GMX V2 - Coding Ingenuity or Prolonged Farce

This in-depth overview will be the first of many projects across the crypto sphere we all hate to love. So step forward GMX, the darling sweetheart of Arbitrum, a decentralised perpetual trading exchange that has spawned numerous forks over its existence, but to date none have matched the success of this innovative behemoth.

We’ll be traversing our way through the recent upgrade to V2, stopping at junctures such as:

• What does this mean for traders?

• How does this affect me as a liquidity provider? 

and

• What are the implications for the wider GMX ecosystem?

But before we get into V2, let’s look back on how GMX V1 pioneered a new breed of DEX, tokenomics and a way of providing liquidity to enable it to compete with the who’s who of DeFi protocols.

Since GMX’s inception on Arbitrum, September 2021 followed swiftly by a second deployment to Avalanche, it has a total volume generated of a staggering $142B of which $224.5M fees have been disturbed to liquidity providers and GMX stakers at a ratio of 70:30. That’s some feat for a community funded project with zero Venture Capitalist backing.  

To illustrate this enormous accomplishment, let’s first compare against all other projects on chain projects over the previous 365 days… 9th with 154m fees generated. Not too shabby aye.

And when we narrow it down to its direct competitors… 1st and it isn’t even close.

How are these fees achieved you ask?

This little known basket of assets called GLP, launched the Arbitrum DeFi season. GLP became the DeFi lego block of Arbitrum with founders racing to build the first autocompounder Stabilize won that battle but unfortunately lost the war. TVL dictates winners and losers… I don’t make the rules. Options trading was built out using GMX liquidity… Bravo Jones Dao and Dopex. Then came the utilisation of GLP as collateral , supply and borrowing markets harnessed the power of this “index” asset. Why so? Well, due to its inherent make up of approx 50% volatile assets (assets that you can freely trade on GMX) and 50% stables, it became somewhat of a stable - ish asset (muted volatility). Finally came yield trading through Pendle, probably the one of the more innovative additions to the GMX wider ecosystem.

See there are two interesting facets to GLP, it’s innately long due to its asset composition and it acts as the counter party to traders. Market goes up GLP price follows and vice versa, now add trader profit and loss into the equation all $17.7m of those trader losses and you have an lp on steroids… well that is, until the market turns bearish and traders become profitable in the extreme… wen GLP drain?

But you see GLP wasn’t the end goal, but it was a game changer propelling GMX into the limelight. Insane APY’s of 130% incentivised with esgmx tokens yet another first for the GMX protocol, unique tokenomics, ensured that liquidity became sticky for the first leg of the journey. 

GLP wasn’t perfect but it served its purpose. It enabled traders to place trades with zero slippage, enter other exchanges routing and facilitate swaps through GLP to provide their customers with the best rates on the market. But it was too inhibiting to the small guy, 0.1% for both opening and closing trades just wasn’t as appealing when the wave of competitors hit the market.

Let’s address the OI caps. GMX’s OI caps are maintained to avoid overexposure to a given side of the market. Ultimately this was a necessary evil due to GMX employing a borrowing rate for both sides of the market. Traditionally, you would see funding rates to incentivise the market reaching parity on the overall OI. All too often one side of the OI was drastically skewed, either by an over emotional market or one mammoth trader. V2 addresses this with isolated markets and funding fees more of that later.

So I hear you ask, V2 why the wait?

10 months in fact (no more 2 more weeks meme). First of all, as we’ve addressed, the GMX team are innovators not copy and paste merchants. This code is brand spanking new and you won’t get to use it without the direct permission of the GMX DAO, as it is housed behind a license wall. 

You see V1 was copied and forked countless times, many ending in some kind of exploit due to the devs not knowing what the code did or what would happen if they added more strings of code to the existing code base. 

Security has always been paramount and a principle held in high regard by GMX and wider contributors to the DAO. Hence the $5M bug bounty on immunefi.com, joint 3rd highest bounty behind Layerzero and MakerDAO. If you’ve a white hat and want to get paid…

So V2, what was the wait all about???

Firstly we’ll look at it from a trader’s perspective. A wider array of tradable assets, made possible by the introduction of isolated synthetic markets collateralised by ETH - USDC. Welcome to the GMX tradable pairs DOGE, XRP and LTC. Oh and did we mention ARB and SOL make an appearance too, but underpinned by their own liquidity pair ARB - USDC and SOL-USDC respectively. You see, with this gentle tweak in theory any asset with a Chainlink oracle has the potential to be listed as a tradable asset. Neat right… 

Combine this with reduced trading fees and you’re starting to concoct the recipe for “CEX Killer”. Let's look at the fees, 0.05% or 0.07% when opening and closing a trade. That's a 30 - 50% reduction which can be further reduced when using our referral code. Why the two differing percentages? GMX are incentivising you to take the opposite side of the market to balance OI, couple this with price impact and funding rates and you have a powerful tool as a trader or even  yield farmers alike.                               

  (left Defillama meta swap, right GMX swap)

What is this sorcery? Positive price impact on simple swaps you say? At the time of writing, there was an imbalance on the ARB - USDC pool heavily skewed to the long side. GMX was incentivising, you good Ser, to re-balance the pools.

(Shout out SniperMonke - The picture of harmony)

Are there any drawbacks to this new liquidity model for you, the trader?

In the main no. But the GMX team has introduced an auto deleveraging function, seen in other protocols such as GNS. Essentially, as the synthetic assets are collateralized by ETH, if say LTC goes on a romping tear leaving everything behind it in the dust the underlying ETH will be such that, there will not be enough underlying collateral value to honour the trade.   

Enough of the mechanisms and assets, does it feel better as a user? Am I guaranteed my position? 

GMX has teamed up with Chainlink to offer traders new low-latency oracles for execution, whereby the oracle system signs prices every block. You want to short ETH at 1850, you’ll get 1850. With this, comes the introduction of limit/stop orders that can be placed at the time of trade. Huge upgrade from V1 where you had to first place the trade then amend once open. The only way you’ll know and feel the difference is by trying it out. The GMX team is always open to constructive feedback to improve user experience. Without users there is no GMX.

Trading strategy options available through V2

Under V1 when longing with any compatible asset via GLP your profit would be paid out in the asset that you long. Similarly when you short an asset, you had the choice to be paid out in any of the available stablecoins within GLP.

With the move to synthetic isolated markets collateralised by ETH/USDC and the introduction of funding rates it’s opened up new opportunities for traders and yield farmers alike on GMX.

First we’ll look at pair trading three that spring to mind are:

• Long/short pair trading 

• Delta neutral pair trading 

• Mean reversion pair trading

In order to carry out these three pair trading strategies in the traditional way you would need to open two trades simultaneously. Well, with the advent of synthetic assets collateralised by ETH/USDC you can simply use ETH as your trading collateral short, say, LTC, to replicate the same outcome. We won’t go into the finite detail on how to execute these strategies, just know those opportunities exist and you can take advantage of them.

Want to be paid for holding the opposing side of the market skew? Well utilise, positive funding rates and skewed OI pairs with less than 1 times leverage. Yes you heard that correctly you can leverage trade with less than 1 times leverage. Be careful though, funding rates are variable and can change from positive to negative in the blink of an eye. 

What’s in it for liquidity providers? How do I even provide liquidity to V2?

GMX has created the GM token to represent your liquidity provided to an isolated market. Synthetic markets, LTC, XRP and DOGE are collateralised by an ETH/USDC liquidity pool with all other tradable assets collateralised by their base asset. 

Choose whether to supply the stable or volatile asset to any GM liquidity pool. Look closely at the underlying dollar value split between assets. Is there a skew? GMX will incentivise you to provide the asset to narrow the difference and bring some balance to the pool.

You can think of GM pools as individual GLP asset baskets when it comes to understanding the dollar value of each individual GM pool with one added benefit… fees are auto-compounded in real time. No more waiting 7 days for an updated APR.

So you’ve had your fill of fees and want to exit the GM pools. In order for GMX to maintain the needed balance in the system when you sell your GM tokens, in return you will receive both assets at the same ratio as the GM pool you are withdrawing from.

$67,760/ $151,203 = 0.4481

$3091 / $6898 = 0.4481 

 So final thoughts…

V1 was a game changer, no one can deny that. In the 10 months it has taken the team to iterate to V2, numerous perp dexs have sprung across the board. Added competition you may say…? Competition is always great for a sector, as it pushes innovation as a whole to the next level, but GMX has a first mover advantage and competitors are incentivizing trading not the underlying component, liquidity. Liquidity is king in this space. You want to be the Binance of on chain perps? You need liquidity.

GMX’s previous liquidity incentives through esgmx are all out in the ether and they still hold C.450m AUM, now imagine if the Arbitrum Poster Boy applies for an Arb grant???? V2 with Arb incentives = game over. V1 was great but unfortunately not scalable. V2 could be greater, if it can see the adoption V1 saw. When other projects start to harness the power of V2, just like MUX has, we will see the cascading effects across the Arbitrum ecosystem and beyond. Remember Arb season round one? Well round two is just around the corner. WEN GM VAULTS, WEN ETH MAINNET?  

If you enjoyed the ramblings of a currently under the weather chap or any of our previous content, drop a follow over on our Twitter account @CryptoRefCodes it would be much appreciated and if you would like to be captivated by the musings of a couple of "normal" blokes, subscribe to our TG group

Until next time folks… Peace and love