Through the utilization of computation proofs, including Zero-Knowledge (ZK) Proofs and Trusted Execution Environments (TEEs), the integrity of transactions can be assured. These proofs, once posted on Ethereum, provide not only a verifiable timestamp but also a secure, unalterable history of computations.

TEEs stand out by offering a means to delegate computations to third-party providers without compromising data privacy or the integrity of operations. It allows code to be executed in a trusted, isolated area of a specialized processor in the server, without the ability for even the owner or administrator of the server to inspect memory, or to tamper with the code.

It enables verifiable confidential computation by providing tools with which anyone can verify that the code is running in an authentic TEE processor. This feature is particularly valuable for decentralized applications (dApps) demanding high levels of integrity, confidentiality, and resistance to censorship. By integrating with blockchain technology, TEEs enrich smart contract environments with secure, off-chain computing capabilities.

The necessity for verifiable off-chain computation stems from the inherent mistrust in opaque, external systems. By submitting computation proofs, whether through TEEs, zero-knowledge, or optimistically, protocols can mitigate trust issues, allowing for off-chain computations without compromising on security or transparency (more info in this post on the different trust models and their tradeoffs). This verifiable approach is critical for dApps requiring off-chain computing resources, ensuring data privacy and integrity without forsaking trust.

While coverage and hype have been around zk-coprocessors, TEEs are currently more practical from a latency and cost perspective, particularly in applications demanding both high performance and privacy, such as AI and ML frameworks. Unlike ZKPs that confirm the correct execution of a program, TEEs ensure that only the authorized configuration and execution occur, offering a unique layer of security for certain dApps.

TEEs also offer a crucial feature known as remote attestation (“RA”), which significantly enhances their utility by enabling trust verification in off-chain TEE nodes. This process allows entities, including smart contracts, to verify the integrity of a TEE node, thereby establishing a secure communication pathway. This mechanism is crucial for thwarting attempts by malicious actors to impersonate TEE operations, thereby safeguarding transmitted data.

An example of TEEs and Remote Attestation in production is illustrated by Puffer Finance's anti-slashing technology, heralded by Binance Research as a "key innovation."

This system includes a "Secure-signer" that utilizes TEEs to prevent slashing events, such as double-signing, and a Remote-attestation-verifier (RAVe), enabling on-chain validation of a TEE's execution environment. Puffer's RAVe facilitates permissionless validator integration by authenticating TEE-produced evidence and managing validator public keys on the blockchain.

I believe TEE coprocessors, and particularly Automata, are undervalued given their multifaceted applications:

1. Trustless MEV-Supply Stack Development with Flashbots: Instead of relying on a trusted service (Flashbots, as a centralized operator, running MEV-share or MEV-boost), the off-chain service can be executed within a TEE. Automata is facilitating the development of a trustless TEE-based MEV supply stack for Suave through its on-chain remote attestation verifier.

2. Enhancing Layer 2 Scalability and Security: Through the implementation of SGX (TEE) alternate proofs, Automata enables a multi-proving model that boosts the liveness and security for optimistic and zk rollups. Scroll's adoption of Automata's technology highlights the growing confidence in TEEs.

3. Enabling On-Chain AI and DEPIN verification: Automata's technology paves the way for on-chain AI by making Large Language Model (LLM) outputs accessible to smart contracts. Furthermore, many DePiN projects need to verify on-chain that their inclusion to the network is being done on a unique and genuine device (attesting that you are using a mobile phone, or a computer).

To achieve this, Automata is launching a modular attestation chain, which will go live as an Eigenlayer AVS. This app-chain will aggregate remote attestations from off-chain computations, making them publicly verifiable and easily accessible. It will also involve the implementation of new tokenomics that will introduce token sink mechanisms.

SUAVE 🤖

The current existing MEV supply stack relies on off-chain trusted providers such as Flashbots (running MEV-share and MEV-Boost) and CowSwap (offering MEVBlocker).

MEV-share (aka Flashbots Protect), for instance, enables searchers to access user orders through an Order Flow Auction (OFA), ensuring value generated from MEV activities is returned to users. A critical aspect of its system is the 'matchmaker' role designed to match user txns with searcher bundles effectively. This is currently a permissioned position held by Flashbots, which aims to prevent bidding wars that could reduce the value returned to users and guarantee that bundles meet specific validity conditions, ensuring MEV gains are rightfully distributed.

Flashbots has a solution for its centralization concerns. They aim to create a platform (called “SUAVE”) where these trusted actors can be replaced by a set of specialized privacy-preserving hardware, TEEs alongside a censorship-resistant chain managing them. It will be something similar to a “TEE Party”.

SUAVE's emphasis on privacy serves dual purposes:

1. Decentralizing the MEV-supply chain by enabling users to internalize the value of their MEV without relying on permissioned environments like off-chain auctions.

2. Introducing "programmable privacy," allowing users to control the release of information to relevant parties during MEV negotiations, promoting scalability and avoiding closed, permissioned markets.

Implementation of programmable privacy involves interconnected TEEs called "Kettles," where users interact with SUAVE's applications via Confidential Compute Requests (CCRs). This approach shifts sensitive data computation off-chain, minimizing trust requirements.

SUAVE's implementation involves two stages:

• Centauri: Initially, execution nodes operate in a trusted or centralized manner, maintaining the status quo.

• Andromeda: Execution nodes transition to TEEs (SGX), reducing reliance on centralized operators and expanding participation. This intermediate layer is expected to launch with SUAVE's mainnet in Q2 '24.

  

  Jon Charbonneau 🇺🇸

  @jon_charb

  

  OH: “SUAVE is DEPIN for SGX”

  53

  7:00 PM • Mar 4, 2024

Future plans for SUAVE include augmenting SGX with cryptography (ZK and FHE) to further reduce trust requirements, although this is not immediate.

This network of TEEs still needs some on-chain implementation to: a) verify that computation and the code being run off-chain on these TEEs nodes is what it is supposed to be, and b) establish a secure channel to them.

Flashbots explains this concept in their latest blog post where they walk through an implementation of a simple TEE co-processor:

We need to use the remote attestation feature of TEE so that our smart contracts can verify that a message came from a TEE. To do this, we add a new precompile so that Solidity code is allowed to generate a remote attestation. It's a bit like Gramine for Solidity, since we are exposing a fairly low-level interface directly to the smart contract environment. While generating an attestation requires an SGX processor, verifying the attestation occurs entirely on-chain. The remote attestation is just a certificate chain, and the verifier is written in Solidity. We specifically use the Automata-V3-DCAP repository, which is an open source Solidity library for verifying these.

I have also included some screenshots from Flashbots´ latest blog writing with highlights of Puffer and Automata mentions:

While both Puffer Finance and Automata are mentioned by Flashbots, Automata's verifier is recommended due to impending deprecation of EPID by Intel. Additionally, there have been suggestions from Phala Network (Jun Jiang) about transforming DCAP verification into a ZK-proof, which, though slower, may offer benefits in on-chain validation. This presents a trade-off spectrum, where while current zk VMs may be slow, you don´t need to verify DCAP quotes frequently.

Chua Zheng Leong

@melynx

Oh wow! Nice work! I'm also working on a version myself. Hopefully I can wrap it up soon and releasing it.

1

1:15 AM • Feb 14, 2024

There's also consideration of using DCAP verification as a benchmark for zk, as suggested by Andrew Miller.

To summarize, SUAVE should be viewed as a big tailwind to put TEEs on the map as they integrate them into their chain and replace current trusted entities. Each SuApp, like Order Flow Auctions (OFA) or blockbuilders, will operate within a TEE forming a network called "Kettles." Verification of TEE computations and code integrity will be ensured through on-chain Remote Attestations (RAs), with Automata being the recommended solution by Flashbots for developers working on SuApps experimentation.

Patrick 🌊

@cosmosp4tr1ck

Our TEE event with @socrates1024, @SecretNetwork, @OasisProtocol, and @PhalaNetwork led to some great conversations!

@MarlinProtocol and @AutomataNetwork also made an appearance, and a new TEE research working group was established. I hope to see something big come out of it!

51

8:08 AM • Mar 1, 2024

Meow no MEV

@meowrpc

Today me together with @0xAnatolii and @melynx @MuratLite @tkstanczak @DelaneSays Chang-Wu Chen participated in a panel discussion about all the mev stuff and how to take care of users in this dark forest. Organized by @AutomataNetwork.
Busy week ahead.

18

7:26 AM • Sep 12, 2023

L2 multi-proving 📜

L2 multi-proving offers a promising solution to the security concerns surrounding Layer 2s, especially as more assets migrate from Ethereum to L2s. Vitalik has emphasized the importance of addressing security issues within these complex systems, highlighting the potential catastrophic implications of bugs within single-prover validation setups. The need for diverse prover mechanisms becomes paramount in ensuring the integrity and security of L2 txns.

Scroll's collaboration with Automata represents a significant step towards enhancing the security of L2 networks through multi-proving. By incorporating a TEE prover as a secondary mechanism alongside traditional ZK proofs, Scroll aims to mitigate the risks associated with bugs and vulnerabilities inherent in complex software systems.

The rationale behind adopting a multi-proving approach lies in bolstering the security of L2s without compromising on finality time or significantly increasing transaction costs. By diversifying the validation mechanisms, Scroll aims to minimize the likelihood of invalid transactions infiltrating the system.

The design space for a multi-proof system on Scroll revolves around three key objectives: enhancing L2 security, maintaining reasonable finality times, and introducing minimal additional costs to L2 txns. While exploring various options for a secondary prover, including fraud proofs and alternative ZK provers, the adoption of a TEE prover emerged as the most viable solution. With negligible overhead associated with proving processes, TEE provers provide a faster and more cost-effective solution, thereby aligning with Scroll's objectives of enhancing security while maintaining operational efficiency.

The specifical implementation of the SGX (TEE) prover leverages Automata´s on-chain remote attestation verifier. It entails using a smart contract as a remote party to create a public and transparent on-chain anchor that enables trust composability — other smart contracts can also trust the computation from the enclave.

Forward-looking other zkL2s like Taiko, Linea, or even opL2s could be integrated and benefit from more diverse prover mechanisms.

Conclusion

Main liquid bet to get exposure to TEE coprocessors and take advantage of a potential rerating arising from ZK coprocessors launching soon at (expected) high FDVs is $ATA.

Disclaimer: The following statement does not represent Spartan's views. It is a personal thesis and post. I welcome discussions on any thoughts and greatly appreciate any feedback received.

I firmly believe Across will be the biggest winner (in terms of growth and share gains) of the bridge landscape in the 1-year horizon (predicting further than one year almost never works, so I´ll leave that to traders, gurus, and threadooors).

STRUCTURE:

1. Bridges are a growing vertical.

2. Bridge aggregators play and will continue to play a big role in the bridge landscape.

3. Across is the most advantageous bridge if aggregators grow and keep taking share from the market.

4. Quantitative analysis.

5. Multiples – Valuing cryptoassets on a relative basis.

6. Absolute valuation based on TAM growth and share gains.

If anyone wants to access the spreadsheet (where I´ll be putting more graphs and stuff) DM me on twitter. I have a paywall of $10 on it, but really think it´s worth it. The data updates daily thanks to Artemis and aggregates KPIs from all of the bridge landscape.

tumilet

@tumilett

Who wants access to this? It´s a fully automated dataset for the bridge landscape, which includes bridge sector-specific metrics that I "created". All data and graphs can be used to extract various insights.

Note that there is a $10 paywall, which IMO is definitely worth it.

49

1:19 PM • Jun 23, 2023

1. BRIDGES ARE A GROWING VERTICAL

Bridges are a growing vertical, which has grown volumes 4x since January 2023 (although primarily driven by Stargate´s L0 airdrop expectations),

and there are many TAM expansions on the short-term horizon:

1. EIP-4844 (reduce L2 gas fees and may drive cost-sensitive users to L2s, for which they will need to bridge to existing L2s (Arbi and Op).

2. Many important L2s are expected to come in the following months: BASE, Taiko, Linea, Scroll, Mantle. All of which need bridge support.

I expect bridge volumes to grow 50% y/y since the L365D. Again, growth comes from 1) existing chain bridge transfers, 2) new ecosystems.

2. BRIDGE AGGREGATORS PLAY AND WILL CONTINUE TO PLAY A BIG ROLE IN THE BRIDGE LANDSCAPE

The most important metric in the bridge landscape that is not being tracked is aggregator penetration: what % of the total bridge volume gets rooted through aggregators.

The market is dormant, and aggregators change the dynamics. Users who were “loyal” or accustomed to using incumbent bridges that weren´t fast and charged high fees, have now discovered a new bridging experience, which is more cost-effective and “democratic”.

For more info on why bridge aggregation makes sense, check this out:

Crypto Aggregation Theory (ft. Bridges)

"" is published by Arjun Chand in LI.FI Blog.

https://blog.li.fi

Anyway, this ratio serves as a metric to gauge the “penetration” of aggregators, made with data from Across, Hop, Synapse, Stargate, Celer, Connext, and Multichain.

Two important things to say about these graphs:

1. In December 2022, there was a notable spike in volume due to the launch of Jumper's V2 and a surge in Bungee's volumes. These protocols attracted significant bridging activity as users started to speculate about a retroactive airdrop of $LIFI and $SOCKET tokens. Currently, the remaining volume consists of organic transactions, which initially appears to have slowed down.

2. In order to provide clearer data, I have excluded Stargate's overall volume from the second graph. Stargate, being a widely used frontend and smart-contract platform, attracted most users for L0 airdrop activities after it achieved a $3B valuation in April 2023. In this graph, we can observe that market penetration has not slowed down; instead, it has increased from 7% to 24% over the past three months.

Having said that, I expect this penetration ratio to grow in the following year (to 40%). As aggregators improve their service offerings and bridge integrations, and as they develop their own network effects, it becomes logical for users to rely on aggregators for their bridging activities. This way, users can abstract the need to know which bridge supports specific chains or if it has sufficient liquidity, among other factors.

3. ACROSS IS THE MOST ADVANTAGEOUS BRIDGE IF AGGREGATORS GROW AND KEEP TAKING SHARE FROM THE MARKET

Across is the fastest and/or cheapest bridge, and as such, gets selected by users and aggregators as the preferred route for most of their transfers. If aggregators penetrate the market, Across wins, and I see that happening.

Based on the last week, 58% of Across´ volume comes from LIFI and Socket.

So, how can someone get exposure to this expected growth in Across? They have a governance token - $ACX, let’s check how has the token performed historically with the evolution of the underlying protocol KPIs.

4. QUANTITATIVE ANALYSIS

Before we delve into the details, let's examine a couple of graphs that highlight an interesting trend. Specifically, focusing on the pair of Across-Hop-Synapse-Connext-Celer, we can observe that Across has been the sole gainer since the beginning of the year.

5. MULTIPLES – VALUING CRYPTOASSETS ON A RELATIVE BASIS

When it comes to valuing cryptoassets, a relative valuation method can be employed. This approach involves determining the price of a target asset in relation to another crypto asset or a basket of assets that belong to the same vertical and exhibit similar characteristics.

Let's focus on bridge multiples as an example. Multiples can be applied to various key performance indicators (KPIs), but it is particularly meaningful to utilize them for sector-specific KPIs in the context of bridges. These include metrics such as total value locked (TVL), bridge volume, and revenue. For more detailed information, I recommend referring to another post I created on my Bridge Token Dashboard:

https://mirror.xyz/tumilet.eth/UnSe-VqOIzfvvNIgUL8UMJHAeesvMcTAolZgq3tEJ0o

The price of ACX (in USD terms) has been tracking relatively close the evolution of the implied price from TVL comps multiple. The correlation is not perfect, but closely correlated, and it has had an average discount to fair value since January 1st, 2023, of 20%.

In my opinion, total value locked (TVL) serves as a lagging indicator for LP-based bridges, as well as for decentralized exchanges (DEXs). TVL tends to grow after the volume and fees have increased, as LPs primarily seek for yield.

Let's now analyze the performance of the implied price derived from Volume and Fees comps relative to the market price of ACX.

Although the results are not particularly revealing, the key insight here is that, despite my opinion that TVL shouldn't be the primary KPI metric for bridges, the market continues to price bridge tokens based on TVL. This is evident from the historical average discount to "fair value" based on comps, which stands at 20%.

The disparity between the implied prices derived from FDV/VOL and FDV/FEES multiples stems from the varying fee structures of each bridge. While Across has witnessed growth in volumes, leading to a higher "fair value" based on this metric, its fee growth has not been as significant.

The reason for this can be found in the daily bridge fees as a percentage of volume, where Across ranks second lowest. Across's users pay relatively lower fees for each dollar bridged, and this trend is continuing, approaching the fee levels of Stargate (fee compression).

6. ABSOLUTE VALUATION BASED ON TAM GROWTH AND SHARE GAINS

This approach allows us to evaluate the prospects of ACX within the bridge landscape and assess its potential as a winner in terms of growth and share gains.

Based on @TheiaResearch and @samuelmandrew GMX valuation framework.

This report presents an analysis and estimate of Coinbase's financial performance during Q1 2023. While most publicly traded companies report their financials quarterly, Coinbase's operations in blockchain-based activities provide a unique opportunity to estimate its revenue in real-time using an automated financial model. By combining on-chain transparent data with some reasonable assumptions, we can predict the company's quarterly results with a high degree of accuracy.

This report was inspired by @kunalgoel's “Spoiler Alert - Coinbase Q4 Revenue Estimates” Messari post. I adapted the SS while adding some additional assumptions to ensure that it updates automatically on a daily basis. Having said that, this report provides a walkthrough of the Coinbase Dashboard 🔵 and highlights the methodology and assumptions used to estimate its different revenue sources to arrive at the top-line revenue.

The historical financial information used in the model was mainly obtained from Coinbase's official website, and annual and quarterly reports, which also provided detailed insights into the company's sources of revenue. For blockchain-related information, I used the Artemis_xyz Google Sheet plugin (links at the end).

The model estimates Coinbase's Q1'23 top-line revenue to be $606 million, with the Fee Take Rate on Trading Volumes remaining the primary driver of overall revenue. A sensitivity table at the end of this report is included, which provides a range of estimates from $602 million to $710 million, reflecting the potential impact of different Fee Take Rates.

TRANSACTION REVENUE

Over the past year, Coinbase's transaction revenues have declined proportionally to the increase in revenues from its subscriptions and services segment. The volatility of the cryptocurrency market and Coinbase's effort to become an all-weather company and diversify its revenue streams are the main reasons behind this decline. As you can observe in the graph below, Coinbase is currently navigating through a transition period as it shifts towards a more stable P&L, providing opportunities to generate strong top-line revenue and EBITDA, even in an ever-changing market.

1. Trading volume and transaction fees

Transaction Revenue remains Coinbase's primary revenue source. To estimate it two drivers have been considered: a) Trading Volume, and b) Fee Take Rate. The methodology used considers the distinction between the volume transacted by retail and institutional investors, given that the company has historically charged different fees to each of these segments.

a) In terms of Trading Volume, the company regularly publishes the total transactional volume on a daily basis. However, it does not distinguish the proportion that comes from retail investors from the proportion that comes from institutional investors. The first assumption we have to make in this case is what % of the total transaction volume comes from retail and what percentage comes from institutional investors since it charges different fees to each of them.

As seen in the chart below, since 2018, the proportion of volume transacted by retail investors has been steadily declining, suggesting this increasing entry of institutional investors into the crypto trading market.

Based on this historical data, I expect the negative trend in retail transacted volume to persist during Q1´23E. Therefore, I predict that the percentage of retail transacted volume will slightly decrease from the previous quarter's figure of 14% to 13%. The remaining 87% of the transacted volume is anticipated to come from institutional investors. Therefore, out of the reported total trading volume of $145B, I estimate that $18B will come from retail investors and $126B will come from institutional investors.

b) In terms of the Fee Take Rate, I assumed Coinbase will continue to charge a transaction fee similar to the historical average for retail, while the fee for institutional investors will either decline or remain similar to the historical average. Thus:

• Retail transaction fees for Q1´23: 1.38% (avg of the last 4Q).

• Institutional transaction fees for Q1´23: 0.017% (avg of the last 4Q).

In summary, Transaction Revenue is estimated to be $290M, with a QoQ decrease of 10%.

Disclaimer: It´s important to consider that during Coinbase's Q4'22 Analyst Call on February 21st, 2023, the company disclosed a Transaction Revenue of $120M for the month of January. As such, the total revenue estimate of $290M takes into account both, the disclosed January revenue and estimated revenues for the months of February and March using the methodology outlined above.

SUBSCRIPTION AND SERVICES REVENUE

2. Blockchain rewards

To estimate 2. Blockchain Rewards, I have calculated the rewards earned separately for ETH staking and other cryptoassets (e.g., Tezos, Cardano, Algorand, Cosmos, Solana, and Aptos). Since Coinbase does not provide detailed information on the revenues of each asset, first, I estimated ETH's reward contribution by considering three key metrics:

• The amount of ETH deposited and staked through the Coinbase platform. Source: Cumulative Eth Staked (forked from @hildobby dune query).

• The daily APR offered by the Eth POS blockchain to validators. Source: Protocol vs Lido APR from @LidoAnalytical dune query.

• The daily USD price of ETH. Source: Artemis Google Sheet plugin.

Combining these metrics enables a precise estimate of ETH's contribution to Coinbase's Blockchain rewards revenue.

To estimate the contribution of other blockchain rewards, including Tezos, Cardano, Algorand, Cosmos, Solana, and Aptos, I assumed that their revenue would grow in line with their average market cap during the analyzed quarter. I obtained this data from sources such as Coingecko, utilizing the Artemis Google Sheet plugin for automatic retrieval. During Q4´22, the sum of the average market capitalization of these cryptoassets was $23.21M, increasing to $25.37M at the end of Q1´23, representing a growth of 9.3%. I used this percentage as a simple proxy to estimate the % increase in staking revenue coming from these other cryptoassets. Find below the table used to make the estimate:

Disclaimer: It is important to consider the accounting of Blockchain rewards that Coinbase uses. Staking is reported as 100% revenue but 75% of this revenue is passed through to users and accounted for in the transaction expense item. Coinbase's real or net revenue from staking is derived from the margin between the rewards it receives and what it pays its users for providing funds, but it reports it on its top-line revenue on a 100% basis.

In summary, Blockchain rewards revenue is estimated to be $74.48M, with a QoQ increase of 19%.

3. Custodial fees

To estimate 3. Custodial fees, two drivers were considered: a) Assets on Platform classified as Institutional vs Retail, and b) Custodial fee as a % of avg institutional assets. Coinbase reports its Assets on Platform quarterly, classified according to the type of client (retail or institutional) and type of assets under custody (BTC, ETH, Fiat, or other cryptoassets). My overall methodology involved estimating first Assets on Platform by the cryptoasset classification (BTC, ETH, Fiat, and others) to arrive at the total estimate. Then, I applied the estimated percentage that historically belongs to institutional investors (based on previous quarterly reported percentages), as institutional investors are charged a fee.

a) To estimate Assets on Platform, based on the cryptoasset classification, the main assumption involved a 10% estimated increase compared to Q4´22 due to the loss of confidence in traditional banks in the last weeks of March, according to SVB and First Republic news. This 10% increase was estimated in terms of the increase in units of assets under custody, regardless of price. Next, each asset was multiplied separately by its trading price at the end of the quarter to arrive at the total dollar estimate of $133B assets under custody following the classification into BTC, ETH, Fiat, and others, as shown in the table below:

The last step would involve dividing the arrived estimate of $133B total custodial assets into the % estimated to belong to institutional investors. In this case, I used the same ratio as Q4´22, which is 50%.

b) In terms of Custodial fees, once the assets under custody from institutional investors have been estimated, I assumed Coinbase will continue to charge a % fee similar to the historical average, which has been proportional to the institutional assets under custody. As observed in the table below, on average, they are charged 0.025%, which is the percentage used for this quarter.

In summary, Custodial fee revenue is estimated to be $13.3M, with a QoQ increase of 16%.

4. Interest on USDC reserves

To understand the USDC interest revenue, we have to introduce the business model of Circle, the company behind USDC. They invest a proportion of the dollar reserves backing the stablecoin through an SEC-registered public money market fund called the Circle Reserve Fund, which holds a portfolio of U.S. Treasury bonds with an average maturity period of one month, custodied at Bank of New York Mellon and managed by BlackRock.

In simplified terms, Circle, Coinbase's partner, is responsible for issuing and redeeming USDC, while Coinbase is one of the few remaining companies capable of holding USDC at scale. This partnership between Circle and Coinbase took place in October 2018 by forming CENTRE, a consortium with the mission to securely and transparently digitize on-chain dollars. Customers can tokenize their dollars and convert them into USDC and redeem USDC for dollars through both platforms, Circle and Coinbase, making USDC the industry's first open, fungible, interoperable fiat-backed stablecoin.

Circle & Coinbase Join Forces to Found the CENTRE Consortium

Coinbase and Circle have joined forces to co-found the CENTRE Consortium. Learn more about what this means for the future of digital money.

https://www.circle.com

Based on this, to estimate 4. Interest on USDC reserves, two drivers were considered: a) the USDC market cap, and b) 1 Month T-bill annualized yield. The methodology involves estimating the daily interest earned on the average 80% investment on US T-bills which Circle typically holds against its USDC circulating supply.

Below is a summary table for the month of January 2023 showing the daily interest rates that have been estimated using the methodology described above.

Finally, for the estimation of total quarterly revenues, it is assumed that they increase with respect to the previous quarter by a % equal to the estimated increase following the methodology described in the previous calculation, as shown in the following table:

In summary, Interest on USDC reserves revenue is estimated to be $205.3M, with a QoQ increase of 13%.

5. Other subscriptions and services

Most of the services that are included inside this category are not material enough for Coinbase to call out significant operational metrics, or revenue, which is why it’s bucketed in the other revenue category. They include:

• Coinbase One: subscription that gives access to $0 trading fees on the Coinbase Platform, $1M account protection, and 24/7 phone support.

• Learning Rewards / Earn Campaign: incentivizes users to learn about crypto with quizzes and tasks to earn cryptocurrency.

• Coinbase Cloud: offers tools for web3 developers to create applications using Coinbase's infrastructure. WAAS, APIs, and SDKs are some examples.

• Other subscription licenses: additional subscription-based services like advanced trading tools, market analysis, research, and tailored features to enhance user experience on the platform.

Based on this, to estimate 5. Other subscriptions and services, due to not having much information, I have followed an approach based on the average of the quarterly variations recorded during the last year. According to historically reported data, a negative trend is observed with an average decrease of -11.59%, which has been used to estimate this line of revenues for the next quarter, as can be seen in the corresponding graph.

In summary, Other subscription and services revenue is estimated to be $23.6M, with a QoQ decrease of 12%.

*BASE L2 🔵

On February 23, 2023, Coinbase announced on its blog the launch of "BASE," a L2 blockchain intended to serve as the bridge that will allow users of its platform to move to and interact with the “cryptoeconomy”.

As the development of the Layer 2 (L2) solution is not expected to have a significant material impact on Coinbase's operational metrics or revenue in the near term, I have included it as a subsection of the Other subscription and services bucket. Despite this, I believe that BASE will have a positive influence on Coinbase's revenue sources as it is expected to be a transversal solution.

Although operational metrics and revenue information cannot be disclosed at present, we can obtain a gauge of the L2 solution's economics by analyzing the revenues and daily active users of other L2 sequencers.

Optimism and Arbitrum, earned gross revenues of $18M and $22M, respectively, in 2022, generating net revenues of $4M and $6M. It is important to note that the difference between gross and net revenue is due to the process these L2s follow and how they interact with their main L1.

L2-L1 interaction*: Users pay a gas fee every time they interact with a dApp or smart contract or conduct a transaction, which is considered the gross revenue of the sequencers who validate these transactions. The validators then collect all these transactions from time to time and record them in the main L1, during which they also pay gas fees. The difference between what the sequencers pay to record these transactions in the main L1 and what they receive from the users in their L2 is what is considered the net revenue.*

This provides a useful framework for estimating the potential revenue and economic impact of Coinbase's L2 solution. Moreover, leveraging the interoperability of its new blockchain platform with its existing products and broad user base, Coinbase's Layer 2 solution presents an exponential potential increase in revenue. For instance:

• By promoting the use of USDC as BASE's native currency for transaction fees and dApps, Coinbase can increase demand for USDC, generating higher interest income and consolidating its position in the stablecoin market.

• Incentivizing the use of cbETH can increase ETH deposits through Coinbase's staking platform, resulting in higher revenue through the 25% fee charged for offering this service

CONCLUSION

In conclusion, Coinbase's Q1 2023 financial performance is expected to be strong, with an estimated base case revenue of $606M. The Fee Take Rate on Trading Volumes is expected to remain the primary driver of overall revenue. Additionally, Coinbase's diversification of revenue streams through its subscriptions and services segment, such as Blockchain rewards, Custodial fees, Interest on USDC reserves, and Other subscriptions and services, will contribute to its overall revenue growth.

Moreover, the launch of BASE, Coinbase's Layer 2 blockchain solution, presents a promising opportunity for exponential revenue growth. By leveraging its new blockchain platform and existing user base, Coinbase can generate economic value by promoting the use of USDC as BASE's native currency, incentivizing the use of cbETH, and providing a transversal solution to move and interact with the cryptoeconomy. As the cryptocurrency market continues to evolve, Coinbase's commitment to diversifying its revenue streams and providing innovative solutions for its users ensures its position as a major player in the industry.

Finally, I would like to express my gratitude to everyone who has taken the time to read and appreciate this report. Thank you!

Appendix:

1. Currently, Ethereum's staking ratio, which is the proportion of staked ETH versus its circulating supply, is around 15.6%. In contrast, other proof-of-stake (PoS) blockchains such as Cardano and Solana have staking ratios of around 70%. With the upcoming "Shanghai" upgrade and withdrawal enablement, the staking ratio for Ethereum is expected to increase (to 40-50%). If Coinbase is able to maintain its LSD market share and capitalize on this growth, it would result in a significant increase in gross revenues for this segment. This assumes that Coinbase's market share will remain constant as the overall market expands, leading to proportional growth.

   If we were to estimate this into the future we could expect a considerable increase in Blockchain rewards revenues. If Ethereum's staking ratio were to increase to 50% (accounting for a reduction in APRs), Coinbase´s blockchain rewards revenue coming from ETH could increase by 25-40%.

2. Data sources used:

   a) Artemis Google Sheets and Excel Plugin

   b) Dune Dashboards: Eth Staked - Cumulative by @hildobby; Lido staking APR by @LidoAnalytical; Rollup economics by @niftytable / @0xKofi on tw.

   c) Ycharts for 1-Month T-bill

   d) Fourth Quarter and Full Year 2022 Analyst Call (Q4´22 Shareholder Letter).

3. Sensitivity table for Institutional and Retail Fee Take Rates.

A month ago, I started creating Google Sheets dashboards to track various narratives I was bullish on. Some of them started to gain traction on Twitter, so I thought it would be interesting to do a small report on every dashboard I have created, so here we are.

The cryptoasset market has grown exponentially in recent years, making it essential for investors to stay up-to-date with the latest developments in the space. To help investors make informed decisions, I have created a Bridge Token Dashboard that offers a complete view of protocols such as $ACX, $HOP, $SYN, $STG, $CELR, and $MULTI.

This report provides a walkthrough of the Bridge Token Dashboard and highlights some of the most important KPIs and fundamental metrics any investor should look at. I will also explain how the metrics were chosen and what insights can be gained from them. Whether you're a seasoned investor or just getting started in the cryptocurrency space, this dashboard will provide valuable insights that can help you make better decisions.

The report concludes that Across and Hop are among the highest potential protocols due to their capital-efficient design. I argue that, in addition to volume, capital efficiency is a more accurate measure of bridge usage and asset utilization, rather than TVL.

1. THE DASHBOARD

As you can see, the dashboard is divided into two main sections:

1. Market Data, Seed & Multiples: This section shows current prices of tokens, market cap, FDV, investor seed allocation, and vesting schedule. It also includes the multiples at which these tokens are currently trading. This is a crucial part of the dashboard as it aggregates information from the two categories and may highlight undervalued opportunities.

2. Fundamental Data: the second section focuses on KPIs that are specific to bridges, such as TVL, Bridge Volume, Capital Efficiency ratio, DAU, Revenue, and Token Incentives.

Let´s now explain each section separately, and look at some interesting data.

Market Data, Seed & Multiples

a) Price, Market Cap, FDV

The first key factor to consider here is the FDV of each project and the ratio of Market Cap / FDV. Generally, a lower FDV indicates less potential for a significant price drop. However, this information should only be the starting point for our analysis and should not be the only data we rely on. Although the crypto market is quite inefficient, there may be a reason why some protocols trade at a higher FDV than others (which we will examine in the Fundamental Data section).

Regarding Market Cap / FDV, except for $CELR and $SYN, every other token currently has a ratio below 20%. This suggests that supply inflation is likely to occur in the future (historical inflation data is provided at the end of the dashboard and in the next section).

b) Seed Investors

Now, let's take a closer look at the Seed Investors and briefly outline how their allocations were determined, as well as when future unlocks might occur that could affect token prices. While each Seed Investor received private funding, their terms, and characteristics varied, including:

• Across: On November 23, 2022, (before TGE) they announced a Strategic Partnership with Placeholder, Blockchain Capital, and Hack VC, which consisted of a $10M investment in exchange for 50M $ACX + 50M Success Tokens, vested until June 30, 2025. They invested at a FDV of $200M, implying a $0.2 seed price.

• Hop: On TGE, Hop announced its tokenomics, where they disclosed a 6.25% allocation to SeedInvestors, with a 1y cliff and a 2y linear vesting schedule, ending on June 10, 2025. The seed price was not disclosed.

• Celer: they disclosed on a blog post 3 different token sale rounds: 1) Seed sale on April 2018, for 11.5% of the total supply, at $0.0065; 2) Private sale on August 2018, for 15.5% of the total supply, at $0.015; 3) Binance Launchpad on March 2019, for 6% of the total supply. They are all fully unlocked and in circulation, so it´s not that relevant.

• Synapse: $SYN was originally distributed in Q3 2021 to $NRV tokenholders. The team states that there is no schedule nor locks left for private investors (if there were any), so it´s not that relevant, and in any case, it´s fully unlocked by now.

• Stargate: their docs state an investor allocation of 17.5%, with a 1y cliff and 2y linear vesting schedule, starting on March 18, 2022 (just 13 days at the time of writing this report).

• Multichain: it was rebranded from Anyswap, and they allocated 20% of their total supply to the shareholders of the Anyswap/Multichain company. I couldn´t find information on their token release schedule or seed price.

c) Comps Valuation Multiples

The purpose here is to have a quantifiable valuation screen to quickly view bridge tokens that may be undervalued in comparison to their peers. There is a multiple for every fundamental KPI.

Fundamental Data

To begin this section, it´s important to state that crypto is one of the least efficient markets that exists right now. This is attractive but should be noted that what drives price appreciation doesn´t always have to be fundamentals. This dashboard focuses on a fundamental valuation thesis, but one should never base their decision of investing solely on this. It helps for high-conviction plays, but in the end, price appreciation occurs when there are more people buying a token than selling it, and there are many variables (apart from fundamentals) that can lead investors or degens to buy a token.

As an example, other drivers include:

• Supply shocks in a token (a protocol suddenly decides to burn 90% of its supply).

• Narrative-driven thesis (dependent on catalysts, like the LSDs narrative and them becoming predictable CF assets with the ETH Shanghai upgrade).

• Community-driven meme tokens (70-90% of the token allocation to retail via an airdrop).

• VC Treasury Management Deals (LIDO did various of these and the announcements were followed by big price rallies because they normally involve large vesting schedules, and help to align incentives).

a) TVL, Bridge Volume, Capital Efficiency Ratio

It’s commonly assumed that Bridge Volume and TVL are the top metrics when looking at bridges, where higher TVL is thought to be indicative of growth. However, I believe that in addition to volume, a more accurate measure of bridge usage and asset utilization is capital efficiency, rather than TVL. Fundamentally, the better bridge is the one that requires relatively low capital to do a lot of bridge volume. This brings a lot of advantages:

• Lower user fees: Since capital has a cost, protocols that require more liquidity to process the same volume must pay higher fees to liquidity providers (LPs). These costs are ultimately borne by the user. In contrast, a bridge that requires less capital can offer lower fees to users.

• Increased LP yields: A bridge that achieves a high volume of transactions with a lower TVL can generate substantial yields for LPs. For example, a bridge with 50% of its TVL utilized daily can offer an APY of approximately 20% (at 0.1% LP fees), while one with 15% utilization has an APY of less than 6%. You can charge lower fees to the bridge user but pay higher fees to the LPs because due to requiring less capital, the total amount of fees coming from bridge volume gets distributed among lesser capital. This enables the ability to retain capital organically (without liquidity token incentives).

• Less security risk: even though LPs still have security risk, the total amount of TVL that is risked for hackers to get is less, as there is less capital at risk. In terms of centralization risk, Hop and Across are both optimistically designed bridges, while in other bridges like Synapse, withdrawals (or “bridge transfers”) are validated by an EOA, which is more centralized by design (L2BEAT).

b) DAU, Revenue, Token Incentives / Inflation

2. THE GRAPHS

CONCLUSION

In summary, this report provides an overview of the Bridge Token Dashboard, which facilitates investor insights into various protocols such as $ACX, $HOP, $SYN, $STG, $CELR, and $MULTI. The dashboard is designed to update automatically on a daily basis, allowing users to monitor any changes in fundamentals that could potentially result in overvalued projects.

Based on the analysis presented, I believe that the protocol that emerges as the most capital efficient in the bridge/cross-chain messaging space will ultimately be the winner. This is because such a protocol would be the most cost-effective for users while also providing organic returns for LPs. Across and Hop are the clear winners here, with Hop showing a higher average annual Capital Efficiency ratio in 2022 and Across exhibiting a higher peak utilization ratio, indicating that it is best suited for high-stress conditions. I think this can serve as a long-term strategic advantage that can help both protocols grow market-share and attract bridge volume and TVL.

Additionally, Synapse is also interesting as they announced the launch of the Synapse Chain, which will create a cross-chain interoperable ecosystem, secured by $SYN, connecting all chains, but that´s still in the works. Furthermore, their analytics page is also noteworthy.

Finally, I would like to express my gratitude to everyone who has taken the time to read and appreciate this report. Thank you!

In December 2020, Lido launched as a liquid staking solution for Ethereum and other proof of stake chains. This allows users to stake their tokens without having to lock assets or maintain staking infrastructure. With the launch of Lido came the issuance of 1 billion LDO tokens, which were distributed among a variety of stakeholders, including Seed Investors, Team and Validators, and the DAO treasury.

This report provides a walkthrough of the LIDO Token Allocation Framework and highlights some of the key players involved in the distribution of the tokens. It also labels some of these investors’ wallets, so you can track on-chain what they have done with their allocation.

The report concludes that Lido is leading the way in achieving "moneyness" in the liquid staking space due to its strong connections within the Ethereum and DeFi community (as an example the stETH token received after depositing ETH on Lido is accepted as collateral in many of the top DeFi blue chips: Aave, Curve, Maker, Uniswap, unlocking capital efficiency).

1. LDO TGE DISTRIBUTION

On January 7, 2020 -3 weeks after issuing the tokens- LIDO published a blog post indicating how were the 1 billion LDO tokens allocated. A snapshot of the reported allocation is included below.

Introducing LDO - The Lido DAO Governance Token

The LDO token governs all Lido DAO governance and network decisions to ensure its prolonged stability and decentralized decision-making to facilitate the growth of fair, trustless and transparent liquid staking.

https://blog.lido.fi

You can reorganize this into 3 categories (which I will be using throughout the report): Seed Investors, which were to receive 22.18%, Team and Validators, 41.5% and Treasury, 22.18%. But if you look at the on-chain allocation there is a discrepancy between the reported allocation in the blog post and the on-chain distribution:

On-chain distribution of the 1 billion minted LDO:

• Seed investors: They ended up receiving 236.4M, not 221.8M (+14.6)

• Team and Validators: They ended up receiving 400.4M, not 415M (-14.6)

• Treasury: It ended up receiving 363.198M, practically the same (363.2M)

Lockup terms: they will have their LDO tokens unlocked on December 17, 2021, followed by a 1y vesting period. These restrictions apply to developers, validators, investors, founders, and employees.

I found it pretty interesting that just at the start there was this 14.6M discrepancy. Having said that, in practical terms, there is not that much of a difference because both, Seed and Team and Validators, have the same lockup terms. Anyway, I think it was worth pointing out.

2. LABELING SEED, TEAM, AND VALIDATORS´ WALLETS

First, the main address to focus on to understand the distribution process is the Aragon Token Manager. It was created from the Lido: Deployer 1 address, and it basically manages the distribution among the different TGE participating wallets.

In this second section, our focus will be to understand the token distribution among Seed Investors and Team and Validators. The reason for this distinction is that these two categories have already been fully unlocked, while the Treasury Deals occurred after the initial token generation event and remain partially locked. My aim is to shed light on the distribution process by highlighting the labeled addresses for each category and presenting a comprehensive table with updated information on their holdings.

Seed Investors

Lido's $2 million seed funding round was supported by a diverse group of investors, including well-established venture capital firms and well-known angel investors.

Based on Messari´s Funding Round information:

Lido raised $2 million in a Dec. 2020 round backed by Semantic Ventures, ParaFi Capital, Terra, KR1, Stakefish, and Staking Facilities, among others. Angel investors, including Rune Christensen of MakerDAO, Stani Kulechov of Aave, and Kain Warwick of Synthetix, also participated in the round.

Once you have information about who were the parties involved, by going on-chain, you can try to match and verify the wallets that received an allocation from the Aragon Token Manager address (here is the Etherscan´s transaction). This process is done manually and involves checking if the wallet has sent funds to a previously identified and labeled address.

Although many of the wallets were difficult to identify, with the help of Arkham, I was able to successfully label some of them. Here is the list of the Seed Investors’ Ethereum addresses, with those that have been labeled highlighted for easier reference:

Team and Validators

Going back to the already mentioned January 7, 2020 blog post:

Founding members are also distributed LDO, with the respective amounts: 221,800,000 LDO for investors, 65,000,000 LDO for validators* and signature holders, 200,000,000 LDO for initial Lido developers, and 150,000,000 LDO for founders and future employees*.

If you sum up on-chain Team (and Developers) 1-8 allocations (label from Token Unlocks) you arrive at 340M, not 350M (-10M); and if you sum up Validators´ allocation you arrive at 60.4M, not 65M (-4.6M).

This goes back to the initial TGE distribution where we discovered that 14.6M discrepancy in favor of Seed Investors, which now can be further broken down into:

a) -10M for the reported allocation to the Team

b) -4.6M to Validators.

Here is the Etherscan´s transaction. In this case, information on the identities of the Team and Developers members and most of the Validators were not publicly available. However, some of them, such as Alex Svanevik from Nansen and Anton Bukov, can be identified thanks to labeled wallets in Arkham.

Seed and Validators´ identified addresses table

Lastly, now that we know the identity of some of these wallets, let's examine the behavior of the TGE participants after their LDO tokens vested on December 17, 2022.

Disclosure: there is nothing wrong with selling vested tokens, as long as they have fulfilled the lock-up terms. Understanding the actions of these individuals in the LDO market can provide valuable insights into the demand for and potential future value of the token.

I have chosen to only include publicly known participants and their labeled wallets in the table below for a specific reason. By doing so, it allows us to gain a better understanding of the behavior of the most recognizable players in the LDO market. This can give us a more accurate picture of how financially incentivized they remain in making Lido the leading LSD solution:

The table reveals that ParaFi Capital, Stakefish, Alex Svanevik from Nansen, and Stani Kulechov from Aave have fully sold their unlocked LDO. This could suggest that these individuals are no longer incentivized to make Lido the leading LSD protocol. It's worth noting that a few weeks ago, RocketPool (another LSD protocol) staked eth, rETH, was proposed to be added to Aave V3.

Here is another research report written by @YohanGGlobal that focuses on wallet-tracking every big dumper of LDO, aside from being identified or overall participants in these allocations.

3. TREASURY DIVERSIFICATION DEALS

The report now shifts its focus to the Treasury Deals, which happened after the initial TGE, where Lido pretty much included every important protocol inside the DeFi space.

The first deal (Treasury Diversification #1), was proposed in the Lido governance forum on April 2021, and ended up being approved on April 6. The objective of the proposal was to sell 100M LDO from the DAO´s Treasury allocation of 363.2M in exchange for 21600 ETH of funding, at $0.7 per LDO, with 1-year cliff and 1-year linear unlocking since the moment of purchase. The purchasers included Paradigm, other institutional investors like 3ac, Alameda, DragonFly, and Coinbase; and other angel investors like Calvin Liu, 0xMaki, Mariano Conti…

The second deal (Treasury Diversification #2), was proposed in the Lido governance forum on July 2022 and ended up being approved on August 4. The objective of the proposal was to sell 10M LDO from the DAO´s Treasury allocation in exchange for 24.27M DAI of funding, at $2.427 per LDO, with 1-year cliff unlock since the moment of purchase. The purchaser was DragonFly Ventures (the same entity that participated in Treasury Diversification #1).

Summary of both deals below:

As you can see, the motivation for both deals was to assemble a diverse group of stakeholders (36 more) with aligned incentives for the success of Lido, by including as many important players as possible in their “cap table”, through direct (angels) or indirect (institutional) investments.

Treasury Deal #1

Something interesting: since approved on the 6th of April 2021 there was a 1-month time frame given to these investors to purchase LDO with ETH through a smart contract. There were 3 Angel Investors who missed the deadline. They were given a second opportunity to purchase (voted again on Snapshot), with essentially the same terms: 1-year cliff and 1-year linear unlocking schedule, from the moment of the purchase.

This is the reason why they started vesting in the months of October-November, instead of the original month of May.

So on this second opportunity and 1-month timeframe deadline:

• Hart Lambur -0xcF4... purchased on the 9th of October 2021. Cliff: October 9, 2022, linear unlock until October 9, 2023.

• Danny Ryan, Core Researcher of Ethereum  -0xf948… purchased on the 15th of October 2021. Cliff: October 15, 2022, linear unlock until October 15, 2023.

• Ameen Soleimani from Spankchain - 0x3B2… purchased on the 1st of November 2021. Cliff: November 1, 2022, linear unlock until November 1, 2023.

Treasury Deal #2

Something interesting: the lockup terms are different between Treasury Deal #1 (the Paradigm + 35 investors round), and Treasury Deal #2 (DragonFly round). In this case there is only a 1y cliff unlock, which ended up being agreed upon by the DAO for two main reasons:

1. DragonFly has publicly (and more importantly) on-chain stated their belief in the long-term potential of the protocol. Despite being 75% unlocked, they haven´t sold any of their April 2021 allocation of 400 ETH worth of LDO, which they acquired at $0.7 per token. The current value of that LDO is 1150 ETH.

2. They decided to invest close to the local top. Which meant they ended up buying LDO at $2.427.

PRICE: 7-day backward-looking TWAP taken at the time of vote completion (August 4, 2022) + 5% premium

The final 7-day TWAP was $2.31 +5% premium: $2.427. DragonFly committed to the terms, and ended up purchasing through this smart contract:

Treasury Deals participants´ identified addresses table

In line with my approach for the Seed and Validator's table, I have exclusively selected to showcase publicly recognized participants and their identified wallets in order to try to gain an actionable understanding into their motives for selling or retaining their vested allocations.

The table reveals that:

• Paradigm has not sold any of their 53M unvested allocation (at least on-chain, we can´t really know if they traded OTC).

• DCG appears to not have sold any of its holdings, but this may not be completely true. They have sent their LDO across different wallets and maybe some of these transactions (despite not going to a CEX or DEX) were OTC trades with other entities.

• Alameda Research and Three Arrows Capital are constantly selling their daily vested Lido in batches (their wallets are on the Google Sheet spreadsheet).

• Institutional Investors overall seem to be selling more of their allocations. The main reason is that many of them end up failing and going bankrupt during the last cycle, and liquidators are in search of assets to sell.

CONCLUSION

In my opinion, the ultimate winner in the LSD space will be determined by the platform that has the most connections to other protocols, rather than just offering high yields for users. This quality is known as “moneyness” or even “capital efficiency”, the ability to use your staked ETH as collateral across different DeFi protocols.

Currently, Lido appears to be in the lead in this aspect due to its strong ties within the Ethereum and DeFi communities. As we have explained throughout the report this is because its funding rounds have involved investment from various important protocols within the DeFi space. And, even though some of these investors have sold their allocations, many other are still holding. This alignment of interests ensures that Lido is well-positioned to succeed, with the ultimate goal of establishing stETH as a form of money, a widely-used derivative for collateral across DeFi. It´s a winner takes all market.

In response to interest from the community, I have created a fundamental dashboard for LSD protocols. It looks at different fundamental and market data metrics of some of these protocols like Lido, RocketPool, Stakewise, Ankr, Frax Finance, and Staked.finance. Although it's not fully completed yet, here is the Google Sheet link. Thanks for reading!