Custom collateral Token? This article sorts out the highlights of GMX’s new V2 version

Author: Leo, BlockBeats

GMX can be said to be one of the most successful DEXs on Arbitrum. Back when interacting with ARB before, GMX was the most important DEX for ARB airdrop winners to interact with, and it was also the DEX that everyone liked after the ARB airdrop.

Although GMX is a relatively strong player, its current version still has some shortcomings. For example, open positions are more risky for LPs. It is friendly to large traders, but charges higher fees for small and medium traders than other DEXs. Fewer tradable assets, capital inefficiency, etc.

GMX also released its V2 version test network some time ago, which can be experienced on the Avalanche Fuji test network, and its V2 version main network may be launched in the near future. Let's take a look at the highlights of the GMX V2 version.

Trader Fee Collection Scheme

Compared with the previous version, the new Fee charging scheme has some changes:

• Opening/closing fees have been halved from 0.1% to 0.05%.

-Introduction of funding rates paid to the weak side.

• Borrowing fees are retained to ensure OI pools are not accumulated for free.

• Implement price impact (unfavorable, especially for medium-large deals)

Compared with other DEXs, GMX’s market impact is quite large. The following are other DEX fee collection plans.

The fee level and mechanism of each DEX are very different, so it is meaningless to compare different types of DEX, but the following Fee indicator data can still be used as a reference:

GMX v1: 0.1% opening/closing + lending fee

GMX v2: 0.05% open/close order + funding rate + borrowing fee + price impact

dYdX: 0.02% for pending orders/0.05% for taker orders + there is a discount when the trading volume is large

Kwenta: 0.02% pending order / 0.06 - 0.1 taker order + $2 execution fee

GNS: 0.08% open/close order + 0.04% spread + price impact

LVL: 0.1% opening/closing + borrowing fee

The V2 update will increase GMX's small-amount transactions from the market's large-amount transactions.

Innovative liquidity provision mechanism

In the new model, the market will use ETH as collateral for long orders and USDC as collateral for short orders to achieve greater flexibility and scalability. For example, to open an order on SOL, ETH is the long collateral, and USDC is the short collateral. For long orders, it is equivalent to holding both ETH and SOL long orders. Traders can also choose their own long-short collateral. In addition, the liquidity of each token pair will be isolated, allowing LPs to choose their preferred liquidity token pairs according to their risk preference/return.

Another interesting change is that traders can choose a leverage of less than 1, such as 0.9 or 0.8.

On the face of it, the returns of a large capital portfolio will be smaller, but safer. Riskier spam coins may have a higher APR due to high trading volume, but if the pool is unbalanced (due to higher volatility relative to ETH), the risk of a trader losing capital in profit or loss is higher.

Introducing a funding rate mechanism will make the new pool more balanced in terms of GLP, which is a great development direction for LPs, as their exposure to trader P&L will be reduced (or even 0 in a fully balanced case ). But the price is that liquidity bootstrapping will be a more difficult process due to decentralization, and some pools may struggle to attract enough liquidity.

Expand the trading market

In the GLP model, all tradable assets need to be included in the GLP pool, but not in the new model. You can add tradable pairs yourself without LP needing to touch the asset, because some tokens will be collateralized with ETH and stablecoins instead of hard assets, while other markets will continue to be collateralized with hard assets (ARB/USDC is backed by ARB, but LTC/USDC will be backed by ETH)

But this market collateral approach has a downside. Such highly profitable trades may end up without collateral if the value of ETH and the selected asset fluctuate widely.

To solve this problem, GMX will introduce a new feature called Automatic Deleveraging (ADL), which will automatically close some trades when a certain profit is reached to avoid collateralization issues. The feature will only affect a small number of transactions, and only in exceptional circumstances.

Additionally, the inclusion of price impact calculations solves the problem of oracle manipulation in a small number of tokens. This price manipulation attack is an important attack method in GLP, and the design of the new model will solve this problem well.

Faster price oracles

Additionally, there are a few features that can significantly improve the user experience. Both are very innovative and can be directly translated into order execution, which is the main pain point of contract DEXs currently priced by oracles.

Low Latency Oracles: Faster oracles will make front-end trading more difficult and make for a smoother experience regarding entry/exit prices and the time it takes to execute orders. The oracles used in the new version of GMX are products from Chainlink, and GMX will be the first protocol to implement them.

Backtrack Orders: Limit/Stop orders will always be executed as soon as the oracle hits the selected price, even if the price changes happen very quickly (in previous V1 versions, some orders were not always Executed during times of high price volatility)

Conclusion

Overall, the GMX V2 version has made some meaningful improvements, which has stabilized GMX's position in the DeFi field. **The new function of GMX V2 completes the balance of efficiency, simplicity and transparency in transactions, and at the same time opens up new ways for traders and liquidity mining to explore. **Of course there will be some risks, such as system development, fee collection, attracting liquidity, etc., but the current public testnet is still in the trial stage. With the completion of the testnet and the launch of the mainnet, the GMX V2 version may usher in more Good development.