Prime Brokerage & Unified Credit — From Wall Street to Crypto
TL;DR
- A prime broker is the "one-stop shop" for institutional investors — hedge funds, asset managers, and family offices get margin lending, securities lending, custody, trade execution, clearing, capital introduction, and reporting from a single counterparty. In TradFi, this means Goldman Sachs, Morgan Stanley, and JPMorgan — each holding ~$1 trillion in client balances.
- Prime brokerage is a $37 billion/year revenue machine — global PB financing revenues are projected to hit $37B in 2025, up 18% YoY, driven by $2.5 trillion in borrowing balances (doubled since 2020) and a $4.5 trillion hedge fund industry.
- TradFi PB concentration creates systemic risk — the top 3 prime brokers control ~60% of the market. When Archegos Capital collapsed in March 2021, it triggered $10B+ in bank losses across Credit Suisse ($5.5B), Nomura ($2.85B), Morgan Stanley ($911M), and UBS ($774M). One family office, six banks, zero transparency.
- Crypto has no real prime broker yet — the market is fragmented across exchange-specific margin accounts, with no cross-venue netting, no unified collateral pool, and no central credit intermediary. You need separate accounts, separate collateral, and separate risk on every exchange.
- Crypto PB players are emerging fast — FalconX ($8B valuation, $1.5T+ volume), Ripple Prime (acquired Hidden Road for $1.25B), Coinbase Prime, Anchorage Digital, and Galaxy Digital are building the institutional plumbing. But none yet match TradFi's integrated model.
- Unified credit is the holy grail — a single margin pool across multiple venues, where your long on Binance offsets your short on Deribit. Copper ClearLoop and Fireblocks Off-Exchange are building this with off-exchange settlement: assets never leave custody, only trading credits move.
- The custody problem is THE problem — TradFi PBs hold your assets; in crypto, leaving assets on an exchange means counterparty risk (see: FTX). Off-exchange settlement solves this by keeping collateral in MPC wallets while mirroring balances on-exchange.
- FTX was the cautionary tale — an exchange that was simultaneously the broker, the clearinghouse, the custodian, AND the hedge fund. $10B in customer funds lent to Alameda. Sam Bankman-Fried got 25 years. The lesson: vertical integration without segregation is fraud waiting to happen.
- Backpack's model — a regulated exchange (MiFID II licensed in the EU) that functions as its own clearinghouse, combining the speed of vertically integrated crypto with regulatory separation. The exchange-as-clearinghouse approach, done right.
1. What Is a Prime Broker?
A prime broker is a financial institution that provides a bundle of services to institutional investors — primarily hedge funds, but also asset managers, pension funds, family offices, and sovereign wealth funds. Think of it as the back-office infrastructure that lets a fund focus on trading rather than operations.
The term "prime" comes from the concept of a primary relationship: one broker that handles (nearly) everything.
The Prime Brokerage Relationship
┌─────────────────────────────────────────────────────────┐
│ HEDGE FUND │
│ (wants to trade, borrow, short, lever up) │
└──────────────────────┬──────────────────────────────────┘
│
Single relationship
│
▼
┌─────────────────────────────────────────────────────────┐
│ PRIME BROKER │
│ │
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌───────────┐ │
│ │ Margin │ │Securities│ │ Trade │ │ Custody │ │
│ │ Lending │ │ Lending │ │Execution │ │ │ │
│ └──────────┘ └──────────┘ └──────────┘ └───────────┘ │
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌───────────┐ │
│ │ Clearing │ │ Capital │ │Reporting │ │ Risk │ │
│ │& Settle │ │ Intro │ │& Tech │ │Management │ │
│ └──────────┘ └──────────┘ └──────────┘ └───────────┘ │
└──────────────────────┬──────────────────────────────────┘
│
Routes to multiple venues
│
┌────────────┼────────────┐
▼ ▼ ▼
┌────────┐ ┌────────┐ ┌────────┐
│ NYSE │ │ NASDAQ │ │ CBOE │
└────────┘ └────────┘ └────────┘Why Funds Need a Prime Broker
A hedge fund could theoretically open accounts at every exchange, set up custody relationships with multiple banks, negotiate borrowing lines individually, and build its own reporting infrastructure. But that's wildly inefficient. A prime broker consolidates all of this:
- One margin account instead of dozens
- One custody relationship instead of fragmented holdings
- One credit line instead of negotiating with each counterparty
- One set of reports covering the entire portfolio
- One legal framework governing all activities
The fund trades wherever it wants, and all positions funnel back to the prime broker for clearing, settlement, and financing.
2. Core TradFi Prime Brokerage Services
Margin Lending (The Big Revenue Driver)
The foundation of prime brokerage is lending. A hedge fund that wants to run a $5B portfolio with $1B in capital needs $4B in financing. The prime broker provides this — secured by the fund's portfolio as collateral.
How it works:
- The fund posts collateral (typically securities or cash)
- The PB extends a credit line, usually at a spread over a reference rate (Fed Funds, SOFR)
- The fund uses the borrowed money to buy more securities
- The PB earns the lending spread — often 50-200 bps depending on collateral quality and fund creditworthiness
For top-tier hedge funds, financing rates might be SOFR + 35-50 bps. For smaller or riskier funds, it could be SOFR + 150-300 bps.
Securities Lending (Enabling Short Selling)
When a fund wants to short a stock, it needs to borrow the shares first. The prime broker facilitates this:
Securities Lending Flow:
Lender (Pension Fund) Prime Broker Borrower (Hedge Fund)
│ │ │
│ Lends shares │ │
├─────────────────────────►│ Lends shares │
│ ├─────────────────────────►│
│ │ │ Sells short
│ │ ├──────► Market
│ │ │
│ Receives lending fee │ Earns spread │ Posts collateral
│◄─────────────────────────┤◄─────────────────────────┤
│ (e.g., 25 bps) │ (e.g., 25 bps) │ (102-105% of
│ │ │ share value)The PB sits in the middle, earning a spread between what it pays the lender and what it charges the borrower. Hard-to-borrow securities (low float, high short interest) can command lending fees of 10-50%+ annualized.
Trade Execution & Clearing
Prime brokers offer execution services — algorithms, direct market access, dark pool access — and clear all trades through the relevant clearinghouses (DTCC/NSCC for US equities, LCH for rates, ICE Clear for energy, CME for futures).
The fund can also use executing brokers — other dealers who handle specific trades — while the prime broker clears and settles everything. This is the "give-up" model: the executing broker "gives up" the trade to the prime broker for settlement.
Custody
The prime broker holds the fund's assets — securities, cash, and derivatives collateral. In the US, SEC Rule 15c3-3 (the Customer Protection Rule) requires broker-dealers to segregate customer assets from the firm's own assets. This is the regulatory guardrail that (theoretically) prevents a TradFi version of the FTX collapse.
Capital Introduction
PBs connect hedge funds with potential investors (allocators, fund of funds, family offices). This isn't a guaranteed fundraise — it's more of an organized matchmaking service. But for a new fund trying to get in front of institutional capital, it's invaluable.
Reporting & Technology
Daily portfolio reports, risk analytics, P&L attribution, regulatory reporting (Form PF, AIFMD), and API-based data feeds. Increasingly, PBs compete on technology infrastructure as much as on financing rates.
3. How TradFi Prime Brokerage Economics Work
Prime brokerage is one of the most profitable businesses in investment banking. Here's how the money flows:
| Revenue Source | Typical Range | % of PB Revenue |
|---|---|---|
| Margin lending (financing spreads) | SOFR + 35-300 bps | 40-50% |
| Securities lending | 25 bps - 50%+ annualized | 20-30% |
| Trade execution commissions | $0.01-0.05/share or 2-5 bps | 10-15% |
| Custody & admin fees | 2-10 bps of AUM | 5-10% |
| Technology & platform fees | Fixed + variable | 5-10% |
The Scale Economics
Prime brokerage has massive operating leverage. The marginal cost of servicing an additional $1B in client balances is minimal once the infrastructure exists. This is why the business concentrates:
Prime Brokerage Market Concentration (2024):
Top 3 (GS, MS, JPM) ████████████████████████████████ ~60%
Top 10 ██████████████████████████████████████████ ~80%
Top 25 ██████████████████████████████████████████████████ 92%
Everyone else ████ 8%Goldman Sachs alone generates ~$1.2 billion per quarter in equities financing revenue from its prime brokerage business. Across the top 3, combined PB-related revenues likely exceed $15 billion annually.
Why Funds Use Multiple Prime Brokers
Despite the "one-stop shop" pitch, most large funds use 2-4 prime brokers:
- Counterparty risk diversification — don't put all your eggs in one bank
- Best execution — different PBs have strengths in different markets
- Competitive financing — pit PBs against each other on rates
- Capacity — single PB may not have enough balance sheet for a mega-fund
Post-2008, using multiple PBs became standard practice after Lehman's bankruptcy showed what happens when your sole PB goes under.
4. The Major TradFi Prime Brokers
The Big Three
| Broker | Est. PB Balances | Key Strength | Notable Clients |
|---|---|---|---|
| Goldman Sachs | ~$1T | Technology platform, securities lending | Citadel, Millennium, Point72 |
| Morgan Stanley | ~$1T | Breadth of services, synthetic PB | Bridgewater, Two Sigma, DE Shaw |
| JPMorgan | ~$1T | Balance sheet depth, clearing | AQR, Man Group, Baupost |
The Challengers
| Broker | Position | Key Differentiator |
|---|---|---|
| Barclays | Top 5 | Strong in European equities |
| BofA Securities | Top 5 | Fixed income PB strength |
| UBS | Top 10 | Swiss wealth management crossover |
| BNP Paribas | Growing | Acquired Deutsche Bank's PB book |
| HSBC | Growing | Asia-Pacific PB expansion |
Concentration Creates Systemic Risk
When three firms control 60% of the market, a failure at any one of them sends shockwaves through the entire hedge fund industry. This was the lesson of 2008 (Lehman) and 2021 (Archegos). Each PB is interconnected with thousands of funds, and each fund is interconnected with multiple PBs. The system is highly networked — and networked systems are vulnerable to cascading failures.
5. Cross-Margining in TradFi
One of the most valuable services a prime broker offers is cross-margining — the ability to offset positions across different asset classes, reducing the total collateral required.
How Cross-Margining Works
Without cross-margining:
Position 1: Long $100M in S&P 500 stocks → Margin required: $15M
Position 2: Short $80M in S&P 500 futures → Margin required: $8M
─────────────────────
Total: $23MWith cross-margining:
Position 1: Long $100M in S&P 500 stocks ┐
├─► Net exposure: $20M long
Position 2: Short $80M in S&P 500 futures ┘
─────────────────────
Total: $4M (risk-based)The PB recognizes that the two positions are partially offsetting — the equity long is hedged by the futures short. Instead of margining each position independently, it margins the net portfolio risk. The savings can be 60-80%.
Cross-Margining Frameworks
| System | Operator | Coverage |
|---|---|---|
| SPAN | CME Group | Futures + options across CME products |
| TIMS | OCC | Listed equity options |
| Portfolio Margin | SEC Reg T / Broker | Equities + options at single broker |
| Cross-Margining Agreement | CME-OCC | Offsets between CME futures and OCC options |
| PB Internal Netting | Each PB | Across all products at that PB |
The key insight: cross-margining only works within a single margin system. If your positions are at different PBs or different clearinghouses, you don't get the offset benefit. This is exactly the problem crypto faces — but at a much larger scale.
6. The Archegos Collapse — A Prime Brokerage Failure
What Happened
In March 2021, Archegos Capital Management — a family office run by Bill Hwang — collapsed spectacularly, inflicting over $10 billion in losses on global banks. It was the largest single-entity trading loss since LTCM in 1998, and it exposed fundamental weaknesses in prime brokerage risk management.
Who Was Bill Hwang?
Bill Hwang was a former Tiger Management protégé (one of Julian Robertson's "Tiger Cubs") who had previously run Tiger Asia Management. In 2012, Tiger Asia pleaded guilty to wire fraud and settled with the SEC for $60 million over insider trading in Chinese bank stocks. Hwang was personally barred from the investment advisory industry in Hong Kong.
Undeterred, Hwang converted Tiger Asia into a family office — Archegos Capital Management — which managed only his personal wealth (estimated at $10-20 billion at its peak). As a family office, Archegos was not required to register with the SEC, file 13F disclosure reports, or submit to the reporting requirements that apply to hedge funds managing external capital.
This regulatory blind spot was critical. Nobody outside the banks knew how large Archegos's positions had become.
The Setup
- Archegos was a family office, not a hedge fund, which meant it was exempt from SEC reporting requirements that apply to investment advisers managing external capital
- Bill Hwang used total return swaps (TRS) — synthetic instruments where the bank holds the shares and Archegos gets the economic exposure. Because the banks held the shares, Archegos never appeared as a large holder in SEC filings
- Through TRS across six different prime brokers (Goldman Sachs, Morgan Stanley, Credit Suisse, Nomura, UBS, Deutsche Bank), Archegos built concentrated positions worth over $160 billion in notional exposure
- The leverage was extreme — estimated at 5:1 to 10:1 (some reports suggest even higher)
- Concentrated in a handful of stocks: ViacomCBS (largest holding, ~$20B), Discovery, Baidu, Vipshop, Farfetch
The Collapse
Archegos Timeline — March 2021:
March 22 ViacomCBS announces $3B stock offering
↓
Share price drops ~30% in days
↓
March 25 Multiple PBs issue margin calls to Archegos
↓
March 26 Archegos FAILS to meet margin calls
↓
Goldman Sachs & Morgan Stanley begin liquidating
(They moved first — and fast)
↓
March 26-29 ~$20 billion in forced block sales hit the market
↓
ViacomCBS, Discovery, Baidu crater further
↓
Credit Suisse & Nomura are slow to liquidate
↓
They eat the biggest lossesThe Damage
| Bank | Loss | What Happened |
|---|---|---|
| Credit Suisse | $5.5B | Slow to liquidate, poor risk controls. Eventually contributed to CS being acquired by UBS in 2023 |
| Nomura | $2.85B | Similar delays in unwinding positions |
| Morgan Stanley | $911M | Moved fast to liquidate but still took significant losses |
| UBS | $774M | Moderate losses despite quick action |
| MUFG | $300M | Smaller exposure |
| Goldman Sachs | ~$300M | Moved fastest — sold $10.5B in block trades on March 26 alone |
Total: Over $10 billion in losses.
Why It Matters for Prime Brokerage
No information sharing between PBs — Each bank only saw its own exposure to Archegos. None knew the total picture. Goldman didn't know about Credit Suisse's positions, and vice versa. Combined, Archegos had built a $160B position that no single bank fully understood.
Total return swaps created opacity — Because banks held the shares (not Archegos), the positions were invisible to the market and to other PBs. The synthetic structure was designed to avoid disclosure.
Family office exemption — As a family office, Archegos was exempt from the reporting requirements that would have flagged its concentrated positions. Post-Archegos, the SEC proposed new reporting rules for large family offices.
First-mover advantage in liquidation — Goldman and Morgan Stanley's rapid liquidation meant they shifted losses to slower-moving banks. This creates a perverse incentive: in a crisis, the fastest PB to dump wins, and the slowest loses. The "prisoner's dilemma" of prime brokerage.
Risk management failures — Credit Suisse's internal review found that its risk systems had flagged Archegos as high-risk multiple times, but the warnings were overridden by relationship managers seeking fee revenue.
The Regulatory Aftermath
The Archegos collapse triggered significant regulatory action:
- SEC proposed Form PF amendments — requiring large private fund advisers to report large positions more quickly, and mandating that family offices above a certain threshold register with the SEC
- Bill Hwang was indicted in April 2022 on charges of market manipulation, fraud, and conspiracy related to a $160 billion trading scheme. In July 2025, Hwang was convicted on federal racketeering and fraud charges and sentenced to 18 years in prison
- Credit Suisse commissioned an independent review (the Paul, Weiss report) that found a "fundamental failure of management and controls" — risk limits were repeatedly breached and overridden. The Archegos losses accelerated Credit Suisse's decline, culminating in its emergency acquisition by UBS in March 2023 for $3.25 billion
- Basel Committee on Banking Supervision issued guidance on managing risks from archetype concentrated positions, and the Federal Reserve increased scrutiny of prime brokerage risk management at US banks
The core lesson: when the incentive to earn fees from a client conflicts with the obligation to manage risk, fee incentives tend to win — until they don't.
7. Crypto Prime Brokerage Today — The Fragmentation Problem
The State of Play
In traditional finance, a hedge fund opens one prime brokerage account and trades across every exchange and asset class through a single margin pool. In crypto, the experience is the polar opposite:
TradFi Fund Experience: Crypto Fund Experience:
┌─────────────────────┐ ┌──────────────────────────┐
│ HEDGE FUND │ │ CRYPTO FUND │
│ │ │ │
│ One PB account │ │ Binance account ($5M) │
│ One margin pool │ │ OKX account ($3M) │
│ One credit line │ │ Deribit account ($2M) │
│ One custody setup │ │ Bybit account ($2M) │
│ One risk view │ │ Coinbase account ($1M) │
│ │ │ Backpack account ($1M) │
└────────┬────────────┘ │ │
│ │ 6 margin pools │
Trades everywhere, │ 6 custody setups │
clears through PB │ 6 risk views │
│ │ No cross-venue netting │
┌─────┼─────┐ └──────────────────────────┘
▼ ▼ ▼
NYSE NASDAQ CBOE Capital trapped everywhere.
No offsets. No unified view.Why Crypto Has No Real Prime Broker
| Challenge | TradFi Solution | Crypto Reality |
|---|---|---|
| Custody | PB holds all assets under SEC Rule 15c3-3 | Each exchange holds assets separately; exchange collapse = total loss |
| Margin | Single margin account, cross-asset offsets | Per-exchange margin, no cross-venue recognition |
| Settlement | T+1 via DTCC/CSD, legally guaranteed | On-chain settlement (variable), exchange-specific internal ledgers |
| Credit | PB extends credit backed by portfolio | No credit intermediary; must pre-fund every venue |
| Reporting | Consolidated via PB | Fragmented; funds build their own aggregation |
| Netting | Multilateral netting via CCP | No netting across exchanges |
| Regulation | SEC, FINRA, OCC oversight | Varies by jurisdiction; many venues unregulated |
The Capital Inefficiency Tax
A crypto fund running a basis trade (long spot on Exchange A, short perps on Exchange B) needs to post collateral on both exchanges independently. There's no recognition that the two positions are offsetting:
Basis Trade Capital Requirements:
TradFi (with PB cross-margin):
┌──────────────────────────────────┐
│ Long $10M spot + Short $10M fut │
│ Net exposure ≈ $0 │
│ Margin required: ~$500K │ ◄── 5% of notional
└──────────────────────────────────┘
Crypto (fragmented):
┌──────────────────────────────────┐
│ Exchange A: Long $10M spot │
│ Collateral required: $10M │ ◄── 100% (spot = full funded)
│ │
│ Exchange B: Short $10M perps │
│ Collateral required: $1M │ ◄── 10% initial margin
│ │
│ Total capital locked: $11M │ ◄── 110% of one leg!
└──────────────────────────────────┘This capital inefficiency is the single biggest barrier to institutional adoption of crypto markets. A fund that can run a strategy with $500K in margin in TradFi needs $11M to run the same strategy in crypto. That's a 22x capital penalty.
8. The Crypto Prime Brokerage Landscape
Major Players
| Player | Type | Key Offering | Volume/Scale |
|---|---|---|---|
| FalconX | Independent PB | Trading, credit, clearing | $1.5T+ cumulative volume, $8B valuation, 600+ institutional clients |
| Ripple Prime (Hidden Road) | Integrated PB | Multi-asset prime brokerage | $3T annual clearing volume, 300+ clients. Acquired by Ripple for $1.25B (2025) |
| Coinbase Prime | Exchange PB | Custody, trading, financing | 275+ assets, up to 6x leverage, qualified custodian |
| Anchorage Digital | Bank PB | Custody, trading, staking, governance | First federally chartered US crypto bank, 500+ tokens |
| Galaxy Digital | Financial services | Trading, lending, investment management | Public company, multi-billion AUM |
| Finery Markets | ECN/Infrastructure | Non-custodial crypto ECN | 100+ institutional clients, 30+ countries |
Different Approaches
Independent PB Model (FalconX, Hidden Road/Ripple Prime):
- Acts as a principal — takes the other side of the trade, then hedges
- Provides credit extension without requiring pre-funding
- Settles T+1, reducing counterparty exposure
- Aggregates liquidity across venues
- Most similar to the TradFi PB model
Exchange-as-PB Model (Coinbase Prime, Backpack):
- Exchange provides prime services directly to institutional clients
- Custody, trading, and margin all under one roof
- Advantage: tight integration, low latency
- Risk: counterparty concentration (all eggs in one exchange)
Custody-First Model (Anchorage Digital):
- Starts from custody (bank charter, regulatory licenses)
- Adds trading, staking, and lending on top
- Strongest regulatory positioning (OCC-chartered bank)
- Lower risk profile but potentially less capital-efficient
Infrastructure Model (Finery Markets):
- Non-custodial ECN connecting institutional participants
- Doesn't hold assets or extend credit
- Provides matching and connectivity
- Lowest risk but also lowest value-add
Hidden Road → Ripple Prime: The Landmark Deal
In April 2025, Ripple agreed to acquire Hidden Road for $1.25 billion — one of the largest deals in digital asset history. The acquisition closed in October 2025, creating Ripple Prime — the first crypto-native company to own and operate a global, multi-asset prime broker.
Hidden Road was already processing $3 trillion in annual clearing volume across FX, digital assets, derivatives, swaps, and fixed income for 300+ institutional clients. By combining this with Ripple's regulatory licenses, RLUSD stablecoin, and XRP Ledger infrastructure, Ripple Prime aims to offer:
- Cross-margining across spot, swaps, and futures
- Multi-asset clearing (crypto + traditional assets)
- Integrated stablecoin settlement via RLUSD
- FINRA broker-dealer license (obtained April 2025)
9. Unified Credit & Cross-Exchange Margin — The Holy Grail
What Is Unified Credit?
Unified credit means a single collateral pool that works across multiple trading venues. Instead of pre-funding every exchange individually, a fund posts collateral once and gets trading credit everywhere. Positions at one venue offset positions at another. Margin requirements reflect the net portfolio risk, not the sum of isolated position risks.
This is what TradFi prime brokers have provided for decades. In crypto, it's the missing piece.
Unified Credit Architecture:
┌──────────────────────┐
│ CRYPTO FUND │
│ │
│ Posts $10M once │
└──────────┬───────────┘
│
┌──────────▼───────────┐
│ CREDIT PROVIDER │
│ (PB / Custodian) │
│ │
│ Mirrors collateral │
│ to trading venues │
│ │
│ Nets risk across │
│ all venues │
│ │
│ Margin = net risk │
└──┬──────┬──────┬─────┘
│ │ │
┌─────────▼──┐ ┌─▼────────┐ ┌──▼─────────┐
│ Binance │ │ Deribit │ │ Backpack │
│ │ │ │ │ │
│ Virtual │ │ Virtual │ │ Virtual │
│ balance: │ │ balance: │ │ balance: │
│ $5M │ │ $3M │ │ $2M │
└────────────┘ └───────────┘ └────────────┘
Fund's actual capital: $10M
Effective trading capacity: $10M across all venues
(vs. $10M needed PER VENUE without unified credit)Copper ClearLoop — The Pioneer
Copper's ClearLoop is the most mature off-exchange settlement network in crypto. Here's how it works:
- Assets stay in Copper custody — protected by an English Law Trust structure, bankruptcy-remote from Copper itself
- Virtual balances are mirrored to connected exchanges — the fund gets trading credit on the exchange without depositing actual assets
- Settlement is near real-time — after trades execute, only the net P&L settles between Copper and the exchange
- Cross-venue capital efficiency — funds can delegate collateral across multiple ClearLoop-connected exchanges without withdrawal/deposit cycles
Connected exchanges: OKX, Bybit, Deribit, Bitfinex, Gate.io, Bitget, Kraken MTF, PowerTrade
Key benefit: The fund never has exchange counterparty risk. If an exchange fails (à la FTX), the fund's collateral is safe in Copper custody. Only unsettled P&L is at risk.
Partnership with FalconX: In 2024, FalconX integrated ClearLoop to provide its clients with off-exchange settlement protections combined with FalconX's cross-venue liquidity aggregation — moving closer to the unified credit vision.
Fireblocks Off-Exchange — The MPC Approach
Fireblocks takes a different technical approach using Multi-Party Computation (MPC) wallets:
- Collateral Vault Accounts (CVAs) — on-chain MPC wallets that programmatically lock assets
- 1:1 trading credit — locked assets are mirrored as trading balances on connected exchanges
- Assets never leave the wallet — only settlement instructions move between the CVA, the custodian, and the exchange
- Institutional custody standards — Fireblocks Trust Company provides regulated custody
Connected exchanges: Deribit, Bybit, OKX, Bitget, HTX, and growing
The institutional crypto custody market is projected to reach $3.28 billion in 2025 and could grow to $15.75 billion by 2034 — a sign of how seriously institutions take the custody problem.
Comparing Off-Exchange Solutions
| Feature | Copper ClearLoop | Fireblocks Off-Exchange |
|---|---|---|
| Custody model | English Law Trust | MPC wallets (Collateral Vault Accounts) |
| Settlement | Near real-time netting | Programmatic lock/unlock |
| Asset protection | Bankruptcy-remote trust | On-chain MPC — assets never leave wallet |
| Exchange network | OKX, Bybit, Deribit, Bitfinex, Gate.io, Bitget, Kraken MTF | Deribit, Bybit, OKX, Bitget, HTX |
| Cross-margin | Yes (via FalconX partnership) | Developing |
| Regulatory framework | UK-regulated | US Trust Company |
10. The Custody Problem — THE Problem
TradFi: Custody Is Solved
In traditional finance, custody is a mature, regulated function:
- Banks hold securities in segregated accounts
- SEC Rule 15c3-3 requires broker-dealers to keep customer assets separate from firm assets
- SIPC insurance covers up to $500K per customer if a broker-dealer fails
- Qualified custodians (banks, trust companies, registered broker-dealers) are legally required for investment advisers under SEC custody rules
- If Goldman Sachs' PB division collapsed, client assets would be identifiable and recoverable (in theory)
Crypto: Custody Is the Unsolved Problem
In crypto, "not your keys, not your coins" isn't just a meme — it's the fundamental security model. But institutional trading requires assets to be accessible on exchanges, creating an irreconcilable tension:
The Crypto Custody Trilemma:
SECURITY
(Self-custody,
cold storage)
/\
/ \
/ \
/ ?? \
/ \
/ \
/ \
/______________\
CAPITAL TRADING
EFFICIENCY SPEED
(Cross-venue (On-exchange
margining) execution)
You can have two. Getting all three requires
off-exchange settlement (ClearLoop / Fireblocks).Self-custody → secure but can't trade quickly or use as margin Exchange custody → fast trading but counterparty risk (FTX, Mt. Gox) Off-exchange settlement → the emerging solution: assets in custody, trading credit on exchange
Pre-Trade Credit Without Holding Keys
The breakthrough concept in crypto PB is extending credit without custodying assets. In TradFi, the PB holds your assets and lends against them — simple. In crypto, the PB needs to:
- Verify the fund has collateral (provable on-chain or in a trust)
- Extend trading credit to exchanges on the fund's behalf
- Settle net P&L periodically (not move collateral for every trade)
- Manage risk in real-time (auto-liquidate if collateral becomes insufficient)
This is technically harder than TradFi custody but potentially more secure — because assets are never concentrated at a single point of failure.
11. Exchange-as-Clearinghouse vs. Independent PB
Two Competing Models
The crypto industry is converging on two structural models for institutional market structure:
Model A: Exchange-as-Clearinghouse Model B: Independent PB
(Backpack, Coinbase, Binance) (FalconX, Ripple Prime)
┌────────────────────────┐ ┌──────────────────────┐
│ EXCHANGE │ │ PRIME BROKER │
│ │ │ │
│ ┌──────────────────┐ │ │ ┌────────────────┐ │
│ │ Matching Engine │ │ │ │ Credit Engine │ │
│ ├──────────────────┤ │ │ ├────────────────┤ │
│ │ Clearing │ │ │ │ Risk Engine │ │
│ ├──────────────────┤ │ │ ├────────────────┤ │
│ │ Risk Management │ │ │ │ Smart Router │ │
│ ├──────────────────┤ │ │ ├────────────────┤ │
│ │ Custody │ │ │ │ Custody/Trust │ │
│ ├──────────────────┤ │ │ └────────────────┘ │
│ │ Margin System │ │ │ │ │
│ └──────────────────┘ │ │ Routes to: │
│ │ │ ┌─────┬─────┐ │
│ All functions under │ │ │Ex A │Ex B │ │
│ one regulated entity │ │ └─────┴─────┘ │
└────────────────────────┘ └──────────────────────┘Model A: Exchange-as-Clearinghouse
How it works: The exchange handles matching, clearing, custody, margin, and settlement as a single vertically integrated entity. This is what most crypto exchanges do today.
Advantages:
- Speed — everything happens in one system, microsecond latency
- Simplicity — one account, one interface, one legal entity
- Capital efficiency within the exchange — cross-margining across all products on the same venue
- Regulatory clarity — one license, one jurisdiction, one set of rules
Disadvantages:
- Single point of failure — if the exchange goes down, everything goes down
- No cross-venue netting — capital trapped at each exchange
- Counterparty concentration — all assets at one entity
Backpack's approach: By acquiring FTX EU and operating under MiFID II, Backpack has positioned itself as a regulated exchange-as-clearinghouse. The MiFID II framework requires client asset segregation, best execution, and regulatory reporting — importing TradFi-grade protections into the crypto exchange model. With nearly $400B in cumulative trading volume and 40+ trading pairs, Backpack demonstrates that the regulated exchange model can scale.
Model B: Independent PB
How it works: A standalone entity sits between the fund and multiple exchanges, providing credit, risk management, and settlement services. The PB routes orders to whichever venue has the best price, and nets positions across venues.
Advantages:
- Cross-venue capital efficiency — single margin pool across exchanges
- Best execution — smart routing across multiple venues
- Reduced counterparty risk — assets don't sit on exchanges (with off-exchange settlement)
- Closest to TradFi PB model — familiar to institutional investors
Disadvantages:
- Adds latency — an extra hop between fund and exchange
- PB itself becomes a risk — if the PB fails, the fund loses its credit facility
- Complex legal structure — multi-jurisdiction, multi-exchange agreements
- Harder to regulate — doesn't fit neatly into existing frameworks
The Convergence
In practice, the two models are converging. Exchanges are adding prime services (Coinbase Prime, Backpack's institutional features). PBs are integrating more tightly with exchanges (Copper ClearLoop, Fireblocks Off-Exchange). The end state likely looks like a network of regulated exchanges connected by off-exchange settlement rails, with institutional investors choosing their preferred access point — direct via exchange or intermediated via PB.
12. FTX — The Cautionary Tale
What FTX Was
FTX wasn't just an exchange. It was simultaneously:
- The exchange (order matching and execution)
- The clearinghouse (margin and risk management)
- The custodian (holding customer assets)
- The prime broker (extending credit, lending)
- And secretly: the hedge fund (via Alameda Research)
This vertical integration — with zero external oversight — created the conditions for one of the largest financial frauds in history.
What Went Wrong
FTX's Structural Conflicts:
┌─────────────────────────────────────────┐
│ SAM BANKMAN-FRIED │
│ (controlled both entities) │
└────────────┬────────────────┬────────────┘
│ │
┌────────────▼──────┐ ┌─────▼──────────────┐
│ FTX │ │ ALAMEDA RESEARCH │
│ (Exchange) │ │ (Trading Firm) │
│ │ │ │
│ Held customer │ │ Received $10B in │
│ deposits │──│ customer funds │
│ │ │ │
│ No asset │ │ Used for: │
│ segregation │ │ - Trading losses │
│ │ │ - VC investments │
│ Backdoor in │ │ - Political │
│ accounting │ │ donations │
│ system │ │ - Real estate │
└───────────────────┘ └─────────────────────┘
The exchange lent $10B in customer funds to its
affiliated hedge fund. No disclosure. No consent.
No segregation. No oversight.The Numbers
- $10 billion in customer funds transferred to Alameda Research
- $8 billion in customer funds were missing when FTX filed for bankruptcy
- $16 billion in total claims from creditors
- November 11, 2022 — FTX filed for Chapter 11 bankruptcy
- November 2023 — SBF convicted on 7 counts (wire fraud, conspiracy, money laundering)
- March 2024 — Sentenced to 25 years in prison
The Lessons for Crypto Prime Brokerage
Vertical integration without segregation is a recipe for fraud. When the exchange, custodian, clearinghouse, and prop trading desk are all the same entity with no external checks, there's nothing preventing commingling.
"Trust me" is not a risk management framework. FTX had no independent board, no external auditor (it used a metaverse-based accounting firm), no regulatory oversight in its primary jurisdiction (Bahamas), and no asset segregation. Institutions need verifiable, not trusted, custody.
Regulatory licensing matters. FTX operated in regulatory arbitrage mode — headquartered in the Bahamas precisely to avoid US rules. Compare with Backpack's approach: acquiring a MiFID II license (FTX EU, ironically), which mandates client asset protection, reporting, and regulatory oversight.
Off-exchange settlement exists because of FTX. The entire category of Copper ClearLoop and Fireblocks Off-Exchange gained urgency because FTX proved that exchange custody is exchange counterparty risk. Institutions learned the lesson: keep assets off the exchange.
The PB and the exchange must be different entities. In TradFi, the broker (Goldman), the exchange (NYSE), and the clearinghouse (DTCC) are separate. This separation exists for a reason — it creates checks and balances. When one entity is everything, the fox guards the henhouse.
13. The Future of Crypto Prime Brokerage
What Needs to Happen
Crypto PB Maturity Roadmap:
2020-2022 2023-2025 2026+
─────────────── ────────────────── ─────────────────────
PHASE 1: PHASE 2: PHASE 3:
Basic Services Off-Exchange Era Unified Credit
• OTC desks • ClearLoop/ • Single margin pool
• Custody only Fireblocks across all venues
• No credit • Credit extension • Real-time cross-
• Exchange- • Cross-venue venue netting
specific settlement • Regulatory
• Post-FTX • Ripple Prime/ framework for
trust crisis FalconX growth crypto PBs
• Institutional • Interop with
adoption TradFi clearing
accelerating • DeFi integrationKey Trends
1. Regulatory convergence — Crypto PBs are acquiring the same licenses as TradFi brokers. Hidden Road got a FINRA broker-dealer license. Anchorage has an OCC bank charter. Backpack has MiFID II. This regulatory convergence legitimizes the space and provides the legal foundation for institutional adoption.
2. Multi-asset prime services — The line between crypto and TradFi is blurring. Ripple Prime clears crypto AND FX/fixed income. Coinbase Prime integrates with traditional prime brokers for cross-asset margining. The end state is a prime broker that handles both crypto and traditional assets in a single margin account.
3. On-chain clearing — Blockchain-based clearing and settlement has the potential to replace legacy infrastructure. Instead of a centralized CCP, a smart contract can perform novation, netting, and settlement in real-time, with full transparency and no single point of failure. Projects exploring this include Canton Network (Fireblocks integration) and various DeFi protocols.
4. The $23B market opportunity — The crypto prime brokerage market, valued at $2.95B in 2024, is projected to reach $23.3B by 2033 (23.8% CAGR). This growth will be driven by institutional allocation to digital assets, the maturation of derivatives markets, and the development of cross-venue infrastructure.
The Endgame
The ultimate vision is a market structure where:
- A fund opens a single account with a prime broker
- That account provides margin across every crypto exchange and every TradFi venue
- A long BTC position on Backpack offsets a short BTC perp on Deribit
- Collateral sits in regulated, bankruptcy-remote custody (never on an exchange)
- Settlement happens in real-time on-chain
- Regulatory reporting is automated and jurisdiction-aware
- The fund's experience is indistinguishable from what it has in equities today
We're in Phase 2 of this roadmap. Copper ClearLoop, Fireblocks Off-Exchange, Ripple Prime, and FalconX are building the plumbing. Regulated exchanges like Backpack are providing the trading venues. The pieces exist — they just need to be connected.
What the TradFi-Crypto Convergence Looks Like
The most telling signal of where crypto prime brokerage is heading: TradFi firms are entering crypto, and crypto firms are acquiring TradFi infrastructure.
TradFi → Crypto:
- Goldman Sachs and Morgan Stanley are offering crypto exposure to wealth management clients
- Standard Chartered launched a crypto custody and trading desk
- Fidelity Digital Assets provides institutional crypto custody and execution
- CME Group has become the largest regulated crypto derivatives venue by open interest
Crypto → TradFi:
- Ripple acquired Hidden Road ($1.25B) — a multi-asset PB clearing FX, fixed income, AND crypto
- Coinbase holds NYDFS Trust Company charter and offers qualified custody
- Anchorage Digital has an OCC national bank charter
- Backpack acquired FTX EU and its MiFID II license
When crypto firms hold the same licenses as TradFi firms, and TradFi firms offer the same products as crypto firms, the distinction between "crypto PB" and "PB" disappears. The future isn't a separate crypto prime brokerage industry — it's a unified financial services industry where digital assets are one more asset class on the prime broker's platform.
14. Glossary
| Term | Definition |
|---|---|
| Prime Broker (PB) | Financial institution providing bundled services (lending, custody, clearing, execution) to institutional investors |
| Total Return Swap (TRS) | Synthetic derivative where one party pays the total return of an asset and the other pays a financing rate. Used by Archegos to build hidden positions |
| Cross-Margining | Recognizing offsetting positions across asset classes or venues to reduce total margin requirements |
| Unified Credit | Single collateral pool providing trading credit across multiple venues |
| Off-Exchange Settlement | Model where assets remain in custody while trading credits are mirrored on exchanges |
| Give-Up | Process where an executing broker transfers a trade to the client's prime broker for clearing |
| Novation | CCP process of inserting itself as counterparty to both sides of a trade |
| MPC Wallet | Multi-Party Computation wallet where private keys are split across multiple parties, requiring threshold signatures |
| Collateral Vault Account (CVA) | Fireblocks' on-chain MPC wallets that programmatically lock assets for off-exchange settlement |
| ClearLoop | Copper's off-exchange settlement network enabling custody-protected trading across multiple exchanges |
| SPAN | Standard Portfolio Analysis of Risk — CME's margining system that stress-tests portfolios across scenarios |
| MiFID II | EU Markets in Financial Instruments Directive — regulatory framework governing investment firms, exchanges, and trading venues |
| Qualified Custodian | Entity legally authorized to hold client assets (banks, trust companies, registered broker-dealers) |
Further Reading
- Portfolio Margin — Cross, Isolated, and Risk-Based Margining — deep dive into margining models
- Settlement & Clearing — how trades actually finish in TradFi and crypto
- Custody & Key Management — the foundations of secure asset storage
- Insurance Funds — how exchanges manage socialized losses
- Liquidation Cascades — what happens when margin systems fail