2023 proved to be a momentous year for network activity, so naturally protocol innovation and growth followed suit. Going into the year, uncertainty around CEX solvency pushed traders to utilize decentralized derivatives, which triggered unprecedented growth for protocols like dYdX and Synthetix. Uniswap continued to innovate on its AMM model through the introduction of Uniswap V4, and Curve switched up the DEX landscape through the launch of its own stablecoin, crvUSD. After the Merge, a surge in Ethereum staking triggered an explosion in LST-fi protocols eager to capitalize on exotic yield offerings, ranging from LST-collateralized stablecoins to interest rate derivatives. We can’t turn a blind eye to the hype seen across consumer-facing dApps, DEPIN, or gaming either. Without further ado, let’s unpack 2023’s protocol highlights. 

Perpetuals

The onchain perps landscape proved to be one of the more bear market resilient sectors throughout 2023, with volume holding strong and native token holders often earning “real yield” from trading fee revenue. Smart contract developers have clearly noticed the cash cow nature of perps protocols, which is made evident by dYdX’s declining volume dominance as newer protocols within the vertical grew in relevance throughout 2023. We believe there is still a large runway for growth given the fact that less than 2% of crypto perps volume occurs onchain, and that the entire pie for CEX and DEX perps volume will continually grow over the coming years. These onchain trading venues are still highly reliant on token incentives to attract users, but we believe the UX gap between CEXs and DEXs has narrowed tremendously over the past year and could serve as a tailwind for these protocols moving forward. We breakdown some of the highlights and trends from 2023 for leading perps protocols below. 

dYdX

In October, dYdX launched its V4 exchange, and trading started in November. This version marked a shift in the protocol’s technology, transitioning from an Ethereum-based zkRollup to a standalone blockchain built on the Cosmos SDK. The key change included decentralizing the order book storage, moving it from centralized servers to being stored in validator memory. Additionally, fee revenue was rehypothecated to provide yield to validators and DYDX token stakers, whereas before, returns were allocated to the centralized company that created the technology. These changes enhanced the protocol's decentralization and made the fundamental value accrual mechanism for the DYDX token significantly more robust.

More nuanced innovations that came with V4 were its novel MEV mitigation through collaborative block building, further asset listings, and a seamless bridging experience through CCTP and Noble. Additionally, the frontend was open sourced, and necessary oracle data used for funding rate calculations was internalized through ABCI++, removing the reliance on Chainlink. 

The dYdX team has done a stellar job at adapting the product to facilitate the best possible trading experience and optimize for decentralization. V4 has yet to see the same level of adoption as V3, but the provision of $20M in trading incentives over the 6 months since launch in addition to DYDX token trading rewards will likely help drive volume to the new version over time. V4 trading volume has picked up since its launch just a month ago, although the maximum daily volume of ~$500M is just on par with V3’s worst days. 

Gains Network

Gains Network had a strong start to 2023 after deploying gTrade on Arbitrum in addition to its Polygon deployment, but has since taken a backseat to other perps protocols when it comes to volume/fee/TVL growth metrics. Unfortunately for Gains Network, it does not enjoy a massive stream of GNS inflation used to incentivize trading activity like many of the other perps projects we will discuss later, so it is already at a disadvantage. Gains Network was also denied an ARB STIP allocation after requesting 7M tokens, which likely played a large role in user volume migrating to other dApps that did receive ARB funding. However, luckily they were able to score 4.5M ARB after a reworked Arbitrum governance proposal passed to backfund more projects that were left out of the initial allocation. This could provide a much needed tailwind for the protocol from now until the end of January 2024.   

Gains Network finally made the jump to a more decentralized architecture in 2023 by establishing governance and grants program processes, thus giving more power/ownership to GNS token holders. The proposals are ultimately vetted by protocol admins, and the votes are not yet enforced onchain, so some more work needs to be done in order to truly decentralize the protocol. The most significant milestone of 2023 was the GNS token redesign, which ultimately increased the amount of revenue that flows to GNS stakers. The development fund and NFT liquidation bots previously earned 16.33% and 14.20% of protocol revenue, but both of which were deprecated with the token redesign and now a vast majority of this revenue flows to GNS stakers. However, this redesign came at a cost to token holders; 4.36M GNS was minted to compensate the development team and original NFT liquidation bot holders. Much of this supply is locked at the moment and will remain a headwind for the protocol as vesting schedules conclude. Additionally, the tokens received by the team were staked and therefore dilute other GNS stakers’ yield. This paired with declining revenue has made the GNS staking yield less attractive when compared to other perps DEX’s native tokens.

While we believe the core development team behind Gains Network is one of the stronger teams in the perps vertical, the project faced an uphill battle in 2023 due to rising competition with newly launched tokens. We do not expect things to get any easier for Gains Network in 2024.  

Vertex

Speaking of shiny new tokens and projects, Vertex took the DeFi space by surprise in the final quarter of 2023 by ousting dYdX as the number one perps platform by volume. The primary catalysts for Vertex were the launch of its token through a liquidity bootstrapping auction, ongoing dual incentives (VRTX and ARB) thanks to the Arbitrum STIP that the project won, USDC.e revenue share with VRTX stakers going live, and an Elixir integration that helped bolster order book depth. Projects like Gains Network and GMX are required to share revenue with LPs in order to ensure ample liquidity for traders, whereas projects like dYdX and Vertex do not have to incentivize LPs in the same way and are better suited to share revenue with native token stakers. Vertex had some very lucky timing between GMX fragmenting its LP and trader user base and dYdX attempting to wind down v3 incentives and get market makers and traders over to v4. The two market leaders ceded ground to Vertex, but it remains to be seen if the volume is sustainable or simply inflated by a low float, high FDV token (VRTX) used for incentives in tandem with ~212k ARB each week. 

It’s quite difficult to predict how this one plays out: On the one hand you have a high quality product in a vertical that has demonstrated strong PMF onchain paired with ~$5.85M of incentives each month. On the other hand, however, you have a highly inflationary governance token with other new perps farms popping up like RabbitX, IntentX, and Aark, as well as incumbents such as dYdX leaning more heavily into incentives as the competition heats up. Many folks are pointing to the dispersion between daily volume and open interest on Vertex, and believe all of the activity is inorganic token farming. We believe this critique is warranted. However, the VRTX staking mechanism increases each user's stake weight every block as long as they do not withdraw any tokens, thus increasing each staker’s share of weekly protocol revenue. If we assume the value of VRTX and ARB incentives holds steady, we believe trading volume will remain elevated through January 2024. One of the most important metrics to track to gauge the overall health of Vertex and its longevity is the total amount of VRTX staked. 10M VRTX tokens will be handed out to traders and market makers every epoch, and what these actors actually do with these tokens (stake or dump) will likely determine Vertex’s long term fate. The thesis behind the token design seems to be that trader’s will continually roll their USDC earnings into more VRTX tokens to compound their stake weight so that staking revenue always outweighs the fees a user has paid to trade on Vertex. Only time will tell if this token design is sustainable, but the tail end of 2023 was certainly huge for the Vertex protocol, which continually notched all-time highs in protocol revenue and TVL throughout the year.

Synthetix

Synthetix began 2023 with an absence of clear product market fit. A lot of their previous experiments, like spot markets and atomic swaps, had waves of strong adoption followed by lull periods of low utilization. Towards the end of 2022, Synthetix tried out a simplified version of a perps product, which saw relatively substantial demand but required some work under the hood. In Dec 2022, they launched Perps V2, using low-latency Pyth oracles, a dynamic funding rate to incentivize net delta neutral open interest, and fees of only 5-10 bps. With that, they now had a product that could heavily compete with other perps competitors. As a result, demand for this product rose quickly, with over $40B of cumulative volume occurring over the course of 2023.

Synthetix’s success can be attributed to a few different factors. First and foremost, great price execution as a result of the oracle based design along with extremely competitive fees make Synthetix an attractive place for traders. With an oracle based design, it’s also important to incentivize LPs by reducing their risk, through decreasing the variance with which traders beat the house, and paying out a high enough yield to justify the risk taken on. The dynamic funding rate does a good jobl at keeping OI neutral and therefore minimizes any potential downside from traders winning big as there will almost always be someone hedging and capturing funding rates through their own delta-neutral position. 

Synthetix also operates as a liquidity back end, meaning that we see front-ends competing to create a sticky relationship with the end user. As front ends were given a fee share earlier this year, which has recently been increased to 20% of all fees generated, these frontends now have an incentive to bring in users, which grows the pie for Synthetix. So far, Kwenta and Polynomial have been the two primary front ends, but a new front end started by Synthetix founder Kain called Infinex is launching and looking to compete heavily on user experience.

There is plenty to look forward to for Synthetix, as they migrate to V3. Under this architecture, Synthetix becomes much more modular to allow for more experimentation at the market level while siphoning risk wherever LPs are comfortable. So, while perps has been the clear case of product market fit, there is going to be more experimentation when it comes to other markets: options, betting, insurance, and much more. The current perps system will be upgraded to V3 and be included in the primary V3 pool called the Spartan Pool. 

The Spartan Council also recently voted on removing SNX inflation, which was paid out to stakers in escrowed format to incentivize liquidity. Given that Synthetix is at a sufficiently mature stage that there is significant revenue already flowing to LPs, and inflation was only 5% due to a very high stake rate, inflation did not make much sense anymore. Inflation in SNX’s history has traditionally been very high and contributed to many people not investing in the token due to either the LP risk or the dilution risk, both of which are now much diminished from even a year ago. This is a huge headwind for SNX adoption and investment going forward, but we will need to monitor the stake rate to see if it brings about any meaningful changes in liquidity. 

In addition, Synthetix is looking to expand across many different EVM chains, using new chains as experiments to see where they can gain the most users and LPs. The experiments will be a part of what they are calling Andromeda, which is the deployment of Synthetix V3 and Perps V3 with only USDC and ETH for LPs. This will help gauge demand from traders for perps on chains other than Optimism, and gauge demand from LPs for assets other than SNX. The first of these Andromeda deployments will be Base, followed by Arbitrum. There are also a few other potential experiments, including an ETH-backed perp on Ethereum mainnet designed specifically for delta neutral stablecoins and protocols like Ethena, and SNXChain on the OP Stack which will be a central hub for governance and SNX LPs. This will make 2024 a very exciting year for Synthetix. 

GMX

2023 marked the introduction of GMX V2 and a falling market share for the project overall. Throughout the year, GMX lost dominance in fees and volume, largely driven by new market entrants and their associated incentive schemes. GMX V2 launched in August to create a better UX for traders and environment for liquidity providers by introducing more appealing mechanisms for them in hopes of better competing with the new exchanges. 

The key updates in GMX V2 include the introduction of isolated pools, allowing liquidity providers to customize their asset exposure and support the listing of long-tail assets. New fee mechanisms were implemented, including reduced execution fees, a funding rate, and a price impact fee/rebate system designed to balance the open interest between long and short positions and mitigate LP exposure to trader PnL. Additionally, upgraded low-latency Chainlink oracle price feeds enhance transaction speed and reduce oracle extractable value. Unlike its predecessor, V2 operates under a business license, making it non-free to fork.

GMX received a notable 12M ARB allotment from the Arbitrum DAO as part of the short-term incentive program to allocate to its users. These incentives started in November and will need to be used by the end of March. They should help drive significant use of the platform until the incentives run out. 

Options

The onchain options vertical experienced a revolution in 2023, with many projects tweaking their mechanisms, introducing new products, and implementing version upgrades. Options protocols have historically struggled to offer a user experience comparable in quality to centralized providers, with many projects having operated rudimentary options AMM models that provided unsatisfactory execution with low liquidity, a lack of tradable underlying assets, and no adequate hedging solutions. During the year, several options AMMs tried improving on this through mechanisms intended to unify liquidity and enable AMMs to perform more complicated actions for hedging purposes such that LPs are better protected from toxic flow. Having said that, the most significant trend for the market during the year has been the introduction of new hybrid CLOB-based exchanges built on application-specific rollups, where orders are matched offchain and trade settlement takes place onchain. In addition to options, these exchanges can offer derivatives products such as perpetual futures. Aevo’s CLOB exchange went live in April but was NFT-gated until general access was enabled in the middle of June (it’s worth noting that the data sets used for the above/below graphs exclude Aevo prior to July 30, 2023), while Lyra is actively developing its own CLOB exchange. Excluding structured options products, Lyra dominated the vertical in terms of notional volume in the first half of the year but has massively lost market share to Aevo’s CLOB exchange. 

A notable market entrant in 2023 was Rysk Finance, which went from a ~1% share of notional onchain options volume traded in October to ~10% in November. It’s worth mentioning that a large portion of the increase in market share likely results from ARB incentives connected to the STIP program that began in early November. Rysk launched its Dynamic Hedging Vault (“DHV”) to the public in July 2023, which strives to maintain market neutrality for LP positions against which both call and put options can be sold and bought. What makes the DHV stand out is its hybrid design—a combination of an options AMM and a request-for-quote system—as well as a mechanism that adjusts option prices depending on the pool’s delta exposure and utilization. Despite the project’s heavy focus on LP yield, the cumulative return for DHV since its public launch is ~(5.5%) as of December 15, which exemplifies the difficulty of building an options AMM that consistently produces stable, positive LP returns.

The volume of onchain options trading is still minimal compared to trading on CEXs. As mentioned earlier, options projects upgraded their protocols’ structures and changed their approaches to the underlying exchange mechanisms during the year to penetrate the offchain market. In addition, new protocols based on peer-to-pool models entered the scene. In general, these solutions are built on top of spot AMMs, with some projects utilizing their own concentrated liquidity pools, offering perpetual options and volatility trading products. To simplify, these next-generation projects enable users to deposit and withdraw liquidity to and from the underlying AMM pools in specific ways to generate payoff profiles that mimic buying and selling options or volatility. 

One of the main advantages of AMM LP-enabled options protocols is that as long as the underlying has a liquidity pool and a seller and buyer can be matched, it’s technically possible to create option-like derivatives on any assets. The aforementioned could help solve a problem the crypto options vertical has struggled with: low or nonexistent liquidity for options on long-tail assets. It’s worth pointing out that projects are also trialing request-for-quote systems to facilitate the trading of long-tail options.

The three projects within the AMM LP-enabled category with the most interest from the market are GammaSwap, Panoptic, and Dopex v2. The two former projects are still in their early days, with GammaSwap going through a soft launch on Arbitrum mainnet with permissioned pools at the end of September, while Panoptic’s gated Beta launch took place in the middle of October (mainnet launch expected for Q1 2024) and the protocol announced a $7M financing round in the middle of November. Dopex is in the process of implementing its v2 upgrade, and as the first step of the overhaul, the protocol released its AMM LP-enabled options in the middle of November. The new product has created over $40M of volume in roughly a month, placing it as the second largest options protocol, but some of this volume is likely driven by the aforementioned Arbitrum incentive program.

These projects are deploying untested mechanisms from a technical perspective. A good example of this is GammaSwap having to freeze its contracts in the middle of October due to a potentially critical vulnerability. It’ll also remain to be seen if AMM LP-enabled options can sustainably penetrate the market. One notable friction for adoption is the complexity of these products. Although projects are abstracting the process of opening and closing positions, traders still have to account for factors such as option Greeks, strike prices, and, most notably, underlying protocol-specific novelties, which could hinder adoption by retail investors, who commonly are the first group of users to take up new primitives. 

While it’s uncertain how well AMM LP-enabled options protocols will do in the upcoming year, CLOB-based exchanges are well-positioned to continue expanding by converting users from CEXs and options AMMs. As mentioned before, the two main protocols focusing on CLOB options exchanges are Aevo and Lyra.

Aevo

Aevo, formerly Ribbon Finance, had quite an eventful year, rebranding the protocol and launching its CLOB-based derivatives exchange, built on a custom, application-specific OP Stack rollup. Formerly, the project’s main product had been DeFi Options Vaults (“DOVs”), which were introduced in 2021. At the time, the product received a vast amount of attention and was thought of as highly innovative, spurring a new class of structured options protocols. It has since become clear that DOVs and principal-protected vaults have several drawbacks, which make them an unattractive choice for most traders. Followingly, it doesn’t come as a surprise that the combined TVL of Aevo’s DOV and Earn products has been down only during the year, dropping from ~$51M at the beginning of the year to ~$16M as of December 15.

The pivot to the CLOB exchange was undoubtedly a success for Aevo—the exchange accounted for ~70% of the onchain notional options volume traded in November. In addition to options on mainstream tokens, the exchange offers an OTC service through which users can trade options on long-tail assets. In October, the transitioning of DOVs settlement to the CLOB exchange began, with options sold by the vaults being fungible with options on the exchange, meaning that vault positions become liquid and users can utilize their positions as collateral for other trades.

Moreover, Aevo has begun focusing on perpetual futures more heavily, offering both traditional perps as well as pre-launch futures. Aevo Exchange has found an advantage in the highly competitive perps DEX vertical by offering perps on a wide range of underlying assets and quickly listing new assets depending on the market’s demand. The pre-launch futures have also gotten a lot of attention from traders, but actual volume and open interest have been relatively low, and it’s unclear how scalable this product will be in the long run. Nevertheless, pre-launch futures seem to be an effective tool for customer acquisition, with Aevo well-positioned to continue penetrating the perps DEX market in 2024. As the exchange has continued innovating, it’s no surprise that KPIs have constantly grown across the board.

Concerning Aevo’s operational side, the project is still in the process of rebranding and implementing new token and governance structures. AGP-1 was approved on November 9, which, most notably, proposed a rebranding timeline for RBN. The new token, AEVO, and associated mechanisms are set to be released no later than February 1, 2024. RBN will be swappable 1:1 for sAEVO, a staked version of AEVO with a staking duration of 3 months. AEVO is solely a governance token, but 75% of protocol revenue will be used to increase AEVO-USDC protocol-owned liquidity once the team’s operating expenses ($5M yearly budget) have been covered. AEVO will begin trading with relatively low liquidity and limited sell pressure, aspects that are covered in-depth in this Flashnote. It’s worth noting that the DAO Treasury currently holds ~110M RBN, all of which will be converted to AEVO at the token generation event. Of the treasury’s AEVO holdings, up to ~35% will be used for incentives, such as airdrops, and up to ~10% will be used for community growth and bounties. These two allocations should accelerate the project’s growth and mindshare even further, at least in the short term. 

Most recently, at the beginning of December, Aevo announced spot swaps within the exchange and implemented a new mechanism that allows users to earn yield on their collateral assets. The passive yield is created through aeUSD, an ERC-4626 asset on Aevo L2, backed by a target of 5% (subject to change) in USDC and 95% in sDAI, which means that 5% of the underlying is instantly liquid. In other words, 95% of the underlying generates yield based on MakerDAO’s Enhanced DAI Savings Rate. In a situation where the USDC buffer is insufficient to cover withdrawals, users will need to wait until sDAI can be redeemed for DAI and swapped for USDC, which could potentially become a bottleneck for redemptions. The rebalancing of the underlying basket takes place daily such that the target weights will be maintained. As no other perps DEX or options protocol currently offers passive yield on deposited collateral, the aeUSD mechanism could possibly help attract additional capital to the platform.

Lyra

Lyra has launched several significant upgrades this year, including its Newport upgrade and V2 upgrade, which is now live. The Newport upgrade allowed the market-making vault to partially collateralize short positions with stablecoins such as USDC and sUSD, removing the need to swap to the base asset to collateralize and hedge. Thus, the vault no longer needs to delta hedge a large amount of the base asset as collateral, resulting in a large amount of fees saved for liquidity providers. In addition, the Newport upgrade allowed the protocol to integrate with any perpetual exchange on any EVM-compatible chain. Along with the Newport upgrade, Lyra expanded to Arbitrum using GMX as a source of liquidity to collateralize and delta-hedge the AMM.

Lyra V2, announced in July 2023, is an expansion of the protocol’s functionality. V2 aims to be a decentralized settlement protocol for spot, perpetuals, and options trading, built as an OP stack rollup. V2 will support portfolio margin, cross-margin, and multi-asset collateral. Lastly, V2 aims to be a decentralized margining and risk engine, along with an open source order matching engine. The order matching service will be off-chain, like many other perp DEX implementations. Lyra V2 went live on December 14, 2023.

As shown in the below graphs, the protocol has seen lackluster growth in 2023 regardless of the metric. Open interest, notional volume, premium volume, and TVL have either been on the decline or stagnated for 2023. This is unsurprising given Aevo’s launch in 2023, where much of onchain options trading volume has occurred. One can also attribute the decreasing TVL to a plethora of alternative yield opportunities, such as stETH, LSDFi, or RWA opportunities that  have overshadowed conventional LP opportunities.

DEXs

Osmosis

2023 was a defining moment for Osmosis as it transitioned from a DEX appchain to a full stack DeFi Hub with borrowing/lending (Mars), perpetual futures (Levana), decentralized stablecoins (Membrane), and other third-party applications launching on the chain. Additionally, Osmosis introduced “supercharged liquidity” to offer concentrated liquidity pools to optimize capital efficiency and improve price execution for larger swaps and released a three-pronged tokenomics revamp that drastically reduced inflation by 67% and redirected a larger allocation of inflation from liquidity providers to Osmosis stakers.

Osmosis also introduced a 0.1% taker fee on all swaps, an EIP-1559-like tx fee mechanism to limit spam, and began internalizing MEV via the ProtoRev module built in partnership with Skip Protocol. With early signs of a revived bull market, Osmosis protocol revenue is going parabolic. Osmosis is currently on pace for over $11M annualized protocol revenue going to OSMO stakers, which is equivalent to over 2.5% non-inflationary yield. 

With Osmosis likely launching its own perps platform in 2024 and a trustless cross-chain swap protocol based on the existing Thorchain codebase, Osmosis is one of the most promising projects in crypto to watch out for. 

Uniswap

In June 2023, Uniswap announced V4, the next major upgrade to the protocol. Uniswap V4 featured a completely redesigned architecture with a singleton contract model, EIP-1153 for transient storage, ERC-1155, and most importantly, hooks. Hooks allow users to launch liquidity pools with custom behavior before or after certain actions, such as initializing an LP position, modifying an LP position, making a swap, and donating to in-range liquidity providers. Importantly, V4 also allows liquidity pool deployers to turn on a fee switch, similar to the protocol level fee switch, that will direct some of the LP fees to the deployer. Uniswap V4 will go live with the Dencun upgrade to Ethereum.

In July 2023, Uniswap also released the whitepaper for UniswapX, a non-custodial dutch auction-based trading protocol. It is an RFQ system that outsources routing and batching to a set of fillers, who can route orders to a combination of onchain and offchain liquidity. As one can see below, UniswapX has been gaining increased adoption, although it still comprises a small percentage of Uniswap’s volume.

Lastly, in October 2023, Uniswap Labs announced an interface fee switch, charging 0.15% on swaps on select tokens through official Uniswap interfaces. Tokens that are subject to the interface fee include WETH, WBTC, USDC, USDT, and a few other stablecoins, and the fee is additive to the existing swap fee. In a flashnote, we outlined how in our view, this decreases the likelihood of a protocol fee switch and is bearish for UNI token value accrual. Since launch and as of December 16, 2023, the interface fee switch has generated $2.93M in total revenue, and extrapolating that, $18M in annualized revenue.

Curve

Throughout the year, Curve released a few new innovations centered around passive AMMs. In contrast to the Uniswap V3 design, passive AMMs allow LPs to simply deposit funds and let the AMM balance their position. The protocol upgraded their AMM design to the next generation (ng) of passive AMMs for three-coin volatile pools (tricrpyto-ng) and stablecoin pools (stableswap-ng). The main focus of the upgrade was to lower gas costs to capture more arbitrage flow, but it also improved the pool oracle to be used as a price oracle within other DeFi applications. The ng contracts are now 40% cheaper than the previous versions yet still require 68% more gas than the cost to process Uniswap V3 swap logic. Since launching in May, the USDC/WBTC/ETH and USDT/WBTC/ETH pools have combined for over $800M volume but the new pools have failed to increase Curve’s share of USDC<>ETH and USDT<>ETH volume relative to Uniswap, likely due to the higher gas costs. 

Even with the improved DEX design, the protocol’s largest innovation of the year was the release of Curve’s lending market (LLAMMA) and stablecoin (crvUSD). The mechanics of the protocol were explained in our extensive report. Since launching in May, crvUSD debt borrowed from the protocol peaked at $160.3M from $280.8M in collateral deposits, and the lending protocol has generated $2.9M in fees for the DAO. Lido stETH is the most popular collateral at 43% of deposits followed by WBTC at 33% and WETH at 13%. Interest rates are specific to each collateral but have hovered between 4% and 9% for the majority of the time. However, a recent governance proposal hiked the base rate and effectively doubled the borrow rate. The aim of the proposal is to activate the peg keepers so there is a buffer against downward depegs, similar to how the PSM helps the DAI peg. Thus far, crvUSD has maintained a resilient peg, but the team is focused on ensuring it remains that way. 

Curve has distributed $123M to its governance token holders since 2021, more than any protocol in the industry. Zooming in on 2023, Curve has distributed $16.3M and saw a material uptrend beginning in November that pushed veCRV yield to 5.4%. Notably, there is a shift in revenue breakdown as lending fees have overtaken DEX activity. This is largely driven by the growth of crvUSD and the larger fee share as 100% of lending fees accrue to the DAO compared to just 50% of swap fees. The outlier week in March was driven by the USDC depeg, which drove north of $8B in swap volume on Curve. 

 

However, 2023 was not without pain. The protocol suffered a $62M exploit due to a compiler-level bug in Vyper, the smart contract language used to write Curve contracts. The bug caused a failure in a security feature that allowed attackers to drain four impacted pools. The exploit left unaffected LPs concerned about their deposits, and Curve TVL is still down ~$1B four months after the exploit. The CRV token took about a 30% price hit which quickly cascaded the problem onto lending markets that accepted CRV collateral. Curve founder Michael Egorov had $110M in outstanding loans against CRV collateral and was forced into OTC selling more CRV tokens at $0.50 to unwind his debt. Over 50M CRV tokens were sold at $0.40 with a rumored six-month lockup, creating a supply overhang heading into the first quarter of next year.

THORchain

THORChain launched new products throughout the year that saw strong adoption, leading to overall market share growth. Focusing on the cross-chain DEX offfering, THORChain trading volume exploded north of $1B per week in Q4 – a 10x increase from the start of the year. New asset listings were sparse in 2023, with the most notable being a few Ethereum stablecoins like USDP, LUSD, and GUSD. All eyes are on SOL to be the next major integration, but there is technical overhead that still needs to be resolved before this is feasible. 

While volume share composition shifted throughout the year, Bitcoin and Ethereum were unsurprisingly the most in-demand networks. BNB Chain chain closed out the year as the biggest loser of volume share, decreasing from 29% to 5%. On the other hand, Ethereum stablecoins gained the most from 4% to 29%. 

In August 2023, the protocol launched its lending platform that allows users to borrow against ETH or BTC collateral with zero-interest and no liquidations. To date, $4.7M and $1.1M has been borrowed against BTC and ETH, respectively. Debt is denominated in USD, but users can borrow any asset supported by the network. Our report from earlier this year discusses the mechanism design and key risks to the protocol. 

Since RUNE is burned each time a loan is opened and minted when debt is repaid, the lending mechanism has a deflationary impact on RUNE when the demand for leverage is high. Market participants are beginning to use THORChain for leverage amid the recent positive price action, resulting in a total burn of 3.4M RUNE or 0.6% of the total supply. 

The biggest risk to THORChain is regulatory pressure as the protocol recently became a popular bridging route for stolen funds. Given THORChain’s wide network coverage, hackers typically bridge exploited funds from the source chain to Bitcoin where the ilicit funds are moved through mixers. Although THORChain does not obfuscate the wallet address or act as a mixer in anyway, there is a view that THORChain is not doing enough to prevent this. THORSwap, the leading DEX on THORChain, recently took action to prevent this by . All in all, there is an increased risk of negative regulatory action against THORChain due to its role in recent exploits. 

Trader Joe

Despite being launched in November 2022, Liquidity Book was one of Trader Joe’s largest innovations and propelled much of its growth in 2023. In addition, TraderJoe expanded to Arbitrum One, BNB Chain, and Ethereum in addition to its core Arbitrum outpost. 

Throughout 2023, TraderJoe launched several new features such as Autopools which are automated strategy vaults for Liquidity Book, permissionless pools, and onchain limit orders. JOE tokenomics underwent a revamp, with veJOE being sunset, and a new staking mechanism, sJOE being introduced, with sJOE receiving a % of trading fees if staked. Lastly, Trader Joe announced Merchant Moe, a purpose-built DEX for Mantle, with 2.5% of the new token, MOE, being distributed to JOE holders upon TGE with a further 5% to be distributed over the coming 12 months. 

Yield Aggregators

Sommelier Finance

Yield aggregators were mostly a dead sector in 2023, with no successful innovations coming from category leaders like Yearn. Prior cycle aggregators generated yield with static strategies that took advantage of unsustainable liquidity mining programs of other DeFi protocols and borrowing/lending rate dislocations in top lending protocols. With little demand for leverage in 2023 and no large-scale liquidity mining programs, these aggregators saw no real TVL growth and also were underperforming simply liquid staking ETH. Sommelier Finance was the major outlier and was the fastest growing EVM yield aggregator, with TVL growing from ~$1.5M to over $60M in 2023. Since yield aggregators are effectively onchain asset managers, TVL is not a vanity metric and is a direct correlation to revenue potential.

Sommelier is the first live Ethereum coprocessor that uses a Cosmos appchain to optimize the EVM DeFi experience. Sommelier’s strategists leverage offchain compute to deloy onchain dynamic yield strategies like UNI-v3 tick optimization and lending optimization. Today, Sommelier offers the best non-inflationary yield for ETH/stETH, wBTC, FRAX, and others. With L2 yield strategies launching in Q1 2024, we expect Sommelier to continue growing and potentially become the largest EVM yield aggregator in crypto.

Pendle

Pendle has cemented itself as the largest interest rate derivatives protocol in DeFi by a large margin, especially in the LSTFi yield trading space. In terms of TVL, it has been one of the outperformers of the year, with TVL consistently up and still growing. Since launching stETH pools in April, Pendle’s stETH liquidity as a % of stETH’s market cap has continued to grow. In September, the protocol made its first foray into RWA markets by listing sDAI and fUSDC pools, expanding the total addressable market. Following that, Pendle has launched a large number of other markets, including sFRAX, crvUSD, ePENDLE, and more. With the rise of stablecoins such as sFRAX/crvUSD and interest-bearing versions of stablecoins such as sDAI in the latter half of 2023, Pendle has significantly benefited from those tailwinds.     

Staking & Liquid Staking

As mentioned in our last report highlighting the state of Ethereum, LSTs saw a huge rise this year following the Shapella upgrade. This was a result of the reduction of risk associated with holding LSTs. Withdrawals being enabled now meant that prices could stay more in line with the intended peg via the ability to arbitrage, and users could receive their entire ETH balance as long as they waited a few days for the protocol’s buffer to handle the withdrawals. 

Lido’s dominance did not change much throughout the year, owning over 85% of all liquid staked ETH. In February, Lido announced V2, which included two important upgrades to Lido’s core tech stack: a new staking router and enabled withdrawals. The staking router is a new modular design for the underlying set of node operators bucketed into “modules” that stakers can pick and choose from. This includes a module for permissionless community stakers, DVT operators, and others alongside the traditional whitelisted node operator set. Withdrawals also come from Lido’s ability to create an ETH buffer where rewards and new deposits sit until they can be deployed. This buffer is then maximized in order to have as much ETH that can be staked while also satisfying withdrawals without needing to exit any validators. 

stETH has also benefited from its use as deposits for other highly anticipated protocols, the two largest being EigenLayer and Blast. EigenLayer continues to raise caps on its  liquidity mining program, where select LSTs are allowed to be used for deposits. Just under 200K stETH has been deposited into EigenLayer, with the second largest deposit being swETH at 46K. Blast, a new L2 whose selling point is “native yield” on the L2, takes all of their bridged ETH and deposits it into stETH with the plan to automatically distribute the ETH yield generated to each user’s wallet on the L2. This chain has seen a large amount of interest, with over 368K stETH sitting in the Blast L1 Smart contract, which will continue to grow until withdrawals are enabled in February.

Rocket Pool

Throughout 2023, Rocket Pool continued to hold its spot as the second largest decentralized LST provider, maintaining a 4.75% market share. The protocol implemented the Atlas Upgrade in April 2023, which introduced LEB8 pools, meaning that a node operator could set up a validator by posting only 8 ETH instead of the previous 16 ETH requirement. As a result, the protocol saw a 50% increase in node operators and an 85% increase in minipools, which further enhanced the decentralization of Rocket Pool’s operations. As an added benefit, the LEB8 pools require a higher amount of RPL collateral, which helped increase RPL demand. 

The protocol has a few catalysts set up for 2024, including NodeSet, which will act as a complementary protocol to Rocket Pool with the introduction of xRPL and xrETH. The upcoming Saturn Upgrade will also reduce the minimum ETH requirement for node operators from 8 ETH to 4 ETH, further democratizing the protocol and increasing potential for added market share. The future is bright for Rocket Pool, and its decentralized ethos cements it as a major contender in the LST market. 

frxETH

Frax launched its frxETH LST in the end of 2022, and its two-token model (sfrxETH/frxETH) quickly grew to become the third largest decentralized LST on the market with a 1.77% market share. In only a year frxETH grew from essentially $0 to ~$500M. frxETH’s composable design allows for it to be integrated with Fraxlend, which enables users to add leverage to their staked ETH positions. The team plans to launch frxETH V2 in 2024, which should assist in scaling node operators and decentralization. An upcoming Frax rollup will also bring more utility to frxETH, which should prove interesting as we have not yet seen much innovation on L2 designs, especially when it comes to an L2 with its own LST behind it. 

Stride

Due to high staking yields in the Cosmos ecosystem, Cosmos DeFi has historically struggled to gain significant traction. Liquid staking has the potential to unlock billions of dollars of staked assets in the Cosmos by allowing users to speculate in DeFi without forgoing staking rewards, promising to be a fruitful venture for whichever protocol holds the dominant market share.

In 2023, Stride became the premier liquid staking protocol in the Cosmos ecosystem and aligned with the Cosmos Hub to become the second consumer chain leveraging the Hub’s security. With nearly $100M TVL and 85%+ market share of Cosmos liquid-staked assets, it is the best index bet on the growth of IBC today.

The recent launches of dYdX and Celestia are likely to become a major catalyst for Stride revenue growth once it supports liquid staked versions of these assets. While other LST providers, including Lido, will attempt to take market share from Stride, LSTs are mostly a winner-take-most market structure. Stride’s early mover advantage, swift coverage expansion, and continued innovation on the tech put it in a strong position to maintain dominance as the de-facto Cosmos LST provider.

Marinade

Currently, Marinade Finance is the largest liquid staking/staking protocol on Solana in terms of TVL. Despite lackluster growth in the first half of 2023 (which included a large 1M unstake related to the fallout of the FTX-Alameda collapse), the protocol made significant technical developments and revamped various incentive programs to attract more user deposits. In January 2023, Marinade introduced an ecosystem incentive program, the Open Doors Program, to increase the adoption of mSOL with protocols and validators. Up to 160M MNDE (16% of the total token supply) was earmarked to be distributed to various protocols, wallets, individuals, and validators who grow the mSOL supply, and is estimated to distribute 10-12M MNDE by completion. In January, the DAO also voted to reduce liquidity mining emissions by 75% and eventually scrapped them completely in favor of a new Earn incentives campaign that launched in October. In addition, the team voted to switch from time-based unlocked to milestone-based ones so the team does not receive any MNDE without TVL growth, and the founding team's 7.5% allocation will be fully vested by January 2024. After implementing these new initiatives, the protocol’s TVL has been steadily increasing, especially in the second half of the year.

On the technical front, Marinade revamped its delegation strategy, improving the performance of the stake pool by narrowing down the validator set to the top 100 validators, identifying and blacklisting commission ruggers, and making high-performing super minority validators eligible for stake. Marinade also released Directed Stake, which allows stakers to support individual validators while holding mSOL, giving users the choice to delegate their SOL stake to a single validator. In July, Marinade also launched Marinade Native, which utilizes the same delegation strategy as the mSOL stake pool and uses the stake bot to create 100+ stake accounts while maintaining zero fees. 

DeFi integrations for mSOL drastically increased in 2023, including the launch of Super Stake SOL from Drift using mSOL, which is a leveraged staking vault, and other lending/borrowing integrations like MarginFi and Kamino Lend. Most recently, Marinade recently introduced Protected Staking Rewards, which will require validators to set up a protection bond to be eligible to receive stake from Marinade depositors. If the node operator’s performance is poor, the bond deposit will be used to compensate for the staking reward losses experienced by Marinade. Lastly, Marinade’s boosted rewards market continues to develop, with validators receiving stake to their node through mSOL/MNDE voting. Validators (such as Solana Compass and Solana Hub) are beginning to boost SOL rewards to Marinade stakers who direct stake mSOL or MNDE stake to them.

Jito

Jito is a protocol building robust MEV infrastructure for the Solana community. Their products, the Jito-Solana MEV client, the Jito Block Engine, and the Jito Relayer, are all now crucial components of Solana’s MEV supply chain. However, their most public product is jitoSOL, a staked SOL alternative that receives a portion of MEV rewards in addition to standard SOL validator rewards. 

In April, Jito introduced a Jito Staking xNFT, allowing Backpack wallet users to manage their staking positions directly through their Backpack wallet. Later in the year, they released a referral and points program, which when combined with MarginFi’s points program, likely greatly contributed to the protocol’s growth. Likely their most important technical announcement of the year, Jito introduced StakeNet, a self-sustaining protocol for intelligence Solana LSTs in October. StakeNet is a network of keepers and onchain programs which allows protocols to fully decentralize stake pool operations. As shown below, JitoSOL has shown an impressive amount of growth over 2023, with TVL in SOL terms being up more than 10x. 

Lastly, on December 7, 2023, Jito’s hotly anticipated token, JTO, went live. With 90M tokens being airdropped and immediately liquid, that meant that roughly $225M (using the ~launch price of $2.50) was airdropped to users, node operators, and MEV searchers. This kick-started significant wealth effects for the ecosystem, and with less than 10,000 recipients, was one of the larger airdrops in crypto history on a dollars received per user basis. 

Stablecoins

The professionalization of fiat stablecoins was a meaningful trend throughout 2023. Tether’s USDT and Circle’s USDC continue to be the dominant offerings, but new players like Paypal have entered the arena with their own stablecoins. Other institutional-grade participants are now experimenting with stablecoins as means to improve settlement of cross-border transactions. SAP announced its USDC payment offering being tested on Ethereum’s Goerli testnet, citing high fees, long transaction times, and lack of transparency as reasons to rethink its existing infrastructure. In continued evidence of the industry’s maturation, S&P Global recently launched a stablecoin stability assessment that evaluates the ability for stablecoins to maintain their peg. USDC, USDP, and GUSD were ranked the highest while USDT, DAI, and TUSD were ranked towards the bottom. 

However, it is impossible to ignore the systemic risk USDC posed to DeFi when it depegged during the SVB Bank Run in March. After hearing Circle held a portion of USDC backing with SVB, token holders sold off the asset to lows of $0.90 over the weekend, anxiously awaiting more news about the safety of SVB assets. The token repegged in the days following the event, but it was a stark reminder of the gap between onchain assets and their real world backing. The depeg put further pressure on the evaporating market cap of USDC, with holders derisking into other stabelcoins or moving offchain. USDC topped at $58.9B in June 2022 and hit a yearly low of $23.8B in November of this year. 

The largest benefactor of the USDC depeg was undoubtedly USDT, which saw meaningful inflows throughout the year and currently sits at market cap all time highs. Tether focused heavily on transparency this year. BDO, a top five global public accounting firm, now publishes attestation reports for the funds backing USDT. The most recent report notes a $72.6B exposure in US T-Bills, highlighting a change to conservative risk management. With regards to the location of USDT, 51% of the circulating supply resides on Tron. The trend began in late 2022, but there is not clear indication of why this flow is occurring. Ethereum holds another 40% of USDT with smaller L1s and L2s holding the remaining $7B stablecoins. 

Frax

2023 was a monumental year for Frax, largely due to growth in its LST, frxETH. In addition to this new product, governance initiated Frax V3, an upgrade to the FRAX stablecoin that utilizes RWAs. The USDC de-peg event in March resulted in Frax governance pivoting from fractional-reserve stablecoin to a fully collateralized stablecoin. Since this decision, the Frax collateral ration (CR) has increased to ~95%. After 100% collateralization is achieved, the protocol will allow veFXS holders to vote on how to use excess protocol revenue, which will likely be used to buyback FXS. 

Frax V3 introduced sFRAX, which lets FRAX holders deposit into a vault to receive a weekly yield, powered by RWAs. The protocol partnered with FinresPBC, an offchain entity, to custody US treasuries. FinresPBC essentially withdraws FRAX from the sFRAX vault, deploys it into RWAs, and then passes the yield back to sFRAX holders by minting fully-backed stablecoins. sFRAX supply currently sits at ~$20M. So far, the protocol has deployed this sFRAX into MakerDAO’s sDAI. FinresPBC has not yet been utilized. In 2024, the second part of Frax V3, Frax Bonds (FXBs) will launch, which lets the protocol issue FRAX bonds, granting participants a discount on FRAX through longer duration yield. Despite the new yield products, FRAX supply sits at ~$650M, which is ~$330M less than its ~$1B supply at the beginning of 2023. 

In addition to the updated stablecoin products and LST, Frax launched the testnet for its own rollup. It is unclear how the network will differentiate itself from other L2s, but it has been speculated that the rollup will provide additional utility for frxETH. The protocol also started a slow rollout of its new frxGov governance module, which further iterates on onchain governance models. The module currently controls Fraxlend, but upon complete implementation it will make the entire Frax protocol completely permissionless. Frax has positioned itself to provide products that stand in any market condition, whether that be through RWA-backed stablecoins in a bear market or ETH-yielding LSTs in a bull market. For a complete overview of Frax’s latest developments, more information can be found in our latest report.

Ethereum Lending Markets

Given Ethereum still holds the majority of onchain liquidity, it is no surprise that most lending activity still resides there. Aave and Compound combine for nearly $11B in deposits and $3.6B in active loans. Of that, Aave V3 is solely responsible for 43% of deposits and 48% of active loans. The latest mechanism design enabled more robust parameter control, allowing the protocol to fine tune its risk controls. Aave also launched its own stablecoin in July to improve its revenue capture and create a more robust service offering. However, GHO got off to a shaky start with questionable mechanism design leading to peg troubles. GHO relies on a governance controlled borrow rate which was set far below the market rate at launch. As such, users continually borrowed GHO against yield bearing collateral to capture the spread, selling GHO on the open market in the process and pushing the peg below $0.96. The problem was exacerbated by a lack of preparation by the Aave DAO to create sufficient liquidity for the new stablecoin. As such, a governance proposal was passed to create and fund a liquidity committee to repeg GHO. After pumping some incentives into a few DEXs, the peg jumped to $0.985. The initial supply cap of $35M has also been hit, which should also help repeg the asset. There will likely not be a supply expansion until the peg stabilizes. 

Spark protocol, a money market protocol built alongside MakerDAO, launched in May 2023 and had an immediate impact on the lending sector. Just seven months since its launch, Spark soared up the lending leaderboard with $853M borrowed on Ethereum mainnet. The protocol provides the cheapest access to DAI loans by charging a slight premium to the DSR set by MakerDAO governance. Spark’s codebase is forked from Aave V3, so the protocol also offers loans denominated in other approved assets. Currently, $539M DAI and $288M WETH have been borrowed from the protocol. 

Solana Lending Markets

Solana DeFi experienced a meaningful shakeup following the FTX collapse. While most protocols died off, the survivors are rising from the ashes alongside a new cohort of protocols that launched this year. While the attention is undoubtedly flowing to Solana, liquidity is still lacking. As such, the lending sector is significantly smaller than Ethereum with the top three players, MarginFi, Solend, and Kamino, combining for $700.8M in deposits and $166.8M in active loans. 

MarginFi and Kamino both launched this year and have points programs that are successfully attracting liquidity into the protocols. MarginFi currently holds the majority of user deposits at $362.9M and has $81.3M of active loans, while Kamino holds $156.9M in user deposits and has $51.9M of active loans. Kamino has seen rapid growth considering it just launched in November. Solend, on the other hand, launched in 2021 and lived through the boom and bust of the previous cycle. The protocol holds $181M in user deposits and has $33.6M of active loans. 

Risk management and product depth will be key differentiators to lookout for as these protocols continue to mature. It is likely that one or more of these protocols launches a stablecoin, similar to MakerDAO or Aave, to internalize the revenue generated from borrowers. 

Cosmos Lending Markets

2023 was the rebirth of Cosmos leverage post-Terra collapse. Mars, once a promising project on Terra, relaunched as a Cosmos appchain with outpost deployments on CosmWasm chains like Osmosis. Additionally, Umee, a cross-chain lending Cosmos appchain that leverages a bespoke version of the Gravity Bridge to tap into Ethereum liquidity, launched as well and has been the dominant lending market in Cosmos due to higher LTVs and historically lower borrow rates for USDC and other major Cosmos assets.

In October, Mars launched v2 of their platform that introduced isolated credit accounts that act similarly to CEX sub-accounts. With these credit accounts, all forms of collateral, including LP positions, can be used for margin trading and leverage LPing to offer maximum capital efficiency onchain. Additionally, Mars has recently decided to deprecate its appchain due to the unnecessary burden of maintaining a validator set and is considering making Neutron its homebase to control other outpost deployments from. Although Mars is behind Umee in TVL, Mars is likely to become the leader in 2024 as demand for leverage and interest in Cosmos increases.  

RWAs

Over the past year, RWAs and tokenization have grown dramatically, with the total value of tokenized products currently sitting at around $74B USD. The majority of these tokenized assets are made up of stablecoins. However, there has also been drastic growth in non-stablecoin RWA products, as the total value locked in RWA DeFi protocols has reached more than $6B USD.

One of the fastest growing verticals amongst RWAs was tokenized debt. Tokenized debt products flourished under the high interest rate environment, as they managed to attract users with lower risk and higher uncorrelated returns compared to DeFi - even those seen from blue chip protocols.

Protocols like Ondo Finance and MatrixDock, which provide access to US Treasury bonds, have seen rapid growth in their AUM. Since April 2023, MatrixDock's TVL has nearly doubled to $108M USD, and Ondo has reached $200M USD in TVL since its launch in February 2023.

Aside from protocols offering access to US debt, we've also seen Asian governments offer treasury bonds onchain. The Hong Kong government issued $800M HKD (worth approximately $100M USD) of tokenized green bonds earlier this year, under their Green Bond Program. The Philippines Bureau of the Treasury, the Development Bank of the Philippines and the Land Bank of the Philippines recently announced that they will be offering 10B pesos (worth $179M USD) in one year tokenized treasury bonds onchain. In addition to treasury bonds, some companies have also begun to experiment with onchain corporate bond issuance recently; for example, Siemens issued 60M euro in digital bonds on Polygon in February of this year.

Tokenized credit protocols similarly have grown significantly over the past year. In particular Centrifuge had a great year, and amassed over $250M in total assets under management. In addition, they have originated a cumulative $482M USD onchain since their launch in late 2020, and have tokenized 1,341 assets in total.

Currently, approximately 47% of loans issued by Centrifuge are originated by BlockTower, and used by Maker to fund investments in real-world assets. BlockTower Series 3 and Series 4 structured credit pools hold $185M USD.

Recently, Centrifuge has focused on partnerships with DeFi protocols and DAOs are part of their newer product Centrifuge Prime, Centrifuge’s infrastructure and services offering for DeFi protocols to onboard RWAs. Recently, they have partnered with Aave to use Aave's stablecoin treasury for RWA investments through Centrifuge Prime, by allocating $1M USDC to the Anemoy LTF, a short-duration US treasury bill pool on Centrifuge. Furthermore, they are exploring backing GHO, Aave's stablecoin, with RWAs.

Apart from Centrifuge, there are also new up-and-coming tokenized private credit protocols to look forward to in 2024, such as Huma Finance. Huma offers comprehensive infrastructure for decentralized risk assessment and lending, backed by income and receivables. It focuses on high-performing, low risk receivables, and is currently one of the leading RWA protocols enabling remittance financing use cases through partners such as Arf, Circle and Stellar, The protocol distinguishes itself through its signal-driven underwriting approach, utilizing diverse data, such as income, assets, and liabilities, as critical indicators for assessing borrower risk. Its signal-driven underwriting approach is enabled by the decentralized signal portfolio, which is at the core of Huma’s lending infrastructure, and gathers and manages a wide spectrum of signals to evaluate borrower risk. 

The DSP collects a variety of information, including income, assets and liabilities sourced from both on-chain and off-chain channels. Its role extends beyond mere data aggregation—it enables authentication of these signals and welcomes the integration of new signal sources through specialized Signal Adapters. Essentially, this portfolio acts as a comprehensive repository, ensuring the collection of critical financial data. 

Since their launch on Polygon in June 2023, Huma has originated nearly $84M in loans, with zero defaults. Huma has recently launched on Celo testnet, and in the future will be integrating with decentralized reputation systems that combine on-chain and off-chain data points. Over the past 6 month, various solutions have been launched on Huma, including Arf Credit, which offers remittance financing for financial institutions, Rain, USDC-backed corporate cards, BSOS Green Financing, onchain credit backed by EV-charging station receivables and ImpactMarket, which offers UBI and microlending using stablecoins. 

Tokenized equity had a difficult year. Current adoption of tokenized equity is relatively low and the available offerings are limited. Swarm has little to no usage (its two most popular stocks, AAPL and TSLA, have $30K USD in combined market cap) and Aktionariat only provides access to Swiss small to mid cap equity issuers who decided to issue their shares onchain. Backed Finance, a hybrid tokenized debt and equity provider, has more assets under management compared to Swarm and Aktionariat, but only offers access to the S&P 500 and Coinbase (COIN). Notably, Swarm recently launched dOTC, their permissionless over-the-counter trading platform for tokenized RWAs on Ethereum. Prior to this, only fully KYC’d users were able to access and trade on their platform - which could’ve been the primary driver for low volume and usage.

On the infrastructure side, we saw many TradFi institutions explore different bridging and chain solutions over the past year. Chainlink recently partnered with the ANZ (Australia and New Zealand Banking Group) to explore how its customers can use CCIP to settle tokenized assets across public and private blockchains. The team has also partnered with SWIFT and large financial institutions including the ANZ, BNP Paribas, BNY Mellon, Citi, Clearstream, Euroclear, Lloyds Banking Group, SIX Digital Exchange (SDX), and The Depository Trust & Clearing Corporation (DTCC) to explore how existing SWIFT infrastructure can be leveraged to transfer tokenized assets across public and private blockchains using CCIP.

Axelar and LayerZero, alongside Avalanche, both participated in a proof-of-concept led by JP Morgan, to explore messaging and asset transfers between private and public blockchains. As part of this POC, Axelar provided message validation and message passing by deploying a permissioned Axelar testnet and private offchain relayers. LayerZero provided an oracle and relayer to pick up an offchain message, and transmitted it to LayerZero endpoint contracts on Avalanche. 

Despite Chainlink’s explicit focus on institutional adoption of CCIP, the participation of Axelar and LayerZero in the JPM Onyx POC shows that the playing field for interoperability providers is still very much even. Given that tokenized asset interoperability between private and public chains is still relatively underexplored and nascent, it’s likely that the ideal solution has not been implemented yet.

MakerDAO

For MakerDAO, the year 2023 can be summarized into three main categories: RWA collateral, DAI supply, and Endgame. For an overview of Endgame-related changes that took place during the year, please refer to the governance section in this report below. RWAs as DAI collateral isn’t a completely new implementation per se, but utilization has been relatively minimal during previous years. This changed with the introduction of MIP65, also known as the Monetalis Clydesdale Vault, at the end of 2022. This was MakerDAO’s first RWA vault to invest in short-dated T-Bills through an allocation of up to 500M DAI. At the beginning of April 2023, the vault’s maximum size was increased to 1.25B DAI, but new capital deployment didn’t notably ramp up before May. Another major T-Bill vault, BlockTower Andromeda, was initiated in July, and its maximum debt ceiling was set to 1.28B. In the middle of October, both of the aforementioned vaults’ maximum debt ceilings were increased to 3B.

Not all of MakerDAO’s RWA investments have succeeded or been deemed appropriate. For example, the Fortunafi, ConsolFreight, and Harbor Trade Credit vault debt ceilings have been decreased to zero during the year, while the HVB vault has been instructed to terminate future investments and return excess available cash to the DAO. Regarding Harbor Trade Credit, its outstanding DAI is supported by loans to a single borrower that defaulted in April 2023, while the ConsolFreight vault’s largest debt position has similarly defaulted. This means that MakerDAO could lose some invested capital, although the project is a senior debt holder and the amount invested is relatively small. This exemplifies the risks associated with deploying collateral assets into RWA loans.

Nevertheless, short-dated T-Bill collateral has become a major revenue source for MakerDAO and facilitates the project’s ability to provide a high yield of 5% through the DAI Savings Rate (“DSR”), at least until the interest rate environment begins reverting. At the end of June, the market began to pick up on the revenue potential deriving from T-Bill vaults, and MKR performed extremely well in Q3. The price of the MKR token outperformed its peers, ending the quarter with a circa 100% increase. Having said that, MakerDAO’s core product and enabler of the RWA investments, DAI, has struggled to attract more demand.

Until the beginning of August 2023, the previous year’s trend of a decreasing DAI supply continued, with USDT taking market share from most of the competing stablecoins. Even as DAI offered a competitive yield through the DSR in June and July, the stablecoin’s supply continued declining. Consequently, the Enhanced DSR (“EDSR”) was introduced, a temporary increase to the DSR when utilization is low. Once utilization passes a predetermined level, the effect of the EDSR also decreases. EDSR is a one-time implementation, meaning that although utilization would begin reverting below a predetermined level, the yield won’t enlarge again. EDSR initially boosted the yield to 8%, which induced a notable increase in the DAI supply, but it quickly became clear that whales were arbitraging the implementation, and the increased yield became a large cost to MakerDAO. As such, the EDSR and associated mechanisms were adjusted, and the yield was capped at 5%, where it has been maintained since then. The EDSR stopped the DAI supply shrinkage, with the total supply having grown by ~1B from this year’s lows, but growth has again stagnated over the past two months.

There has historically been a vast amount of discussion regarding whether or not DAI should be backed by other stablecoins. As it currently stands, there are three Peg Stability Modules (“PSMs”) through which DAI can be minted 1:1 against the other stablecoin in the module regardless of at what price the stablecoins are trading at. The most significant module is the USDC PSM, responsible for ~8% of the total DAI supply. MakerDAO leverages the USDC PSM to deploy capital into RWAs. This is done by minting DAI, swapping against USDC in the PSM, and finally, the USDC is redeemed and invested. When USDC lost its peg to the U.S. dollar due to the bank run on Silicon Valley Bank in March, DAI faced major sell pressure because investors swapped USDC for DAI through the PSM, whereafter DAI was sold on the open market. The aforementioned caused DAI to shortly trade at a price as low as $0.88, but the stablecoin quickly recovered as USDC regained its peg.

In addition to being an investment vehicle for MakerDAO, the USDC PSM functions as a stabilizer for DAI’s peg, which is why the community chose to maintain USDC as an asset backing DAI after the depeg event. The community’s opinion at large is that it’s important that there are enough onchain pairs with sufficient liquidity for DAI such that the price can quickly be arbitraged if needed. Accordingly, the project has implemented a directive stating that when aggregate USDC in the USDC PSM and two Uniswap V3 DAI/USDC pools (0.05% & 0.01%) falls below 300M, USDC is withdrawn from the Coinbase Custody RWA vault into the USDC PSM such that aggregate USDC increases to 400M. Conversely, when aggregate USDC in the PSM and the two Uniswap V3 pools exceeds 500M, USDC is withdrawn from the PSM and moved back to the Coinbase Custody vault such that aggregate USDC decreases to 400M.

The process of moving capital between the vault and the module is inefficient and takes place with a delay. This is because the trustee is alerted manually by the CEO of Monetalis, who has historically reacted relatively slowly when thresholds have been breached, under this forum post when funds need to be moved, and Coinbase has to move USDC from cold storage wallets to its hot wallet before the capital can be sent to the PSM. In Q4, the PSM has experienced increased volatility and relatively large withdrawals of USDC, forcing MakerDAO to heavily draw down capital from the Coinbase vault, leaving it practically empty occasionally. As of December 15, the Coinbase vault only holds ~70K USDC. If the USDC PSM faces continued withdrawals and MakerDAO wants to maintain the module’s liquidity, the project will likely have to draw down capital from its T-Bill RWA vaults next.

In December, the Coinbase vault’s debt ceiling was increased from 500M DAI to 1.5B DAI, likely in preparation for Endgame, where SubDAOs currently will be required to hold at least 18% of their total portfolio in Cash Stablecoins (defined as USDC and USDP). It seems that MakerDAO is beginning to prepare to rotate capital away from other RWA vaults into the Coinbase vault so that SubDAOs can quickly access stablecoin liquidity. This would mean that the project’s blended interest rate on RWA vaults decreases notably since T-Bill vaults are returning ~5% in the current interest rate environment, while the Coinbase vault yield is a constant 2.6%. 

However, this might inadvertently be a decision that decreases MakerDAO’s liquidity risk. The heavily scrutinized L2, Blast, and Aevo will use the DSR to create passive yield on assets deposited into the aforementioned projects. Blast is a novel implementation that has already amassed a notable amount of TVL, and users may suddenly want to withdraw funds en masse when the ability to do so is enabled. Since this implementation will use the USDC PSM for withdrawals, MakerDAO could run into liquidity issues if a majority of its assets are in a lower liquidity tier.

Lastly, it’s worth mentioning that MakerDAO reactivated a value accrual mechanism for MKR with the implementation of the Smart Burn Engine (“SBE”) in July. MakerDAO’s profits are sent to the Surplus Buffer, an insurance fund containing DAI in case the protocol accrues bad debt. Before the SBE was implemented, capital in the Surplus Buffer hadn’t been actively utilized since January 2022, when the buy-and-burn mechanism for MKR was deactivated. With the SBE, when DAI in the Surplus Buffer passes a predetermined level, half of the additional capital from the buffer is used to periodically acquire MKR tokens from the Uniswap V2 DAI/MKR pool. The acquired MKR is combined with the other half of the additional DAI, and the position is LPd back into the same Uniswap pool. As it currently stands, when the Surplus Buffer reaches over 50.03M DAI, the SBE activates and uses 30K DAI to acquire MKR every ~4 hours until the buffer decreases back below the threshold. The Surplus Buffer has recently experienced some notable increases in DAI from time to time because certain RWA profits are paid in lumps.

NFT Marketplaces

Blur

Blur dominated NFT volume on Ethereum for all of 2023. The Blur team not only built a superior product that engaged with collectors, but also more sophisticated NFT traders who trade NFTs in a similar manner as altcoins. Blur was also able to incentivize NFT traders with a steady stream of points and airdrops. Despite OpenSea having a first-mover advantage, since February 2023, Blur has dominated the NFT marketplace space on Ethereum. For example, over the past few weeks, Blur has consistently facilitated over 70% of NFT volume on Ethereum.

In May, Blur announced the launch of Blend, a peer-to-peer lending protocol for NFTs - built in collaboration with Paradigm. Blend allows NFT owners to borrow against their NFTs or lend them out to earn interest, with no fees for borrowers or lenders. Since its launch, Blend has facilitated nearly 400K loans, with over 1.3M ETH (nearly $300M USD as of 12/24) borrowed.

OpenSea

While Blur had a good year, OpenSea struggled. Back in November, Coatue, an investment manager, reportedly wrote down their investment in OpenSea by 90%, indicating a potential valuation for OpenSea at or below $1.4B- a considerable drop from its valuation of $13.3Bearlier in 2022.

Over the past few months, OpenSea has introduced some meaningful changes to onboard more users. In August, OpenSea announced that they are developing a new open standard for redeemable NFTs - to be used to redeem other digital or physical items. In October, OpenSea introduced OpenSea Studio, which is a no-code platform for creators to easily launch and manage NFT projects, including creating and managing drops, minting NFTs directly to wallets, and enabling buyers to mint from a collection using a credit card. Despite these updates and an attempt to appeal to NFT traders through their acquisition of Gem last year, OpenSea has continued to steadily lose market share to Blur.

Tensor and Magic Eden

Tensor, tailored for NFT traders (similar to Blur), saw a breakout year in volume, likely due to its reward program and increased interest in Solana. Since the middle of October, Tensor has dominated NFT trading volumes on Solana, and at peak held 80% of market share. 

Tensor implemented small quality-of-life updates over the past few months. In October, they introduced a new offer system and a new alert system, and at the end of November, Tensor started supporting Solana inscription and implemented referrals. Although Tensor is a strong product, and reflects what current participants want out of an NFT marketplace (similar to Blur), it’s likely that a lot of their volume comes from their points/rewards system. Currently, Season 3 of the Tensor rewards program is live.

Consumer-facing dApps

Earlier this year, friend.tech became one of the most widely adopted consumer apps in crypto, and has seen over 525M USD in cumulative inflow since its launch and had nearly 73K active users at its peak. The app did particularly well since it streamlined many aspects of the signup process - bypassing Apple’s App Store limitations by being a progressive web app, and abstracting away wallet creation by allowing users to onboard  using a phone number, Google or Apple account, and immediately getting a wallet provisioned. Following the success of friend.tech - not only in terms of popularity but also the revenue (nearly $27M) that the treasury earned through its fee system - there was a month-long period where every other L2 and chain saw a new friend.tech fork launch.

Apart from friend.tech, two other notable decentralized social media apps are Farcaster and Lens. Both are similar to Twitter, in that they allow users to post text and media publicly, for their followers or other users to see. Lens launched Lens v2 earlier this year, which transitioned the app to a profile-centric approach and disallowed following via addresses. NFTs have been optimized to offer a better user experience, where both unwrapped and wrapped states are allowed now for flexibility and compatibility with other protocols. Additionally, Lens V2 implemented an Advanced Referral System, Publication Action Modules, and Delegated Executors.

Farcaster reached 219K users on their app, with 5.9K unique active users and 9.8K casts on average per week. Farcaster has had a few interesting AMAs on their platform - one with Vitalik Buterin and one with Balaji Srinivasan. Recently, Farcaster saw a surge in interest following the launch of a Farcaster native memecoin, called POINTS. Since then, the volume of posts has normalized.

Gaming

Crypto gaming faced significant challenges earlier in the year, partially due to the dwindling activity across crypto, and partially because many users (particularly those in NA and EU) view gaming projects skeptically, often perceiving them as vaporware with little genuine interest. Unfortunately, this was the reality for quite some time: P2E games often suffer from unsustainable tokenomics or unreasonable promises (which they fail to deliver upon).

Parallel TCG

More recently, we've seen crypto gaming gain some steam, as risk-on conditions have returned. Parallel TCG had a particularly good H2 2023, and has become potentially the most popular GameFi project this year. Parallel Alpha cards (which are used to play the game) have had over 80,500 ETH in volume on OpenSea (worth approximately $185M USD as of 12/22). Parallel also launched the pack opening for Planetfall (their first expansion set) earlier in December. The expansion pack contained 120 new playable cards and saw nearly $500K in trading volume. Parallel has also recently been teasing the launch of their mobile game soon. In addition, they are set to launch Parallel Colony, a game that uses AI, allowing users to interact with their AI avatar which acts autonomously based on its memory and biography. Parallel Colony is set to also use the PRIME token, and will allow users to use existing Parallel Avatar NFTs.

Star Atlas and Aurory

Apart from Parallel, there have also been a few important updates in older games. Star Atlas launched early access for their SAGE Labs game in September. SAGE Labs is a fully onchain open-world game built on Solana, which allows users to extract resources and craft in deep space. Since the launch, Star Atlas has steadily grown to 6K daily users, and has seen 150M total transactions over the past 4 months. Another popular game on Solana, Aurory, announced the launch of avatar customization, land exploration, full blitz combat, a new currency, and more with Seekers of Tokane. Aurory launched Seekers of Tokane in alpha for Aurorians and early access code holders in November. 

Apart from the projects above, there are also a few interesting games coming up, such as Project Awakening (developed by CCP Games, the publisher of EVE Online), Shrapnel (an Avalanche Subnet FPS), and Pudgy World (a zkSync-powered open world game by Pudgy Penguins) that should be followed closely over the next year.

Gaming Infrastructure

At the start of December, Xai Network, an Arbitrum L3 developed by Offchain Labs, had their node sale - allowing users to contribute to the running of the network. The Xai Network mainnet is expected to launch in early 2024. Beam, a new subnet built on Avalanche for gaming, launched their mainnet in Q3 2023. Aside from providing gaming infrastructure and tooling, Beam also offers an AMM (Beam Swap) and an NFT marketplace (Sphere). As of now, Beam only offers a dozen games (including upcoming launches) which are mostly smaller card dueling games, but has announced some titles that are more interesting; such as Walker World, an open-world game built on Unreal Engine 5. Lattice is currently developing Sky Strife, a fully onchain real-time strategy game, and Primodium, a fully onchain factory simulation game, using MUD, their Ethereum-based gaming application framework. Previously, Lattice used MUD to build OPCraft, a voxel crafting game (similar to Minecraft) on the OP Stack , back in October 2022. Lastly, Argus Labs announced the World Engine (WE), a blockchain designed for onchain games, in June 2023. WE aims to provide tooling and infrastructure for game developers to launch interoperable games, using a layer 2 sharding architecture that allows games to have their own scalable chain. 

DePIN

Akash

For Akash, 2023 was a breakthrough year with the launch of its GPU marketplace where providers can now offer their underutilized GPUs for LLM training and inference workloads. By expanding its marketplace to include GPUs, Akash can create network effects and “grow the pie” by onboarding new providers who are willing to offer extra compute resources across the range of offerings. Despite a bear market, Akash provider sales are at all-time highs with no signs of slowing down.

Additionally, Akash is currently undergoing a a multi-phased plan that seeks to bring value accrual to AKT stakers, incentivize the growth of the network, and abstract away the need for AKT to procure resources. In Q1 2024, Akash will begin a pilot provider incentive program to onboard highly in-demand GPUs like Nividia A100s, H100s, and L40s. At current AKT prices, this incentive program will likely be over $300M and span multiple years, potentially becoming the largest AI-related incentive program in crypto. Akash’s services are perfectly positioned at the intersection as a general-purpose compute platform and is in the early innings of scaling its GPU marketplace. 

Five High Impact Governance Proposals in 2023

Sushi Kanpai 2.0

The successful Kanpai 2.0 proposal completely changed SushiSwap's tokenomics. Before Kanpai, a portion of swap fees (5 basis points out of the total 30 basis points fee) generated by the DEX was allocated towards purchasing SUSHI tokens on the open market. These tokens were then distributed pro-rata to holders who staked their SUSHI for xSUSHI, creating a direct demand for the SUSHI token. 

This mechanism served as a fundamental link between the value of SUSHI and the growth of the protocol. However, under the new Kanpai 2.0 proposal, all revenue generated by the Sushi protocol is redirected to the treasury instead of being used for market buying SUSHI. This change represented a shift away from rewarding SUSHI holders and aimed to provide necessary resources for the dwindling treasury and team. 

While this decision impacted SUSHI's short-term price action negatively, it was deemed necessary due to the urgent financial needs of the project in order to have any runway. This situation highlights the significant influence governance proposals can have on tokenomics and prices, demonstrating the power of governance over a crypto investor’s portfolio. 

Tornado Cash Governance Attack

The Tornado Cash governance attack on May 20th showed why DAOs need sophisticated actors as delegates and the woes of crypto governance. A seemingly innocent proposal to slash relayers in the system acting in bad faith had a malicious executable hidden in the proposal’s attached code. Buried in the code was a function that gave the exploiter the ability to use 1.2M votes in Tornado governance, more than the true 700,000 votes outstanding. As it passed, the exploiter gained complete control over governance and many of the protocol’s contracts. The bad actor proceeded to dump a large amount of TORN tokens and pocket their ill-gotten bounty. This proposal acts as a cautionary tale of governance attacks. 

MakerDAO Endgame

In October 2022, the founder of MakerDAO, Rune Christensen, successfully pushed through a set of governance proposals to initiate the controversial Endgame plan. Endgame, initially introduced by Christensen in May 2022, has since developed and been heavily modified, with Christensen posting an extensive, updated overview of the upcoming changes in May 2023. Despite the adjustments proposed to the Endgame plan throughout the year, the core idea has stayed the same: decentralization through smaller, specialized entities called SubDAOs. This vision split the community in half back when governance originally voted on the proposal, with proponents favoring a more traditional corporate structure through a board of directors. Many notable governance participants and core developers opposing Endgame consequently withdrew from MakerDAO, which might explain why Endgame still seems to be some time away from launching.

The plan consists of five phases and will establish, among other things, a new brand, governance token, stablecoin, proprietary blockchain, and six SubDAOs, each with their respective native tokens. MakerDAO has prepared for the launch of Endgame throughout 2023, but no specific date has yet been set. However, a notable change connected to Endgame that materialized this year comprises the deployment of Spark Protocol, a money market forked from Aave v3, introducing a use case for sDAI—the tokenized version of DAI deposited in the DSR. Spark Protocol is connected to one of the new SubDAOs called SparkDAO, and users are currently able to pre-farm the airdrop of SparkDAO’s native token by either supplying ETH or borrowing DAI on Spark Protocol. This airdrop will run either until the beginning of SubDAO token farming (Phase 2) or May 2024. Spark Protocol is currently the fourth largest lending protocol by TVL, having grown notably since the end of October.

Based on governance forum activity, most current Endgame development efforts seem to focus on general documentation changes and Phase 3, which will see MakerDAO introduce AI tools meant to enhance and optimize governance tasks. Phase 5—revolving around the introduction of a proprietary blockchain used to host the backend logic for governance security and SubDAO tokenomics—was also widely discussed during the year. In September, Christensen published a forum post announcing he thinks Solana is the most promising codebase to be used for MakerDAO’s new chain. MakerDAO is one of the most established DeFi projects on Ethereum and is seen as highly aligned with the ecosystem, so unsurprisingly, the Solana announcement caused controversy within the community. For a relatively up-to-date deep dive into Endgame, please see this research report on MakerDAO we released in August.

Uniswap Bridge Debacle

The Uniswap BNB bridge debacle perfectly exemplified the politics rampant across many DAOs throughout 2023 and is likely to continue into the future. The story starts with a proposal to launch Uniswap on BNB Chain. It passed Snapshot, but before the proposal was ready to go live onchain, the DAO needed to decide which bridge would be used to facilitate cross-chain messaging so that UNI on BNB would be able to vote on governance proposals. 

A subsequent proposal offering Wormhole, LayerZero, deBridge, and Celer as options went up on Snapshot; Wormhole won the contentious vote. Notably, a16z, an extremely large holder of UNI, could not participate in the vote given their custodian had not integrated with Snapshot. a16z is a large backer of LayerZero, and if they had been able to vote in the Snapshot, would have swayed the vote toward their bridge portfolio company. As the proposal hit onchain voting with Wormhole as the selected bridge, a16z swiftly voted no. Other VCs holding UNI, like ParaFi, are backers of Wormhole. The ensuing lobbying and arguments in the forum, on Twitter, and in private chats would be one for the DAO history books. The other bridge providers in the Snapshot were mostly overlooked. 

Despite the 100+ comments in the forum, many who had voted for LayerZero in the initial Snapshot chose to support the onchain proposal for Wormhole so that Uniswap could launch on BNB with haste before the BSL expired. If the proposal had not passed, it would have been likely that a fork of Uniswap V3 would front-run the real protocol to BNB Chain and cement itself with a first-mover advantage. That said, this situation is a great example of the politics and importance of DAO governance. 

End SNX Inflation

On December 17, SNX governance voted to eliminate SNX inflation. While the inflation was only 5% heading into this vote, the growing supply had been a substantial reason in the past for people's hesitancy to buy the token. Therefore, removing this burden made it easier for people to feel more comfortable holding SNX in their portfolio. Governance proposals can spur activity, change tokenomics for better or worse, and can significantly alter investment theses. 

Final Thoughts

Despite the overwhelming and relentless infrastructure investment from VCs, 2023 was truly the year of the protocols rather than blockchains and infra. dApps across every ecosystem and sector were forced to innovate as users left and native token prices plummeted, and much of the development work that occurred will lay the foundation for the next bull market. 

Perps continued to prove its product market fit by showing resilience through the bear market, with the competition intensely heating up and with diverging architectural decisions between dYdX V4 as an appchain, CLOBs with various components offchain, and various AMM LP models. Other derivatives such as options built back from what was quite frankly, a disastrous 2022. The DEX landscape continued to ship as we seek the answer to LVR, with Uniswap releasing hooks and UniswapX as a potential solution, along with the rise of intent driven solutions. Staking/liquid staking was one of the largest winners across DeFi as TVL almost doubled and LSTs became widely adopted across DEXs and money markets. That growth was not limited to Ethereum and also included other ecosystems such as Solana and Cosmos. Stablecoins struggled this year as we saw USDC supply continue to decrease while grappling with a depeg after the failure of SVB, and other stablecoin innovations failed to deliver on growth promises. The largest innovation in the lending sector goes to LLAMMA and its soft liquidation mechanisms, with existing incumbents making small incremental improvements to increase capital efficiency. Protocols had no shortage of innovation this year. 

2024 stands to be the year where dApp developers will see whether their mechanisms and upgrades will function as intended, and which of them stand to win market share in their respective sectors as users and degens flock back. May the best protocol (“incentives”) win.  

This report is sponsored by Huma.

Re-imagining private credit.

Built by serial entrepreneurs, fintech, and credit veterans, Huma Finance is the leading credit platform, where high performing receivables from real world businesses connect with global capital on-chain. Since our launch in June 2023 Huma has originated over $80M loans with zero default, and enabled partners to launch credit solutions in cross-border payments financing, payment settlement financing, supply chain financing, and green financing. Learn more at https://huma.finance/.