A prevalent lesson from the FTX saga has been that it was a TradFi issue rather than a DeFi one. It is correct that custody concerns only apply to CEXs and not DeFi. Nonetheless, given the severity of the collapse, we believe it is vital to examine why, despite all of its inherent custodial risks, CEXs continue to attract money at such large scales.

Reflecting honestly on this subject reveals that present infrastructural rails are immature on many fronts. When these factors are coupled, the user experience is inferior to that of centralized systems.

Unsurprisingly, the present limitations of these sectors also point to some of the most promising potentials in crypto. In light of this, we'll begin by summarizing the existing situation across the four verticals: access control, interoperability, scalability, privacy, identifying current pain spots and highlighting prospective solutions/projects that are striving to improve them.

Access management

One of the key reasons why individuals choose custodial solutions is the difficult UX of self-custody. Access control in blockchains is currently static and binary. A typical user has either total and exclusive control over her money or none at all. If her private key is lost, there is no remedy. Many people find this to be a highly uncomfortable experience.

Smart contract wallets and MPC wallets are two solutions to enhance the present UX of self-custody. This article is not intended to be a complete comparison of MPC and smart contract wallets.

In this article, we will cover some of the primary emphasis areas and trends in these verticals, as well as highlight solutions that are gaining traction.

Smart contract wallets

The adoption of smart contract wallets is greatly reliant on the underlying chain's VM and/or consensus layer architecture.

Smart contract wallets are now considered second-class citizens in Ethereum. They cannot function on their own; thus need an externally owned account (EOA) such as MetaMask to initiate a transaction and pay for gas in order to activate an action, and they use more gas.

In order for smart contract wallets to be widely accepted, chains must be modified properly. For many years, the Ethereum community has been actively exploring answers here. These attempts are referred to as account abstraction.

Account abstraction (AA)

Account abstraction is a game changer in the making. To understand it, we must first review how Ethereum accounts function.

On Ethereum, there are two kinds of accounts: externally owned accounts (EOAs) managed by private keys and smart contract accounts controlled by code. All transactions must be initiated by an EOA and may result in the execution of arbitrary code by smart contracts.

Account abstraction allows previously segregated EOAs and smart contract accounts to be combined. In other words, they are abstracted away. Let's have a look at how this works at a high level.

AA is essentially the addition of programmability to transaction validity rules. Every Ethereum transaction may execute arbitrary code encoded in smart contracts. Transaction outcomes are, therefore, entirely configurable. However, in order for transactions to be mined on-chain in the first place, they must follow specific requirements of validity. They must have the correct nonce, gas amount, signature, and syntax.

These validity criteria are currently fixed and not configurable.

Account abstraction also enables the programming of validity criteria. Smart contracts, under AA, may not only identify the results of transactions but also whether or not they are legitimate and so operate as the agents that allow them.

Account abstraction motivations

For the end user, AA represents a significant increase in user experience. In the lack of native AA support, even the most basic UX requirements are now out of reach for on-chain users. A typical Ethereum user, for example, must sign three separate transactions using MetaMask to LP on Uniswap - two to authorize tokens and a third to deposit them.

Smart contract wallets paired with native AA eliminate these and many other frictions that now make our life more difficult. AA will enable an on-chain UX approach comparable to Web2 — without losing self-custody.

When does account abstraction occur?

AA is not a new concept. It has been in existence since 2015. However, the Ethereum development community has yet to agree on a specific solution. This is due to the fact that it is a difficult issue to solve and might result in unforeseen second-order effects.

The network is protected by transaction validity constraints; gas prevents spam, nonce prevents replay, and signature prevents theft. It gets more difficult to reason about these safeguards under AA. Proper gas accounting, in particular, might become difficult, exposing the network to griefing and/or DDOS attacks.

Another issue that arises while introducing AA is the loss of backward compatibility with current programs.

Given these hurdles, it's reasonable to anticipate Ethereum to support AA natively in the near future. The good news is that L2s do not have to be backward compatible. They also have a lower bar for social cooperation. This enables them to respond quickly and benefit from all of the effort done on AA over the years.

Account-abstracted layer 2

zkSync and StarkWare were among the first L2s to have native AA support in their protocols, giving their smart contract wallets a significant advantage. Argent has been by far the most active promoter and advocate for AA for many years. When it is released, it will support StarkWare as well as zkSync. Cartridge is an additional wallet that uses AA to improve the on-chain gaming experience on StarkWare.

Source: Cartridge

Fuel is another chain that supports native AA in its VM. FuelVM is intended to enable the previously described UX-enhancing features like batched transactions, native multi-sig support, privacy-improving features, and so on.

Cross-chain interoperability

Today's multi-chain environment demonstrates that a single chain cannot meet everyone's demands. Thus, cross-chain interoperability is important to scalability.

Unfortunately, bridge security remains a key source of concern for the sector. Bridges are dangerous, as the previous two years have shown. Bridge hacks have cost about $2.5 billion in total damages.

The majority of these hacks were caused by implementation errors, but some were caused by key compromises that were part of the bridge's security architecture (no fallback mechanism).

Bridges that are Natively Verified vs. Bridges that are Externally Verified

When analyzing bridge designs, it's easy to separate the security concept and implementation. Moving ahead, we would suggest that the better (and maybe more practical) strategy is to evaluate them both.

Bridge designs may be divided into three categories: third-party, light client, and rollup bridges. Third-party bridges are validated externally, while light client and rollup-based bridges are validated locally.

The majority of bridges built in the previous two years have been externally validated. These bridges are often built and maintained by haphazard teams, each with its unique design. In terms of code examination, they have little to no collaboration. Furthermore, since businesses fight for market share, they are more likely to enhance TVL during the initial few days of introduction by offering incentives. This creates an ideal setting for black hats.

This is why we no longer believe it is wise to evaluate bridge designs in isolation from how they will be executed. We've established that the threat model consists of nation-state-level black hats. While we can never be confident that no vulnerabilities exist, the successful solutions will most likely be those that promote the greatest collaboration and code examination. This takes us to the topic of natively validated bridges.

Bridges with native verification are becoming more common as standards.

Standardization is the cure to bridge hacking. Natively validated bridges (light client and rollup-based) have distinguishing qualities that make them more likely than externally verified bridges to emerge into standards.

Natively verified bridges not only serve as general message-passing protocols, but also as primitives for mobile wallets, quick synchronization, decreasing dependency on centralized RPC services, and so on. Natively validated bridges are depended on and evaluated by a greater number of teams and developers in their particular communities since they are ecosystem-centric. Their roadmaps are centered on building ecosystem-wide norms that maximize the mutual benefit of ecosystem members.

IBC is unquestionably the most successful example of this. IBC is not immune to implementation issues, but its development benefits enormously from the mindshare and attention it gets from a varied set of Cosmos teams. Given how soon a significant event was identified and fixed, we may consider this a crucial determinant in its survival.

The traction of IBC

In 2022, IBC was undeniably successful, becoming the definitive bridge for Cosmos. IBC is now used by 53 chains to communicate with one another. IBC contributed more than $0.7B in cross-chain volume in the previous 30 days, ranking third behind Multichain and Arbitrum.

Because of its current network and lindy effects, there is growing interest in expanding IBC to Ethereum and other ecosystems.

From IBC to Ethereum

The cost of verification is a significant problem for light client-based bridges. The key limitation here is that the destination chain must validate the source chain's header for each block when the source chain delivers a message. This entails validator signature verification.

For instance, if the source chain has 150 validators, the destination chain checks 100+ signatures (>2/3 validator votes) on-chain for each block where cross-communication occurs. This may be very expensive if the destination chain does not natively support the source chain's signature method and has expensive gas.

This is the main reason why the Cosmos>Ethereum IBC link has always been impossible. However, thanks to recent zk-tech breakthroughs, there is now a plausible avenue to introduce IBC to Ethereum!

IBC with Zero Knowledge

Instead of Ethereum smart contracts directly confirming the header on-chain, off-chain provers employing beefy processors execute the processing. These provers then create a concise validity proof that Ethereum smart contracts may easily verify. Consensus proofs are the name given to this approach. Electron Labs and Polymer Labs are two notable pioneering teams in this field.

To be clear, the usage of zk-proofs does not alter the bridge's current trust architecture. The proofs are used to lower the verification costs of demonstrating consensus without adding additional trust assumptions. For the protection of money, our light client-based bridge would still depend on source-chain consensus (>2/3 validators). As a result, it would still be less trustworthy than rollup bridges (which prove every state transition). However, because of its restricted scope, demonstrating consensus may acquire traction much sooner with the widespread use of zk-proofs.

The addition of zk-proofs to IBC bridges is highly interesting since it can introduce IBC to any smart contract blockchain that lacks native IBC capability. Using recursive zk-proofs, the consensus from many chains may potentially be aggregated into a single proof in the future. This would lower the amortized costs of cross-chain communications and become the final answer to IBC's relayer incentivization dilemma. Perhaps this offers us an idea of what IBC will look like in its final form.

Ethereum light clients using ZK

So far, we've spoken about how Tendermint consensus can be inexpensively validated in various ecosystems. But what about Ethereum consensus verification?

For PoW Ethereum, an on-chain light client was out of the question. However, with Ethereum's shift to PoS and advances in zk-proofs, this is now conceivable.

Indeed, pioneering teams such as Succinct Labs and zkBridge have made remarkable progress this year. Both teams are nearing the completion of their goods. In testnet, Succinct Labs has previously shown a two-way on-chain light-client bridge between Ethereum and the Gnosis chain. Gnosis was selected because it uses the Ethereum consensus system. Similarly, zkBridge has created a preliminary version of an Ethereum lite client based on SNARK.

Cross-Chain applications

Cross-chain applications are one trend we anticipate seeing next year as we shift our emphasis to the application layer. The application layer is currently multi-chain but not cross-chain. The first significant cross-chain applications are expected to be released next year. This is manifesting itself in a variety of ways.

Siloed deployments are collaborating with one another.

Major dApps may now be found on many chains. This is notably frequent in the EVM realm, where blue chips like Uniswap, Aave, Synthetix, and Compound are used. However, most of these occurrences are compartmentalized.

This is due to a variety of factors. For starters, it is mostly a relic of the bull market, when teams sought market share above all else, with little attention paid to cross-chain composability. DeFi protocols seeking bridge partners have in-depth conversations but were unable to commit due to the lack of obvious winners. Bridge hacking aggravated the situation even further. However, given that the bull market frenzy seems to be gone, the scenario may shift next year.

Chainlink's cross-chain interoperability protocol, CCIP, will be a driving force in this regard. Because significant communities like Aave, Synthetix, and Compound already depend on Chainlink for price feeds, it's simpler for them to choose Chainlink as a cross-chain interoperability partner. This would strengthen Chainlink's position in DeFi.

ccip of chainlink

LayerZero is another protocol that fits in well here. Their new integration with Chainlink oracles may be a trigger since apps may spin up bridges using Chainlink oracles without requiring authorization. Sushi is already utilizing LayerZero thanks to its connection with Stargate earlier this year. Another intriguing derivatives platform that will employ LayerZero for cross-chain services is Rage Trade.

THORChain is a cross-chain protocol that is ideally positioned to leverage its previously proven market fit. THORChain is expanding its partnerships with current DEXs and wallets. The idea here is to abstract THORChain from user flow. Users may use their preferred wallets/DEXs without understanding that their cross-chain swaps are being routed via THORChain. Pangolin, Trader Joe's, and Kyber were all integrated early this year. Recently, a significant integration with Trust Wallet, one of the largest wallets in terms of user population, was unveiled.

THORChain is increasing its cross-chain capabilities in addition to integrations. It just released a savings product that may be compared to a decentralized version of BlockFi. It is the only significant DeFi product that presently provides yield on native Bitcoin. THORChain intends to provide a new cross-chain lending/borrowing solution with zero liquidations and interest next year. THORChain is making critical efforts toward being a commercially viable protocol by adding additional integrations and products.

Axelar is another protocol that is well-positioned in this space. Axelar is a hub for general message-passing rather than an app-specific protocol like THORChain. Axelar is in a unique position since it supports Ethereum, other EVM chains, and IBC. So far, it has made a significant contribution to Cosmos by serving as a direct link between Ethereum and a number of Cosmos appchains. Axelar was chosen as the canonical bridge for Osmosis shortly after its inception early last year, and it has since been among the top chains in terms of IBC traffic. Axelar might gain substantially from higher Cosmos adoption next year.

Finally, a new group of dApps will begin cross-chain operations right away. Some famous examples are Mars Protocol, Prime Protocol, Composable Finance, and Rage Trade.

Both Mars and Prime are cross-chain lending/borrowing programs that allow users to borrow on one chain against collateral on another. Mars will be a Cosmos appchain that will use sophisticated IBC features including interchain accounts and queries. Prime will provide a comparable use case, but will target EVM L1s and L2s and will leverage Axelar for cross-chain interoperability.

Scalability

Scalability is an elusive goal. Blockchains, like the internet, will continue to struggle to grow as more users join.

To put it another way, scaling solutions take many shapes and forms depending on the amount of customer demand. To handle the present user load, many blockchain designs coexist today.

Nowadays, general-purpose monoliths dominate user adoption. As time goes on, we anticipate that this will progressively alter and that user acceptance will shift toward appchains and modular chains. Given their maturity, we anticipate appchain adoption in the near future, maybe within the next 1-3 years, and see modular chains as the end-state architecture in which blockchains become ubiquitous.

This is not to argue that general-purpose monoliths will go away anytime soon. While we see a movement toward modularity and appchains over time, we acknowledge that general-purpose chains provide synchronous composability and permissionless innovation in sandboxed contexts. We do not anticipate things to go gone totally.

Indeed, we anticipate that these many designs will coexist for a very long period. Each category will have winners and losers. In this part, we will identify ecosystems and initiatives that we believe will prosper in their respective categories.

Despite all of the blows it sustained during the FTX meltdown, we still believe Solana to be the best general-purpose monolith.

What happens next for Solana?

The latest FTX collapse had the greatest impact on the Solana environment. To say Solana DeFi has had a difficult few months is an understatement. Solana's biggest DeFi protocol (Mango Markets, >$100M TVL at the time) was exploited in October, followed by the FTX collapse a month later. There were also issues about FTX-led Serum during the aftermath, so the community banded together to fork it and seize control, installing a new CLOB DEX named OpenBook. Because the majority of Solana DeFi was based on Serum, several DeFi protocols were forced to pause/go down owing to a lack of liquidity. Solana DeFi has been completely reset, with NFT activity responsible for keeping Solana users for the time being. So, should we write off Solana as dead, or are there catalysts in the horizon?

To begin, we should note that Solana is becoming a more modular environment in and of itself, not in terms of the Solana chain, but with the Sealevel Virtual Machine (probably Solana's biggest invention) becoming a rollup standard through Eclipse and Nitro. If you think that the EVM is a large component of Ethereum's moat, then the SVM getting widely accepted should be no different for Solana, and the evidence supports this.

The Solana Web3-developer package is the most extensively used outside of Ethereum. Adoption has also increased week after week, rising from 87k weekly downloads a year ago to 200k in June, 365k in the first week of November, and 420k in the first week of December. Despite the crash of FTX, it has maintained its constant ascent. Aside from SVM, there are reasons to be optimistic about Solana L1.

First, network performance has improved over time. Over the previous half-year, average block times on Solana have decreased dramatically. While part of this may be attributed to a larger vote/non-vote transaction ratio (votes are less computationally intensive), the network has also received various software updates. The NFT minting/spam problem that brought Solana down numerous times has not had the same effect on the chain, and it should only get better with solutions like QUIC, fee markets, Jito, and Firedancer all coming online.

Source: Solana breakpoint

During the FTX crash, there were still some performance difficulties (such as publisher oracle updates not coming through), but the chain performed considerably better than it would have six months ago and remained operational despite the pandemonium and enormous bot liquidations that stressed it. While non-vote TPS is down, this is mostly due to decreased general activity, which we are witnessing across all chains, not just Solana.

Despite this, Solana is substantially ahead of all other chains, with 300 non-vote TPS and eliminating the 100 from oracles. When compared to other blockchains, their 15-20M non-vote transactions per day are still in a class of their own.

TPS is far from an ideal number, even when votes and Oracle transactions are excluded, but it is still a rough indicator of total demand. We also shouldn't dismiss oracle transactions entirely since Solana's sub-second block times + low fees allow these quick oracle updates (which no other chain has live today) and, as a result, unique products — particularly in the derivatives market.

We would like to calculate units per block as a demand measure since it represents both the number of transactions and their complexity. Seeing this number rise over time will be a solid indicator of actual Solana use. With Serum down and general demand for block space declining in recent months (see Ethereum gas costs), this measure has reached a new low. Solana blocks have a maximum of 48M compute units per block, as seen in the chart below.

It's no secret that Alameda/FTX was a huge part of Solana's growth, and we can definitely state with certainty that a lot of the SOL price activity in 2021 was due to them rather than real/organic factors. However, although the loss of Alameda would have an immediate impact on on-chain liquidity, it is not a fatal blow. Solana is unusual in that it was created from an engineering standpoint rather than an academic one. While some of these design choices have damaged them in their early years, they have also opened unique use cases that not all chains can reach. TPS (excluding votes) are still much greater than any other chain, and the aforementioned enhancements should continue to improve network stability. Still, Solana will almost certainly go down again at some time.

Cosmos is gaining momentum in the Appchain World.

Restriction of access to the state is a key component of appchains. In contrast to general-purpose chains, which enable anybody to start new apps without authorization, appchains pick which programs to run via social coordination. Likewise, cross-chain communication necessitates social cooperation. To build IBC connections, chains shake hands.

There is an unavoidable trade-off here. Appchains lose the agility of permissionless innovation while gaining control over their apps. Furthermore, companies get to optimize their chain for the specific use case(s) they wish to solve, resulting in a more dependable UX.

Incentives are another important part of the appchain concept. Revenue often leaks to the underlying gas token on general-purpose networks. Even the most popular dApps fail to capture the majority of the revenue they create.

We can go no further than CoinGecko to demonstrate this argument. It's amazing how uncommon it is for pure app tokens to be in the top 100. By integrating app and gas tokens, appchains address this incentive imbalance.

Cosmos comes to mind when we think about appchains today since the Cosmos SDK is the most production-ready toolbox for the permissionless building of a new chain from the start. In the long term, we anticipate that appchains as rollups will acquire traction as the technology matures. Today's significant live and forthcoming appchains include

dYdX: A pure DEX appchain, it is now operational as a StarkEx Ethereum L2 and will migrate to Cosmos as an independent chain in the new year. Through dYdX, there is a big potential for the Cosmos ecosystem in 2023, and FTX's collapse (although clearly tragic) could not come at a better moment. Trading is being "pushed" on-chain as centralized entities all across crypto proliferate, and dYdX has the most benefit as the leading DeFi derivative DEX today. The launch of dYdX will coincide with the introduction of the Cosmos Hub's "asset issuance" consumer chain. Circle will issue USDC on this chain, with dYdX being the biggest initial client (dYdX presently holds $430M USDC on its StarkEx L2). This will significantly increase the quantity of native USDC on the Interchain, which has previously been absent in Cosmos and IBC. To summarize:

On Cosmos, dYdX growth is beneficial to native USDC growth.

The rise of native USDC on Cosmos is beneficial to the expansion of activities via the Hub's economic zone as an Interchain Security provider of the asset issuance chain.

The Hub's expansion as an ICS provider is beneficial to the adoption of additional consumer chains, and native USDC enables bootstrapping them faster.

Native USDC development and the expansion of the Hub's economic zone are beneficial to general liquidity and the growth of other appchains in the ecosystem.

Sei and Injective are "DeFi-optimized" chains that fall between pure appchains and something more general purpose. Orderbook logic is incorporated into the basic layer, and permissioned applications are developed on top. The main distinction between dYdX and these two is that dYdX is a pure appchain focused on derivatives, while Sei and Injective are merely the infrastructure layer on which things like perps, options, stablecoins, and more may be created. Sei aspires to be the foundation layer for both the SolanaVM and MoveVM rollups.

Privacy

The transparency of public blockchains is a two-edged sword. Openness is a characteristic and one of blockchain's major benefits in terms of solvency and transparency. Being able to audit procedures in real-time and assess their solvency in a matter of seconds eliminates the black box opaqueness that plagues centralized institutions – a bitter lesson learned in 2022.

Following the FTX crash, we saw DeFi protocols like dYdX provide proof of reserves on Twitter, emphasizing why DeFi is worth constructing. With completely collateralized on-chain protocols, you can't get into a Celsius or FTX scenario. A one-click audit of dYdX's full reserves will never be possible under CeFi, regardless of how rigorous the PoR procedure is or how diligent and reputable the audit company is.

However, a completely transparent financial network is not without drawbacks. When it comes to personal financial privacy, openness is a bug, not a feature. Having your history 100% on-chain not only exposes winning trading tactics to the public (leading to alpha decay), but it may also be hazardous to your personal safety. With the collapse of FTX being a recurring topic in this article, we go no further than prominent media sources such as Bloomberg, The New York Times, and Financial Times filing to publicly dox all creditor information.

In this case, publicly revealing creditor details is extremely dangerous because it has the possibility to leak not only customer information and account balances at FTX, but also on-chain addresses used, giving bad actors the power to tie someone's entire financial background together, especially when this data can be cross-referenced with other leaks. We shouldn't have to explain why it's an issue for someone's whole financial history and home address to become public. It's evident that we need to find a middle ground in which protocol audibility and solvency are accessible 24 hours a day, but personal data remains private and hidden from public scrutiny. We anticipate that privacy will resurface as a prominent subject, with the conflict being both technological and legislative in nature. There are similarities to the fight against encryption.

Penumbra: As a completely private DEX, Penumbra serves as a safe haven inside the Cosmos/IBC ecosystem. Their shielded swaps avoid MEV-like front-running and sandwich attacks and use sealed-bid batch execution to clear orders at a single price. Their v1 AMM has a robust liquidity architecture, allowing market makers to implement strategies privately without exposing them to the public. Future use cases (such as private governance voting on Cosmos chains with Interchain Accounts) will enable users to vote secretly from their Penumbra accounts on any linked chain, thanks to the IBC connection. While public trade/transaction history is entirely secret, users may decrypt their whole history on demand. Penumbra is set to launch in 2023.

Aztec: An Ethereum-based privacy-first zk-rollup, Aztec differs from existing zk-rollups such as StarkNet, Scroll, and zkSync in that it leverages zero-knowledge technology for privacy rather than scalability. Aztec Connect, their main product, is distinct from previous rollups in that, rather than having its own smart contracts, dApps, and liquidity, it just operates as a VPN to access the Ethereum L1. To enable Aztec users to utilize their protocol discreetly from Aztec, Ethereum protocols may interface with the Connect SDK. Connect functions as a private pool; several users on Aztec submit encrypted transactions like swaps, which are then batched together and processed on Ethereum. This batching not only provides secrecy but may also lower gas expenses by spreading the cost among several users and transactions. Because all liquidity remains on Ethereum, there is no fragmentation, contract redeployment, or need to re-audit any code. Their product, zk.money, began with basic transfers but has now developed to include a wider variety of DeFi activities such as staking, swaps, and borrow/lend. Expansion to private NFTs and the debut of their new zero-knowledge language Noir are now in the works.

Source: zk.money

Conclusion

The negative externalities of MEV, bridge exploits, and the FTX collapse took center stage in 2022, making it a difficult year for the sector. On the other hand, considerable progress is being made across all important infrastructure fronts, including UX, scalability, interoperability, privacy, and censorship resistance. Today's crypto infrastructure is not yet ready for broad use, but it is on its way, and 2023 will be a watershed moment.

DeFi was in 2020, NFTs were in 2021, and reality hit in 2022. We feel that 2023 will be a year for infrastructure to return to its origins, with a renewed emphasis on the infrastructure layer. How else will we scale cryptocurrency to the billions?

Disclaimer:
This article is intended for and only to be used for reference purposes only. No such information provided constitutes advice or a recommendation that any investment or trading strategy is suitable for any specific name.

Crypto infrastructure narratives in 2023