Knowledge
Introducing vCeFi
Apr 17, 2025
Foreward - CeFi is a Trust Wrapper
A Tale of Two Realities
Did you know that at large traditional financial institutions, there are entire teams dedicated to trade reconciliation? Because traditional assets are recorded independently across different databases and settle on varying schedules (T+n), internal records of trades and positions must be constantly and manually compared to external sources. When mismatches occur, discrepancies must be corrected before settlement, or the firm may face financial and legal penalties. This process unfolds over hours or days and often requires teams working primarily in Excel and CSV dumps.
Manual systems exists because humans are error-prone.
On the other side of the looking glass, a teenager in a developing country can create a memecoin with just a few dollars and a smartphone. Millions of traders can then trade it on a DEX within seconds. Solana smart contracts facilitate price discovery and settlement, indexers extract real-time data, and Dune dashboards generate and maintain live reports. There is no back office. No spreadsheets are being emailed back and forth.
Code is determinstic. There are no mistakes.
This contrast isn’t new. When I interned on Wall Street, one of our summer analyst activities included a trip to the NYSE trading floor. The etymology of common trading terms reflects how tradition lingers:
Ticker / Tape - Real-time market data for prices and volumes. Machines using Morse code printed prices on paper tape, producing a “tick” sound (Video)
Front-running - Traders exploiting trade information quickly. Brokers would physically sprint from phones to the pit to execute trades faster than others
Spread - Difference between bid and ask prices. Originally described the physical distance between best bids and asks posted on chalkboards
You’d think electronic trading systems would have quickly replaced these arcane methods, but the CME didn’t close its trading pits until 2021. The NYSE still uses an order type (D-quote) that must be placed physically through a broker at the NYSE trading pit at 18 Broad Street, New York. Sometimes, against all altruistic instincts, “good enough” is, well, good enough.
When electronic trading eventually overcame manual trading, the exchanges did not immediately recreate the entire trading industry on computers. Objectively speaking, the technology could have improved and replaced every facet of the financial industry, but it was initially driven by need. Known as the Paperwork Crisis of 1968, due to a rise in trading volume, from 1968 to 1970, NYSE was closed every Wednesday to process trades. Settlements were delayed for weeks or months, each trade had to be hand-recorded and stock certificates physically delivered.
This led to the founding of NASDAQ and the Deposit Trust Company (DTC) to digitize trade records and securities ownership. But it wasn’t until the 1990s that electronic order books started to replace phone-based voice trading (which still happens to this day).
Lessons Learned, Use Decentralized Finance
There have been countless crises in crypto, but none have been caused by the underlying base-layer blockchain technology. Instead, the vulnerabilities have come from human errors and centralization, bugs in smart contracts, compromised signing mechanisms, flawed market logic, and private key leaks.
Just like the Paperwork Crisis, each crypto crisis pushes the industry one step closer to the endgame: Decentralized Finance.
NYSE trading floor in 1969
However, unlike the digitization of securities, this transition hasn’t been top-down. Large centralized financial institutions aren’t adopting decentralized technologies at scale. Instead, the migration has been bottom-up, driven by individual users.
After every major exchange collapse, more people have moved toward self-custodial wallets and DEXs, adopting the mantra: not your keys, not your coins.
When FTX collapsed, proof-of-reserves emerged as a new expectation for CEXs, but also opening the door to CEX-like DEXs like Hyperliquid that blend performance with transparency.
So far, the great migration to DeFi has been slow and steady:
Centralized exchange (CEX) volumes still far exceed DEX volumes.
Institutions still prefer custodians like Anchorage, which may use MPC technologies, but are still relevant due to the regulatory and operational fit.
Direct-to-consumer options like Gnosis Safe haven’t yet become industry standard.
The core reason? Trust and regulation still matter for many users.
In practice, CeFi has become an onion of regulated Web2 wrappers around DeFi technologies.
Most people’s journeys begin in CeFi because regulated CEXs offer fiat on-ramps. And they often end in CeFi too, because the Web2 tech stack that powers CEX trading still provides the smoothest and most familiar UX. While CEX reserves may be backed on-chain, it’s all wrapped in layers of CeFi that are, frankly, “good enough.”
Under Collateralized Lending, The Exception
Source Galaxy Digital
In many areas of finance, CeFi has been “good enough.” But there’s one clear exception: under-collateralized lending.
The CeFi lending crisis of 2022 started with the collapse of Terra/Luna and soon took down Voyager, Celsius, and BlockFi, all of which were essentially trust wrappers for retail crypto lending. In the institutional B2B space, the collapse of Genesis followed.
These platforms facilitated under-collateralized loans from lenders to institutional borrowers. When $UST depegged and $LUNA crashed to zero, contagion spread quickly. Lenders rushed to withdraw, and rapidly falling asset prices exposed poor risk management practices across the industry.
In all cases, CeFi lending failed because it was a black box of trust wrappers who should not have been trusted.
This time, users didn’t return to CeFi. The collapse triggered a shift to DeFi that stuck. Unlike trading, lending emphasizes safety over UX. The best UX in lending is simply: don’t lose your money—and that’s a function of transparency and verifiability. CeFi just wasn’t “good enough” in this case.
Source Galaxy Digital
In contrast to CeFi loans, DeFi loans are the opposite of black boxes:
Transparent - lenders can observe how funds are being used and monitor collateralization and risk in real time.
Verifiable - no subjectivity; all asset values and actions are on-chain.
Automated - smart contracts manage enforcement based on deterministic on-chain data.
But DeFi lending also comes with trade-offs, especially for borrowers.
Too Transparent - under-collateralized loans in DeFi often require revealing trading positions and identities to some degree.
On-Chain Only - it's difficult to take an under-collateralized loan off-chain. Flash loans are possible, but must be repaid in the same block. This limits DeFi under-collateralized lending to strictly on-chain use cases.
Onion of Trust Wrappers
CeFi overcomes these barriers using trust wrappers. A lender gives funds to a regulated institution, which in turn distributes loans to trusted, KYC’d borrowers. Risk is periodically reviewed via audits, but the ultimate fallback is legal recourse.
In institutional lending (e.g., prime brokerage), the hedge fund trusts the PB to keep their strategy confidential, and in turn, the lender trusts the PB to run proper due diligence and manage risk.
In DeFi, liquidations can take seconds. In CeFi, they can take weeks or years.
3AC “risk assessment”
We're all part of this onion, whether we realize it or not.
Take your Chase bank account. Your money is held at JPMorgan Chase, and some of it is lent to the JP Morgan investment bank, which then lends to hedge funds. That hedge fund may use your money, tak on leverage of 4x (or more), for discretionary - albeit, net non-directional bets. The hedge fund may pay 5% APY to take this loan, but each wrapper in the trust stack takes a cut. You get 0.5% APY.
But loans go bad in TradFi too.
In 2022, Bill Hwang of Archegos Capital built up $160B in gross exposure with only $36B AUM, using leverage across multiple PBs. When the underlying stock prices dropped and margin calls came, Hwang failed to meet them. When the PBs compared notes, they realized he held the same positions at each of their accounts. If they sold, prices would crater.
Goldman got out first. Credit Suisse got out last.
Credit Suisse lost billions. A year later, the almost 150 year old bank was sold to UBS.
vCeFi - Verifiable CeFi
Trust, but Verify
There’s a fundamental gap between CeFi and DeFi:
In the context of smart contracts, CeFi is a black box. Once funds are onboarded via fiat or on-chain deposits, the state and flow of those funds become opaque.
Transparency in CeFi is achieved through archaic, error-prone Web2 processes: Excel spreadsheets, audits, screenshots and manually updated reports.
There’s no question that DeFi represents the next generation of digitization, one that could eventually replace many trust-based wrappers with verifiable proofs and deterministic smart contracts. But until then, verifiable CeFi (vCeFi) offers a flexible, hybrid solution.
vCeFi is a decentralized protocol designed to address the underlying limitation that DeFi smart contracts cannot interact with centralized exchange (CEX) systems.
DeFi is powered by on-chain data - balances, transfers, trades. Smart contracts can access this information natively. But because CEX data isn’t on-chain, CeFi is effectively invisible to DeFi.
vCeFi explores a new paradigm: making CeFi transactions verifiably available on-chain, so DeFi can interact with them.
The ultimate goal of vCeFi is to go beyond proof-of-reserves by introducing a richer set of verifiable metrics, such as proof-of-trading and proof-of-risk, such that DeFi dApps can build trustless protocols on top of CeFi data.
Proof of Trading - CEX Data On Chain
When putting off-chain data on-chain, the first concern is GIGO, garbage in, garbage out. If the data isn’t provably accurate, then any smart contract logic built on it is compromised from the start.
vCeFi solves this in two ways:
verifiable witness attestations
Immutable on-chain data permanence
Verifiable Witness Attestations
When a trader executes a trade on a CEX, that trade appears in two places:
Anonymized public stream - live ticker and order book updates
Private user stream - UI, real-time websockets, and REST APIs tied to the user’s account visible only to the user and those who have the user’s API keys
vCeFi uses a Minimally Necessary Set (MNS) of permissioned witnesses who monitor the trader’s websocket feed using a read-only API key. They observe the trades in real-time, batch and hash the data into Merkle trees, and post attestations to an on-chain contract (on Monad) for a given user and epoch.
Since all witnesses are observing the same live data, and all post the same Merkle root for each time window, the result is a verifiable record of the trader’s activity, anchored on-chain.
On-Chain Data Permanence
A luxury of DeFi is that once a transaction is on-chain, it's there forever; auditable, inspectable, and irreversible unless explicitly modified by another transaction.
But in CeFi, you can’t assume permanence. For example, Binance could change its historical trade API format, invalidating previously posted hashes or Merkle roots. That would break your proof.
vCeFi mitigates this risk by storing immutable encrypted copies of trade data on IPFS/Arweave at the moment they are observed. The data is encrypted, not for technical necessity, but to preserve privacy, a core feature of CeFi. This allows traders to post verifiable trade data on-chain while keeping visibility limited to themselves, the exchange, and the attestors.
Imagine giving a group of trusted friends access to read your emails. Every time you send or receive one, they encrypt the contents and post a hash to the blockchain.
One day, your cryptography professor says you didn’t submit your assignment (which you know you did). You could forward the email, but he has no reason to trust that you didn’t change the date.
Instead, you point him to the on-chain hash posted by your friends at the time you sent the email. Since that hash was posted independently, by multiple people, at a fixed time and your email still matches it, you've proven your submission.
The only ways to fake this would be:
- Create a different email that hashes to the exact same value (second preimage attack)- Manipulate the blockchain to forge the timestamp (51% attack)
Both are computationally and economically infeasible, so your professor accepts it.
Congrats, you’ve tokenized your emails. Now replace email with your Binance trades…
Proof of Risk - Aggregating CEX Data
At this point, two datasets exist on-chain representing a verifiable record of CEX transactions:
Encrypted, immutable trade data - stored immutably via IPFS/Arweave
Witness attestations - posted on-chain by permissioned observers
Together, these datasets resemble DeFi transactions, but with privacy preserved. Normal observers can't see the data’s contents, but they can verify its integrity and timestamped origin.
But data by itself isn’t useful until it’s made actionable.
To derive meaningful insights, vCeFi introduces a second layer of permissioned participants who are authorized to decrypt the encrypted trade data and aggregate it into useful metrics.
How it Works
Each participant takes the posted trade data for a given epoch and calculates metrics like:
Volume traded
Profit and loss (PnL)
Assets under management (AUM)
Gross exposure (long exposure + short exposure)
Net exposure (long exposure – short exposure)
Net Leverage
Since this computation is deterministic, multiple independent parties can calculate and reach the same result. They then post:
The aggregated metrics on-chain, publicly legible
A zero-knowledge proof (Merkle root proof) that confirms the metrics were derived from the correct, complete and unaltered data, without revealing the underlying trade data
This architecture mirrors how zk-rollups work in scaling various L2 protocols, but for off-chain trade data verification.
Key Benefits
Traders can now prove their positions, performance, and risk profiles to third parties, without disclosing individual trades.
Smart contracts can consume these metrics and act on them in lending, governance, or insurance applications.
Auditors and regulators can verify inclusion and completeness of data without full visibility.
This makes it possible to bridge the gap between off-chain activity and on-chain trust.
Your cryptography professor was impressed with your tokenized email proof, so he levels it up.
Now, whenever you submit an assignment, professors from multiple schools grade your work independently and post the result on-chain. Each grade also comes with a Merkle root proving it was computed based on all previous assignments, none were left out.
You’re on a scholarship that requires a 90% average. Suddenly, the program is shut down due to widespread fraud.
You meet the dean and present:
- Immutable, timestamped records of your assignments- Matching grades attested by multiple independent parties (consensus)- Merkle roots proving all past work was included in your final average
The dean doesn’t need to see every assignment. The timestamped grades and Merkle proofs provide cryptographic assurance that your 90% is real.
Now swap grades with CEX trading metrics and that’s vCeFi.
Tying it all Together
Let’s revisit the fundamental gap between CeFi and DeFi:
In the context of smart contracts, CeFi is a black box.
Transparency in CeFi is provided via outdated, Web2-era processes.
vCeFi changes this paradigm by bringing verifiable CeFi data on-chain, where DeFi smart contracts can interact with it directly. Think of it as doing for CeFi what Real World Assets (RWAs) are trying to do for off-chain physical assets, but with stronger cryptographic assurances.
vCeFi doesn't fork DeFi into permissioned walled gardens. Instead, it extends CeFi into the DeFi world using:
Witness attestations
Merkle root proofs
zk-rollups and on-chain aggregation
Rather than compromise decentralization to gain institutional legitimacy, vCeFi introduces trustless logic into CeFi, one layer at a time.
The long-term vision for vCeFi is to make CeFi lending great again, but this time, powered by transparency, automation, and smart contracts.
By plugging vCeFi into dApps, platforms like tread.fi can build lending vaults that are soulbound to a user’s CEX account. These vaults allow DeFi lenders to:
Lend securely into CEXs through smart contracts
Transparently monitor how their funds are being used
React to risk events or changes in exposure in real time
In effect, the vault becomes a peer-to-peer, permissionless smart prime broker (PB), automating due diligence, enforcing risk thresholds, and enabling permissionless lending workflows, even if the assets live off-chain.
vCeFi on tread.fi
Tread.fi is a trading platform offering algorithmic trading tools across CEXs and DEXs. Serving both retail and institutional users, it processes over $1B in weekly trading volume.
Soon, we’ll launch vCeFi as a native feature, enabling users to bring transparency to their CEX trading activity.
Our goal is to promote vCeFi as an industry-wide standard. Since the proof-of-* system is modular, it can be tailored to specific use cases, solving some of CeFi’s most critical transparency gaps.
Movement Labs, Foundation conducts ‘investigation’ into ‘market maker abnormalities’
With vCeFi, market makers can post their trading data privately on-chain while publicly sharing net exposure metrics (e.g., long minus short). A legitimate MM should have little to no net exposure.
Mantra's OM token crashes 90%, founder blames exchanges
Team wallets can use vCeFi to publicly prove that:
They did not sell tokens (no proof-of-trading)
Their CEX balances remain intact (proof-of-balance)
Any OTC buybacks were genuine and verifiably executed
With this level of transparency, CEX accounts can be held accountable, without compromising privacy.
Closing Thoughts
The next chapter of DeFi isn’t about abandoning CeFi, it’s about making it verifiable.
vCeFi is a bridge, a toolkit, and a standard. One that doesn’t force users or institutions to choose between usability and transparency, but allows them to benefit from both.
We're just getting started.
Please reach out to david@tread.fi if you have any questions about vCeFi.
Check out tread.fi’s trading platform at app.tread.fi
