure will still be possible on v2. With the launch of many new projects working on volatility strategies and structured products on Arbitrum, like Opyn, Rysk, Smilee, and Gammaswap… it is reasonable to expect an increasing number of integrations. This will drive more volume and open up a constant flow of capital into the protocol.

Key Takeaways

• GMX v2 has undergone significant upgrades and improvements to address the limitations of GMX v1 and maintain its competitive edge.
• The transition to GMX v2 offers a more capital-efficient and flexible way to get exposure to decentralized perpetuals trading.
• GMX v2 introduces isolated markets, allowing liquidity providers to choose what assets they want to provide liquidity for and offering better hedging opportunities.
• Low-latency Chainlink oracles are integrated into GMX v2, reducing latency and improving the execution of Stop Loss and Take Profit orders.
• Funding fees have been introduced in GMX v2 to balance long and short positions and protect liquidity providers’ funds.
• The fee structure in GMX v2 has been reduced to incentivize liquidity providers and make trading more cost-effective for users.
• Composability remains a key strength of GMX, with both v1 and v2 running in parallel and providing opportunities for integrations with other DeFi protocols.
• The adoption of GMX v1 in the Arbitrum ecosystem has been significant, and the transition to v2 is expected to attract more volume and capital into the protocol.

Conclusion

With v2, GMX has addressed the limitations of its previous version by introducing isolated markets, low-latency Chainlink oracles, funding fees, and a more user-friendly fee structure. These improvements have enhanced capital efficiency, reduced risk for liquidity providers, and improved the overall trading experience for users.

Research Links

• GMX v2 (beta)
• GMX stats
• Trading guide
• Providing Liquidity

Revelo Intel Links

• GMX v2 vs Synthetix v3
• GMX project breakdown

Data Sources

• Coingecko GMX
• DefiLlama

Disclosures 

Revelo Intel has never had a commercial relationship with GMX and this report was not paid for or commissioned in any way.

Members of the Revelo Intel team, including those directly involved in the analysis above, may have positions in the tokens discussed.

This content is provided for educational purposes only and does not constitute financial or investment advice. You should do your own research and only invest what you can afford to lose. Revelo Intel is a research platform and not an investment or financial advisor.