A VC would gladly pay you $6 to buy a five-dollar footlong.
Why? Because after he’s wasted $1m of Subway’s money, he can say, “You see how popular that sandwich is? Your user base grew by 1 million people. Bet you could sell these loyal customers a sub for $7.50 now.” Congrats. You’re now a venture capitalist.
It’s no different for blockchains - you’ve gotta spend money to make money.
For those of you who were around in 2021 (please take me back), you’ll fondly remember the ecosystem incentive programs of yore. VCs were playing a game of chicken with each other to the tune of hundreds of millions of dollars. All in the pursuit of users.
In Q3 of 2021, it seemed like there was no end to the money faucets for these VCs.
- First, Avalanche Rush kicked off a $180m stimulus campaign
- Followed closely behind by $100m from Celo
- $300m from Harmony
- $800m from Near
- And $314m from Fantom
These liquidity incentives were an afterburner up the ass for these ecosystems. Up-only market conditions may have fanned the flames, but VC money was the gasoline.
While you may think, “OK, just ape the native token when these things come along,” this isn’t always the best strategy.
These programs are generally funded with the ecosystem’s native tokens, inducing a natural sell pressure. So, how do you profit from these incentives? Generally:
- Farming - farming and dumping the crazy APYs was the name of the game for 2021 incentive programs
- Ecosystem tokens - sometimes there are beta plays to be made on the top eco tokens (e.g. Pangolin for Avalanche)
And while today we’re far from the pent-up mania that was the 2021 bull cycle, we just got our first taste of what incentive programs may offer this time around. Were you paying attention, friends?
Arbitrum’s Short-Term Incentive Program (STIP) was our first real taste of what incentive programs may look like over the coming months and years. The STIP may come off as an experiment into how incentives are distributed, but this one was primed to achieve a few objectives: supporting the growth of the network, experimenting with the program, and finding new models for grants and incentives. The total allocated budget was $50M. There were four grants available:
- Beacon Grants: up to 200K ARB; require the projects to be live for a minimum of 2 months on Arbitrum and reach either $1.5M TVL or $2M 30-Day cumulative Volume.
- Siren Grants: up to 750K ARB; require the projects to be live for a minimum of 4 months on Arbitrum and reach $4M TVL or $40M 30-Day cumulative Volume.
- Lighthouse Grants: up to 2M ARB; require projects to be live for 6 months on Arbitrum and meet $15M Arbitrum Network TVL or $100M 30-Day cumulative Volume.
- Pinnacle Grants: above 2M ARB; require the projects to be live for at least 12 months and reach $30M Arbitrum Network TVL or $200M 30-Day cumulative Volume.
Almost all protocols on Arbitrum were eligible for the grant only if they provided information around how they were planning to use the incentives, sharing contract address details, the grant’s objective etc. The protocols that ended up winning the grants were a combination of established players as well as new players that had shown extraordinary promise in their journey. While some protocols (such as GMX, MUX, Vertex) offered fee rebates, others funneled those incentives for liquidity and volume acquisition via incentives. Most of these incentives were to be used to create a flywheel effect to attract the initial set of users who hadn’t ever used them before and convert them to users who ended up opening positions across different DeFi protocols and getting locked in via composability. But was that really achieved?
This report aims to address some important questions about the STIP:
- Do the old rules still apply this time around?
- Which players benefitted the most from STIP? What were their strategies?
- Incentive programs will return. What should we remember for next time?
To kick us off, let’s look at the strategies attempted by some of the players. Which did you have your money on?
GMX is a perpetual futures DEX built for markets on Arbitrum and Avalanche that offers up to 50x leverage on select tokens. On GMX, you can open long and short positions on the available assets. If you want to learn more about what GMX is, check out this guide created by Grant. GMX currently has the lion’s share of the market on Arbitrum, with the TVL being over half a billion at $539M.
GMX v1 was a showstopper thanks to its GLP token, an index token composed of ETH, BTC, LINK, USDC, USDT, DAI, and FRAX. GLP was considered one of the safest index investments for liquidity providers that provided real protocol yield. In essence, the LPs to GMX make a profit when GMX’s traders make a loss and vice versa. In addition to that, they also get a share of the fees that the protocol generates.
V2 improved on some limitations that v1 had, such as diversified asset pools to protect against directional bias, giving more control to LPs to control their asset exposure, and so on. One of the critical differences was the introduction of GM pools in v2, which consist of a long (ETH) and a short asset (USDC). Despite the improvements, v2 was not as aggressively successful in the beginning as v1; and this was because by the time it was introduced, several competitors had already come into the picture offering perpetual trading services to users. The trading volume on GMX had fallen after mid-2023 for a brief period.
Enter $ARB Incentives
Perpetual protocols in general have largely benefited from the additional ARB incentives with a crazy amount of liquidity flowing in the ecosystem. GMX had received about $13.8M in incentives from Arbitrum via the STIP, of which 10.2M ARB was equally distributed between trading and liquidity incentives.
GMX utilized the $ARB incentives to accelerate driving liquidity to their V2 pools, and this strategy worked its magic from the beginning. Their v2 pools saw a massive rise in WETH deposits across the ETH/USD pool.
Even the number of positions being opened on GMX skyrocketed, with several new ETH longs and ETH shorts positions now being opened. This signaled a rise in trading activity on GMX, which is a natural result of the liquidity pools ballooning.
Why should you care?
- GMX received the highest allocation of ARB incentives (12M ARB), and saw a renewed interest in their v2. While token price saw some attention (+25% vs ARB), the real winners were the GM Pool farmers.
- Short/medium term incentivized growth can potentially set a floor for long-term sustainability (at least, that is the hope), as more users become comfortable with their application.
Camelot is a DEX and a launchpad built specifically for Arbitrum that aims to cater to projects being built on the network. A key distinguishing factor of Camelot over other DEXes is that it supports a dual-liquidity model: one is for volatile pairs (like Uniswap V2), and the other is for stablecoin swaps (like Curve). This is possible thanks to their implementation of a Uniswap V2-like model as well as implementing a Solidly curve to facilitate stablecoin swaps (and/or swaps between highly correlated assets).
In a nutshell, they offer seamless (low slippage and highly efficient) trading of almost all types of pairs.
Camelot is also a launchpad that facilitates and helps bootstrap initial liquidity for new and emerging projects. They have partnerships with most of the leading Arbitrum protocols, such as WINR Protocol, GMX, Pendle, etc. This is made possible thanks to two tools they offer.
- Nitro Pools: These are pools that help new protocols target a very specific group of stakers (based on the requirements of the protocols). This way, the protocol can set requirements like the minimum amount required, duration, whitelist etc. (NB: More on spNFTs later).
- Dynamic Directional Fees: These are the fees that can be set by individual protocols deploying their liquidity pools to suit their own needs, such as growing the size of the pool (lower fees) or slowing the growth of the pool (higher fees). Furthermore, the V3 directional and dynamic volatility fees adjust according to the market conditions (i.e., lower fees during low-volume days and vice versa).
The introduction of STIP incentives drove massive volumes to Camelot. The protocol received 3.09M ARB (i.e., $3.553M) as part of the STIP. You can see in the data below that cumulative volume roughly doubled from $3.5B to $7B in under three months. This was due to the fact that Camelot allocated 3.09M ARB for liquidity incentives, which incentivized deeper reserves for trading activity.
That said, Camelot is not the biggest DEX on Arbitrum (lagging behind Uniswap in terms of the amount of swaps and market share). However, it’s only slightly less than one-half of Uniswap v3 in terms of TVL.
The DEX allocated most of these tokens to incentivize their partner pools (with partners like GMX, Pendle, Radiant, Dopex etc.).
Camelot V3 offers dynamic volatility fees, i.e., they adjust the fees on their pairs according to the market conditions on the DEX. Moreover, it offers pool creators the ability to set different parameters for fees based on the buy/sell activity of a particular asset in that pool.
The native token of the protocol is $GRAIL, which has an extremely limited supply of 100K tokens overall. In the public sale, 15% of this supply was up for grabs. Of this, 5% was distributed via xGRAIL, and the remaining 10% via GRAIL. The current circulating supply of GRAIL is 16K, with the remaining set to be released over a period of 3 years.
When you stake GRAIL, you earn the escrowed xGRAIL. These tokens can then be allocated to what are known as plugins (which are essentially contracts) where users earn a share of the protocol’s revenue (22.5% to be exact). A major part of these earnings are distributed to liquidity providers (60%).
Why should you care:
- Camelot is the biggest Arbitrum-native DEX and one of the go-to protocols for new launches in the ecosystem.
- Camelot received a large allocation of ARB incentives when accounting for their market cap (#3 overall in Grant/MC). This coincided with $GRAIL appreciation of >50% vs ARB. Perhaps this is a correlation we should watch closely (hint, hint).
MUX is another perpetual protocol on Arbitrum that offers zero price impact trading, leverage up to 100x, and no counterparty risks for traders. Its key defining feature is that it’s the first protocol to offer multi-chain unified liquidity across all the deployed chains.
One of the biggest issues in a multi-chain world is the fragmentation of liquidity, and this can especially be pernicious for perpetual traders since leverage trading necessitates DEEP liquidity.
This is where MUX’s “Universal Liquidity” mechanism helps. It works in a simple stepped process:
- When an order is placed for a long, the “broker” checks for the total liquidity across all chains (such as Arbitrum, Avalanche, BSC, and Fantom).
- It checks if the available liquidity from the total liquidity can fulfill a user’s order. If it can, it proceeds to fulfill that order.
Just like GMX v1, MUX also implements a counterparty to traders, the MUXLP - a multi-asset pool that comprises blue-chip assets and stables. It works similarly, too - when the traders are in profit, the pool is in loss, and vice versa. Currently, it comprises ETH, BTC, ARB, AVAX, OP, BNB, and more. More on MUXLP composition here.
Since there are caps to open interest for each of the markets, MUX ensures that traders are able to realize their profits.
MUX received a total of 6M ARB tokens (i.e., $6.9m as part of STIP). From the total incentives they received, they allocated 55% as grant fee rebates in GMX V1, V2, Gains, and MUX native pools. This led to a rise in growth since November, with both daily and cumulative volume skyrocketing.
Why should you care:
- The 7-day volume on the protocol peaked in December, but since then has seen slower growth.
- MCB underperformed expectations, despite the large grant value. The winners in the end were the farmers, collecting huge amounts of ARB rewards from the MUXLP.
Radiant aims to be an omnichain liquidity hub for major protocols creating a seamless multi-chain experience for both lenders and borrowers. They achieve this by enabling deposits across multiple chains (ensuring that borrowers can seamlessly borrow cross chain too). They launched their v1 on Arbitrum because it offered lowered transaction costs without compromising on security.
Radiant v2 is a step forward in unlocking the full potential of DeFi and “ushers in the generation of DeFi 3.0”. The core objective of v2 is to a) create RDNT as an omni-chain token, b) implement a linear scale for users who want to exit their RDNT positions, and c) increase the vesting time.
Radiant has introduced a liquidity model that rewards long-term contributors to the protocol. As part of it, users must lock their dLP tokens to benefit from the RDNT emissions.
This was implemented thanks to their unique proposition of dynamic liquidity provisioning. This ensures that RDNT emissions are proportional to the qualified deposits for dLP, making dLP yields higher and incentivizing users to participate even more.
The Flash Loan Attack on Radiant
Radiant Capital recently suffered from a flash loan attack of $4.5M where the attacker managed to exploit a “rounding issue” in the codebase. The attacker exploited the time window of when a new market was created. This led to the pausing of all markets on Arbitrum as the protocol did not want the issue to penetrate any other markets and/or lead to loss of funds for the user. Since then - as of writing this piece - the protocol has resumed lending/borrowing on all markets. The users who were wrongly liquidated because of the incident are being compensated.
While the protocol saw a small dip in deposits during this period, they will resume as Radiant is a strong DeFi money market on Arbitrum. That said, users are advised to exercise caution.
Why should you care
- Radiant’s v2 changes the game around protocol-native token emissions. Most of the protocols with their own tokens often end up increasing their emissions in an attempt to acquire mercenary capital, which later gets dumped on by the LPs anyway. This is where Radiant stands apart.
- Radiant distributed 2.8M ARB tokens (i.e., $3.279M as part of STIP). A significant portion of this was reserved for dLP Locks for the 6 to 12 months duration. Other incentives were reserved for GMX v2 BTC and ETH GM lending pools, Camelot v3 and Dopex V2 RDNT/ETH liquidity pools, and some for PlutusDAO incentives.
- Radiant’s token price underperformed lower-market cap lending protocols such as Lodestar and Silo following STIP disbursements, perhaps showing that speculators preferred higher beta opportunities.
Pendle offers you the ability to split your yield-bearing tokens (like Liquid Staking Tokens) into two assets. The PT (Principal Token) and YT (Yield Token), in this case, are redeemable at the end of the maturity period for the original asset. For instance, if you staked 100 DAI and got a 5% yield on that - you get 100 DAI (from PT) and 5 DAI (from YT) at the end of the maturity of the period. Pendle allows for the yield to be claimable in real-time.
- Principal Token (PT) — which will mature into the yield-bearing asset upon maturity. For those familiar, this is equivalent to a zero-coupon bond in TradFi. You remember those $50 bonds that your grandma gave you at your baptism/bar mitzvah/blót? The ones that weren’t worth $50 until you were 25? Same idea here.
- Yield Token (YT) — which gives the holder the right to yields generated by the underlying asset until maturity. This is equivalent to coupon payments on a bond.
The prices of the PTs and YTs are determined by market demand and supply up until the time of maturity. This allows market participants to speculate on the rise and fall of interest rates and yields for different assets in DeFi.
Users also get to speculate on the yield-bearing tokens by purchasing YT at a discounted rate. For profitability in the case of YT, they look at the implied APY on Pendle, which is the expected APY of the YT at the time of maturity. If they expect their overall returns to be higher than the implied APY, then they buy the YT token at the discounted price. The implied APY, in this case, is a direct correlation to the underlying asset’s APY being offered on its own platform.
Now one might ask that the implied APY would be - in most cases - closest to the APY that the underlying project is offering, correct?
That may be true. But in the case of LSTs, for example, this can be different because there can come a time when the LST in question gets depegged or the protocol issuing the LST undergoes an isolated slashing event which impacts the yield of both the YT and the PT. In the case of LPs of lending/borrowing protocols, the yield would fluctuate based on the utilization rate of the underlying protocols and of the assets in question.
If the actual APY turns out to be greater than the implied APY at the end of maturity, the YT holder ends up receiving significantly higher returns than someone who is just holding the underlying asset (such as stETH itself).
Any user can attempt to time the market in a way that they receive yields from when buying both the PT and YT tokens at different times, depending on when the implied yield on either is low.
Lower implied yield = Expensive PT but cheap YT
Pendle’s AMM uses Notional Finance’s AMM model as its baseline to allow users to trade their PT and YT tokens. It is a fascinating AMM that - based on market demand for PT and YT - sees fluctuations in the Implied APY. The higher the demand for PT, the lower the implied yield because the market is - effectively - full of participants shorting yield on that token (thus, the value of the PT is higher).
Why should you care:
- The PT tokens on Pendle can be thought of as fixed-income assets. This is especially true for LSTs because the principal on those assets doesn’t generally change. Hence, a user can easily buy an LST at a discount (such as buying 1 stETH for 0.9 stETH) and sell it for a profit at the end.
- Pendle was offering 2M ARB tokens (i.e., $2.3M). It offered four-figure Long Yield APYs on select trading pools! This was thanks to their allocation of 1.1M ARB for Pendle pools and the remaining for increasing trading activity and volume and for using Pendle partner protocols.
Frax is a decentralized stablecoin that started its journey as a fractionally-backed stablecoin. Additionally, Frax also has an ecosystem of its own, which has another stablecoin (FPI, which is pegged to a basket of consumer goods), frxETH and sfrxETH (which is a replacement to WETH and a liquid staking token), FraxLend (money market for ERC-20 tokens), FraxSwap (AMM with a TWAMM for large trades), and FraxFerry (a protocol-native application that allows you to bridge over FRAX tokens).
While the protocol began its journey as a part-collateralized and part-algorithmic stablecoin, it pivoted to becoming a fully collateralized stablecoin in early 2023, seeing the downfall of several algorithmic stablecoin projects.
- Frax V1 was a fractional-algorithmic stablecoin where a part of it was backed by collateral, and a part of it was fractionally managed via the FXS shares. The plan for v1 was to fully stabilize the USD peg of the stablecoin algorithmically, allowing for rapid expansion.
- Frax V2 introduced the ability for the protocol to enact monetary policy to perform open market operations (governments use this to buy and sell securities in the financial markets). These operations would keep the FRAX peg stable. This is called the Algorithmic Market Operations (AMO) controller. Unlike Central Banks, however, AMOs do not arbitrarily print FRAX to keep the peg stable.
- Frax V3 aims to utilize AMOs and its subprotocols like FraxLend and FraxSwap to fully collateralize the FRAX stablecoin. The collateral, in this case, is RWAs that are approved by the Frax Governance Module (frxGov).
For the RWAs that help collateralize the FRAX stablecoin, the protocol currently considers those that yield close to the Interest on Reserve Balances (IORB) rate of the Fed. Thus it currently only considers short-dated US Treasury bills, Federal Reserve Overnight Repurchase Agreements, USD deposited at Federal Reserve Bank master accounts, and select shares of money market mutual funds.
One of the critical offerings of Frax is the FPI - which is a stablecoin that is pegged to a basket of real-world consumer items - defined by the US CPI-U average. This is made possible using the Chainlink oracle that works in coordination with Frax.
This effectively means that FPI is pegged to “inflation”. Thus, as the CPI increases, so does the FPI making it - more or less inflation-protected. It tracks the 12-month unadjusted inflation rate.
Frax received 1.5M ARB tokens (i.e., $1.725M) as part of the STIP. These incentives were distributed across sfrxETH, frxETH, FRAX, and FPI. Thanks to this, the protocol saw an uptick in growth - with the 7-day MA (moving average) volume hitting a $10M mark and 542 DAUs (Daily Active Users).
Why should you care:
- Frax offered its 1.5M ARB (that it is receiving via phased disbursements) for three of its assets: sfrxETH, frxETH, FRAX, and FPI.
- Frax aims to offer highly consistent returns when you stake the FRAX tokens (to get sFRAX). This is because when users stake their FRAX, it is deployed to yield an APY that is as close to the IORB as possible. This ensures consistency in the returns for users.
- Frax’s FXS underperformed ARB in the timeframe following STIP. A couple reasons: a) One of the reasons is likely that Frax is not an Arbitrum-native protocol and b) FXS has a very large market cap compared to other entrants
It is rumored that truly understanding women is one complex problem that we haven’t been able to solve. The only problem that comes close to that is reducing impermanent loss.
Despite the best efforts of Uniswap V3’s concentrated liquidity model (and later that of Sushiswap, Pancakeswap, and Solidly), the problem of IL still remains. Why? Because to counter IL effectively, you need to hyper-actively manage your liquidity across concentrated liquidity offering DEXes. That creates a very poor UX for users and LPs to DEXes. Thus, we need something better.
Gamma attempts to solve all of these problems through its vaults, which support a plethora of different platforms and blockchains. The core problem that Gamma solves is that it actively manages liquidity in a way that it generates the highest fees for your position.
The protocol employs several different strategies for LPs that deposit their assets to the protocol. Some of these include:
- Dynamic Range (Wide/Narrow): This is a passive rebalancing strategy that gets automatically implemented when certain thresholds in LPing are hit. This is specifically for long-term LPs who have deposited their liquidity across a wider price range. The narrow range, on the other hand, is suited for short-term LPs and caters to their liquidity preferences.
- Stable: These are for stablecoin pairs or highly correlated asset pairs. The accrued fees, in this case, are compounded back into the positions on behalf of LPs.
In addition to this, Gamma also supports the creation of custom strategies. The protocol currently supports all major L1 and L2 blockchains and also works on major DeFi protocols like Uniswap, Thena, QuickSwap, Camelot, SushiSwap, PancakeSwap, and other such popular DEXes.
The $3.4M Hack on Gamma
Gamma recently suffered a hack on its deposit vaults that reached a cumulative of $3.4M. The Gamma team was quick to respond and shut down its deposit vaults immediately. They also reached out to the attacker to negotiate to get the funds back. The team has assured users that the protocol’s code will undergo a third-party code review to ensure that the error/bug is mitigated before the deposits are reopened for users.
Why should you care:
- Gamma sits atop major DEXes and pools that offer users the ability to concentrate their liquidity. Given it offers active management of concentrated LP positions, it is at the center of a growing and promising sector within DeFi.
- The native GAMMA token can be staked to partake in a portion of the revenue that the protocol generates from managing liquidity across various applications. They are also introducing multipliers to offer greater returns to users who stake their GAMMA tokens for longer.
- Gamma protocol received 750K ARB tokens (i.e., $863K) via the STIP program. It offered the boosted yield from that to users across various pools on Arbitrum; all of these incentives were purely given out to LPs. Some of the pools on Camelot were offering as high as 120% APR.
- Like Frax, Gamma doesn’t spring to mind when thinking about Arbitrum protocols, so traders faded this one as well. For future incentive programs, pay attention to native products first if you want to snipe a token.
Lodestar Finance is an algorithmic money market protocol on Arbitrum that offers users the ability to lend and borrow their assets. Additionally, Lodestar offers liquidity without creating taxable events, leverage trading strategies, and unlocks the liquidity of yield-accruing assets. The protocol was initially launched to support MAGIC and DPX tokens, because while they had strong communities on Arbitrum, there were no money markets servicing them.
The core features of Lodestar allow users to:
- get access to liquidity without creating a taxable event (by collateralizing their assets) and
- open leveraged long and short positions.
In addition to this, the protocol also allows users to collateralize positions with liquid-staked assets such as drMAGIC and plsDPX.
These long positions are opened when a user deposits collateral to mint USDC and uses that borrowed amount to buy more of the token whose price they expect to increase. Similarly, to open a short position, they sell the asset that they feel will decrease in price.
When someone lends their assets to Lodestar, they receive a representative token in return that rises as the assets accrue interest. For borrowers, the amount of assets they can borrow from the collateral depends entirely on the collateralization ratio (CR).
Lodestar takes inspiration from Compound’s jump rate model and utilizes five models with varying parameters depending on the type of assets being deposited in the protocol.
- Model 1 (Minimal Risk): This is for USDC as it is one of the most widely used stablecoins in the market. The interest rate parameters are at par with that of other money markets in DeFi.
- Model 2 (Low Risk): Here, ETH is considered the native asset; ETH deposits enjoy a higher rate of return, and borrowers can get more from the collateral they deposit (this has a collateral of 80%).
- Model 3 (Medium Risk): These are for blue chip assets such as wBTC, plvGLP, DAI, and FRAX. Here depositors to the protocol earn an increased yield.
- Model 4 (High Risk): This model is suited for assets like MIM that are backed by volatile assets. This model also covers USDT. The rate earned on lent collateral, in this case, remains close to that on USDC deposits. Here, borrowers pay a much higher interest rate if the pool’s utilization rate hits 80%.
Why should you care
- Lodestar is one of the leading money markets on Arbitrum. Thanks to its introduction of different interest rate models for varying assets, it caters to a wide spectrum of users.
- Lodestar received a grant of 750K ARB tokens (i.e., $862K) via the STIP incentives, out of which 95% were reserved for market participants (i.e., lenders and borrowers). In terms of market cap, Lodestar’s ARB grant was the largest of any project.
- Lodestar was a great pick for an outperforming alt, given it: a) Received a large amount of ARB given its small market cap and b) Is an Arbitrum-native project
Jones DAO is a yield aggregator protocol that manages users’ portfolios via institutional grade management strategies. Additionally, it unlocks the liquidity for assets that users deposit with the protocol, thereby offering a much-enhanced user experience.
The protocol generates returns for its depositors through its Vaults. There are three different types of vaults that the protocol Offers.
- OpFi Vaults: These OpFi vaults and the tokens deposited in them are used to write/sell options and take advantage of inefficient options pricing.
- Metavaults: These are LP vaults that perform best during periods of peak volatility. They also leverage the price action on the tokens that are LPed to maximize profits. There are two types of Metavaults available: DPX-ETH and rDPX-ETH.
- Advanced Strategy Vaults: These are the vaults that offer users superior USDC.e yield and amplified GLP yield.
To unlock users’ deposits in Jones DAO vaults, the protocol issues jAssets - which are representations of user deposits. These assets have their own liquidity pools enabling anyone to exit their jAsset position by exchanging it for a stablecoin in the market.
Their core offering is the vault strategies. Users provide LP tokens in the Metavaults, which are then staked in Dopex farms, and the yield gets generated in DPX and rDPX tokens. A portion of this yield is used to buy options, and the remainder is auto-compounded in the LP farm. Thanks to this, the protocol is able to offer high yields during bull and bear markets.
The jGLP and jUSDC vaults offer auto-compounding yields for users. When users deposit their GLP (or any of the basket tokens of GLP) and/or USDC.e into the vault, it is used to borrow USDC collateral from the jUSDC vault to mint more GLP. This effectively opens a levered long position on GLP.
This way - the jGLP vault delivers high yield to users while the jUSDC vault delivers the borrow yield.
Why should you care
- JonesDAO received 2M ARB tokens (i.e., $2.3M) as part of STIP. It allocated 1.3M for jGLP and jUSDC, 350K for wjAURA, and 350K for MV2 and smart LP strategy. The TVL in these pools saw a major inflow right after the incentives were announced but seemed to plateau after.
- JonesDAO’s grant/MC was the second-highest in STIP-1
- JonesDAO was also great pick for an outperforming alt, given it: a) Received a large amount of ARB given its small market cap and b) Is an Arbitrum-native project
Dolomite is a DeFi chad aggregator combining the strengths of a money market protocol as well as a DEX. It offers users the ability to take out loans, do margin and spot trading, and utilize other financial instruments. Dolomite is a fork of dYdX and has two modules, one which is immutable (i.e., cannot be changed and the other that can be changed). This is the core differentiator of the protocol.
The modularity of the protocol ensures a certain part of it remains fluid and adaptable to change, while the other part remains immutable and beyond the scope of any governance-inspired changes. Thanks to this modular structure, it can offer a suite of different products and services to its users. And that’s exactly what makes Dolomite unique.
Some of the features that can be built on Dolomite are:
- Trading between two users
- Trading versus a liquidity pool
- Trading within Dolomite itself
- Opening overcollateralized positions on Dolomite
All of this is made possible thanks to its unique adoption of capital efficiency which effectively implies that all the deposited internal liquidity in Dolomite can be used in multiple yield-generation methods. When you deposit your assets to Dolomite, they are counted as Dolomite Balance. This effectively makes the entirety of your deposited assets to be used across the entire suite of products that the protocol offers. This balance is considered “virtual” because none of this liquidity is used on-chain; this enables spot trading on the protocol itself. The main service they offer to users is the ability to borrow assets. For this, users can deposit any asset of their choice and open a borrow position. The core difference from Aave is that users can open multiple Borrow Positions using their Dolomite Balance by supplying different types of assets.
This means you can use your Dolomite Balance to open multiple borrowing positions without having to interact with multiple pools all at once. More details about borrowing on Dolomite can be found here.
Dolomite received a 1M ARB (i.e., $1.15M) grant as part of the STIP to drive liquidity and growth to the protocol.
These incentives are being utilized to aid the protocol to underwrite loans for larger positions - enabling more strategies for the users. This incentive program rewards those who hold USDC, ETH, WBTC, ARB, LINK, or GMX on the protocol; they are rewarded with oARB tokens. These oARB tokens are then vested into ARB tokens at a discount to the prevailing market rate.
Why should you care
- The protocol distributed 54,000 oARB weekly to users with 38.5% going to USDC LPs, 24% going to ETH LPs, another 24% for WBTC LPs, and so on. The exact distribution can be found on the Rewards page on Dolomite.
- The total deposited liquidity to the protocol has seen a hockey-stick growth ever since they received the STIP rewards from Arbitrum. Thanks to these incentives, the growth has remained almost steady since then.
- In addition to the oARB incentives, the protocol is also offering as high as 47% supply APR on some of the coins (USDC) making Dolomite a strong contender for any degen investor’s portfolio.
STIP retrospective and how we’re playing future incentive programs
For the STIP, yields were flowing and farm-and-dumpers were eating well. But, were there other opportunities to be found?
Arbitrum’s native token, ARB, appreciated modestly from 1-3 months after the STIP announcement. Notably, however, it underperformed ETH (+32% over 3m), a natural benchmark. As with other native tokens from prior incentive programs, buying ARB upon the announcement did not outperform the broader market.
This makes sense if you remember that STIP distributed ARB to dozens of protocols to use as incentives. And what did farmers do with this ARB after earning it? They dumped it. That’s 50m ARB worth of sell pressure.
Which beta plays outperformed?
So, what about Arbitrum alts? If we look back at the performance of all the grant-winner’s tokens, we can draw some conclusions.
First - let’s consider the following timeframes for our analysis:
- Time frame between the STIP announcement and grant votes. This would probably only work if you had a crystal ball and knew which protocols would win the grants at the time of the announcement (and if this is you, please slide into our DMs).
- Time frame between the grant vote and 1 month post-vote. This is probably the most relevant, since you would know which projects had won the grants by then.
- The entire time frame from STIP announcement to 1 month post-vote. This, again, would require some guessing, but gives us the longest time frame to consider. Perhaps you have a really strong hunch based on some early governance chatter?
Second - how to standardize the grant amounts? A 1m ARB grant means a lot more to a small protocol than it does to a larger one. We will consider the following standardization techniques:
- Grant / Total Value Locked. This will compare the grant value to the TVL of the project.
- Grant / Market Cap. This will compare the grant value to the current circulating market cap of the project’s token.
- Grant / Fully Diluted Value. This will compare the grant value to the market cap of the project’s token if the tokens were all in circulation.
Comparing these, we arrive at the following correlation table (where values close to 1 or -1 are significant). Note that we compared the price appreciation of these tokens against that of ARB - we are looking to outperform ARB, after all.
We see that our second column (Grant/MC) shows the highest correlation by far. This means that if we wanted to make an educated guess about which tokens ought to perform well, filtering by Grant/MC is a great place to start.
To visualize this correlation, we’ve graphed the middle case (where you track the token price appreciation vs ARB from the time of the vote to 1 month post-vote) as it is likely the easiest to reproduce for future incentive programs. We see that while most Arbitrum ecosystem tokens ended up outperforming ARB, the winners: had high Grant/MC values and were Arbitrum native. In fact, Arbitrum-native protocols outperformed ARB by an average of 14%, while non-native protocol tokens only outperformed by 1%.
While Arbitrum’s STIP allowed the L2 to temporarily steal the spotlight away from the consistent launches of new chains, was it “worth it?”
Incentive programs of 2021 propelled chains like Avalanche into the vernacular, minting portfolio all-time highs for early rotatooors and catapulting Avalanche’s TVL ~10x even when denominating in their native token. In comparison, Arbitrum’s growth during their incentive period has been, well… muted by the general market conditions.
It’s true that Arbitrum’s TVL increased since the announcement of STIP, but when compared to the overall market, STIP doesn’t appear to have had a massive impact. This is likely due to the Player vs Player (“PvP”) nature of the 2023 market. While the second half of 2021 saw a huge inflow of fresh stablecoins (56% increase), the second half of 2023 netted to basically flat flows. 2021’s influx of new capital to the market catalyzed the incentive programs at the time, leading to gargantuan opportunities.
All this being said, it’s encouraging to see Arbitrum attempt to take a more-measured approach to doling out their millions of dollars worth of tokens. While this can be an arduous process, it shows the maturation of the industry that a DAO is able to decide the worthiness of protocol proposals as opposed to a single player picking winners. The STIP process can definitely be improved; in fact as part of its next Long-Term Incentive Program (LTIP), Arbitrum DAO has already decided to make changes such as stationing Application Advisors to work closely with grant recipients and guide them with the necessary feedback on best utilizing the incentives and more. The LTIP aims to improve on the flaws of the STIP as well as serve to collect feedback prior to the launch of an year-long incentive program by the Arbitrum DAO. One of the problems observed with STIP was that several protocols missed the deadline for it and/or were intentionally waiting for the upcoming round. Some of the improvements in the upcoming round include:
- Setting up a Council that selects which protocols deserve the incentives
- Setting up impartial Application Advisors that will provide detailed feedback on how applicants can improve their applications
- More innovation on incentive structure will be allowed
These are some of the significant improvements suggested for LTIP. We’re super excited to watch the program unfold and the projects that come out of it.
The hope (for us) is that these future incentive programs will coincidently coincide with a well-capitalized market, where one investor’s gains are not coming at the expense of another’s loss. If the market conditions are more favorable during future implementations, we are very bullish on Arbitrum ecosystem tokens.
If this occurs, be on the lookout for the fastest horses and the juiciest yields. Generally these will be the native projects that are able to secure lucrative grant values with respect to their Market Caps.
Newer ecosystems will also be important to watch. They have fewer bagholders, less price history, and generally use new gas tokens.
Prime candidates for future incentive programs: new L2s
As the playing field gets more and more crowded, expect rollups to incentivize activity on their networks in order to stand out and attract users. VC-involved L2s such as zkSync, Linea, and Starknet need eyes on their product if they ever want to dump their bags on retail.
Be careful of unlocks (check TokenUnlocks) to ensure you aren’t becoming exit liquidity and ape as appropriate.
Mantle is somewhat of a dark horse in this category. It’s a DAO-operated L2 that has amassed a frankly hilarious warchest. Mantle is already running incentive programs that have been quite successful, growing it to become the sixth largest Ethereum L2, and we expect the gravy train to continue. Be on the lookout for larger incentives when the market gets frothier.
Prime candidates for future incentive programs: new L1s
This cycle will likely debut new entrants to the field of crypto with the advent of Move-based chains (such as Sui and Aptos), parallelized EVM chains (Sei, Monad), and modular chains (Celestia, Eclipse, Dymension, AltLayer, lots more). Each of these narratives is lauded by the “smart money” in this space, so it’d probably behoove us to help separate this smart money from these fine folks once they turn on the money faucets.
The 2021 winners rode the “ETH Killer” narrative to multi-billion dollar valuations. Do you think this time will be different?
Remember, VCs have already dumped millions of “real” dollars into these products. They won’t mind spending their token allocations (or convincing a DAO to do it for them) to pump their bags before exiting.
Regardless of who’s paying, we’ll have our hands out. Here’s the playbook we’re running:
- Farm the yields - safest. Beware of rugs, but generally the safest to farm-and-dump the free incentives.
- Buy the beta plays - riskiest. Determine your strategy (such as native + Grant/MC, mentioned above) and place calculated bets on protocol tokens that meet your criteria. Have an exit plan.
Here’s to a hopefully-bullish 2024, friends.
DISCLOSURES: BrownBacon, the author of the report, has exposure to GMX. 563, who assisted with this report, has exposure to Pendle, Frax, and GMX. Please be aware of these biases while reviewing this report.