How TradFi Money Moves Into Crypto, and Where It Stalls

I wanted a better mental model of how traditional finance money actually gets into crypto. Not the headline version, where an ETF has a big week and everyone declares that the institutions have arrived. The plumbing version. Where does the money enter, what does it pass through, where does it sit, and where does it stop.

So we mapped it. This article is that map, and it ends with an observation I did not fully appreciate before doing the work: the money is arriving in size, and almost none of it can reach on-chain markets yet, because the layer that would carry it there has not been built. Exchanges exist in abundance. The intermediary that institutions actually trade through does not.

Follow the money and the shape of the missing piece becomes hard to ignore.

Where the money enters

TradFi capital does not enter crypto by buying coins. It enters through wrappers.

A wrapper is a regulated legal structure, a fund, a trust, a listed futures contract, a money market share, that holds or references the underlying asset and issues the allocator something familiar in its place: a security or a contract that fits inside an existing brokerage account, an existing mandate, existing accounting, and an existing audit. The wrapper does the crypto-specific work, custody, key management, on-chain settlement, inside the structure, so the allocator never has to touch any of it. This is not a convenience feature. Most institutional mandates prohibit holding assets outside qualified custody at all, so for a large share of the world's capital, the wrapper is not the preferred way in. It is the only permitted way in.

As of mid-2026 there are five main doors, each a different wrapper.

The ETF door. Two terms first. An exchange-traded product, an ETP, is any security that trades on a stock exchange and tracks the value of something else. An exchange-traded fund, an ETF, is the most common kind of ETP: a fund whose shares trade all day on an exchange like a stock. A spot bitcoin ETF holds actual bitcoin with a custodian and issues shares against it, which means a pension or an advisor can buy bitcoin exposure in the same brokerage account, under the same rules, that they use for equities.

Spot ETFs are the dominant entry point, and this is no longer a matter of interpretation. In the 2026 EY-Parthenon and Coinbase survey of 351 institutional decision-makers, 66% already hold spot crypto through ETFs or ETPs, and 81% say a registered vehicle is their preferred way to hold spot exposure at all. The flows move in institutional size: US spot Bitcoin ETFs absorbed $1.7 billion across three days in mid-January 2026, including a $648 million single-day inflow into BlackRock's IBIT.

The more telling milestone is in derivatives. In late April 2026, open interest in Nasdaq-listed IBIT options reached $27.6 billion, passing Deribit's bitcoin options open interest for the first time, per CoinDesk. Deribit had an eight-year head start. IBIT options needed roughly eighteen months. Regulated money did not learn to love crypto options; crypto options moved inside a wrapper regulated money was allowed to touch, and the flow followed immediately.

The CME door. CME Group, the Chicago-based operator of the largest derivatives marketplace in the world, is where institutions have traded futures and options on interest rates, equity indexes, currencies, and commodities for decades. Its crypto contracts are regulated by the CFTC, the US derivatives regulator, and settle in cash against reference rates, so a desk can hold bitcoin or ether exposure without ever holding a coin. For institutions that want derivatives exposure directly, CME is the venue of record. Its crypto futures and options traded a record $3 trillion notional in 2025, and 2026 volumes are running 46% above that pace. In May 2026, CME moved to 24/7 trading for crypto futures and options, pending final approval. That decision is worth pausing on. The most traditional derivatives venue in the world concluded that crypto's market structure, the always-on part, was the standard to converge toward. The venue adapted to the asset, not the reverse.

The stablecoin door. The GENIUS Act, signed in July 2025, gave fiat-referenced stablecoins a federal framework: one-to-one reserves, redemption rights, no rehypothecation. Final implementing rules from the banking agencies are due July 18, 2026, days after this article was written. The institutional response has been immediate in intent if not yet in size: 86% of surveyed institutions are using or exploring stablecoins for cash management and money movement. Stablecoins are how TradFi cash learns to settle on-chain, and regulated cash rails are a precondition for everything else in this article.

The tokenized treasury door. Tokenized US government debt grew from roughly $1 billion in mid-2024 to more than $15 billion by May 2026, per CoinDesk. In May, Moody's assigned its highest money-market-fund rating to BlackRock's BUIDL and Fidelity International's FILQ, the first time on-chain funds carried that assessment. In a GDF and ISDA survey published this month, 66% of financial institutions plan to launch tokenized money market funds before the end of 2027, and 44% expect to accept them as collateral. That last number matters more than the AUM. Collateral is the quiet center of institutional trading: ISDA counts roughly $1.6 trillion in non-cleared margin posted at end-2025. When the collateral itself lives on-chain, the case for settling the trade anywhere else weakens.

The retirement door. In March 2026, the Department of Labor proposed process-based safe harbors for 401(k) fiduciaries who include alternative assets, crypto among them, in plan lineups. The rule covers plans serving more than 90 million Americans. Even a 1% allocation across the several trillion dollars of 401(k) assets would represent tens of billions in structurally long-term flow, arriving through the most conservative fiduciary channel that exists. This door is only now being unlocked, and it opens slowly by design.

Five doors, one pattern. Every successful entry point wraps crypto exposure in a structure institutions already trust: a fund, a cleared future, a regulated liability, a money market share, a retirement plan menu. The capital is not resisting crypto. It is resisting unfamiliar market structure.

How the money moves once it is inside

Entry is only the first step. The more instructive question is what the money passes through after it commits, because the plumbing being built for it tells you what institutions actually require.

Custody concentrates, and gets rated. When an ETF takes in cash, an authorized participant sources the underlying, and the coins land with a qualified custodian. Both of those roles deserve a definition, because they are the two workhorses of the entire wrapper economy. An authorized participant, an AP, is a large trading firm or broker-dealer contracted by the ETF issuer with the exclusive right to create and redeem ETF shares; firms such as Jane Street, JPMorgan, and Virtu appear as APs in the spot bitcoin ETF filings. When demand pushes the ETF's price above the value of its holdings, APs create new shares and buy the underlying bitcoin; when it falls below, they redeem. That arbitrage is what keeps the wrapper honest, and it is the actual mechanism by which ETF inflows become crypto buying. A qualified custodian is a regulated trust company or chartered bank that holds client assets under fiduciary custody rules, segregated from its own balance sheet: in crypto, that means firms like Coinbase Custody Trust, Fidelity Digital Assets, BitGo, and Anchorage Digital, operating under state trust or federal bank charters.

That custody business has consolidated dramatically: Coinbase Prime custodies more than $350 billion, including over 80% of US spot BTC and ETH ETF assets, per CoinDesk, which is roughly 12% of the entire crypto market cap under one roof. The interesting development is not the concentration but the scrutiny. Agio Ratings now publishes default-risk ratings on custodians the way agencies rate corporate credit: Fidelity Digital Assets at 0.39% twelve-month probability of default, Anchorage and BitGo at 0.46%, Coinbase Prime at 0.49%. Custody has matured into rated, chartered, boring infrastructure. This is what an asset class growing up looks like.

Banks are re-entering the pipes. For years, accounting guidance known as SAB 121 made it commercially impractical for US banks to custody digital assets. The SEC rescinded it. The OCC, the Office of the Comptroller of the Currency, the Treasury bureau that charters and supervises national banks, then confirmed in Interpretive Letter 1184 that those banks may provide and outsource crypto custody and execution, and the Fed and FDIC withdrew their restrictive statements. The largest balance sheets in finance now have a clear path into crypto plumbing, and the survey data suggests their clients expect them to take it.

Prime brokerage is being rebuilt, venue by venue. This is the layer where the story gets specific. In late 2025, Ripple launched Ripple Prime on its acquisition of Hidden Road, offering US institutions OTC spot execution that can be cross-margined alongside swaps and CME-listed futures. In April 2026, Coinbase's institutional head described cross-margining between spot and derivatives, which cuts client capital requirements by 10-20%, as the final component that made Coinbase a full prime broker, while noting that rivals still force clients to assemble trading, custody, financing, and derivatives from separate providers.

Notice what the competition is about. Not listings, not leverage, not incentives. The race is to give institutions one account, one collateral pool, and execution across many venues, with someone on the client's side of the trade. Traditional finance solved this decades ago and called it brokerage. Crypto is rebuilding it under other names, and every serious firm in the market is converging on the same job description.

And the derivatives flow itself? Derivatives are most of crypto trading, roughly three-quarters of total volume. As of April 2026, per CoinGecko's State of Crypto Perpetuals report, centralized exchanges still carry about 90% of perpetuals volume, with the leading venue alone near a third of it. The on-chain share peaked at 13% in November 2025 and sat near 10% in April.

But volume is a lagging picture. Open interest, the capital that stays committed overnight, tells a different story: perp DEXs' share of open interest grew from 3.6% to 13.5% in roughly a year. Hyperliquid entered the global top ten perp venues, centralized exchanges included, in the first quarter of 2026. Positions are migrating on-chain faster than trades are, which is exactly what you would expect if the traders moving first are the ones holding risk rather than churning it.

So here is the state of the plumbing in one paragraph. Custody: institutional, rated, solved. Cash rails: regulated, arriving on-chain as stablecoins and tokenized funds. Regulated derivatives: growing at record pace, onshore and 24/7. On-chain derivatives: structurally attractive, visibly gaining committed capital, and still carrying a tenth of the flow. The money gets all the way to the last step and stops.

Who allocates next

The forward-looking data comes largely from the EY-Parthenon and Coinbase survey fielded in January 2026, and it is unusually consistent: 73% of institutions plan to increase crypto allocations in 2026, and 74% expect prices to rise over the next year. The question is not whether more TradFi money comes. It is which segments move next, and through which door.

The early adopters, hedge funds and family offices, are already inside and moving past spot. They are the natural users of the options, basis, and funding-rate trades that the ETF and CME complexes now support, and the likeliest first institutional users of on-chain venues.

The middle of the adoption curve is the wealth complex: RIAs, broker-dealer platforms, private banks. Their constraint was never conviction, it was approved product. Registered vehicles solved that, which is why 81% of institutions prefer them. As wirehouse platforms complete diligence, this channel becomes a steady bid rather than an event.

The far end holds the fiduciaries: pensions, endowments, insurers, sovereign wealth. They move last and slowest, but two 2026 developments were built for them specifically: the DOL safe-harbor proposal for retirement plans, and Moody's top ratings on tokenized money market funds. Fiduciaries do not buy narratives. They buy rated products inside safe harbors, and both now exist.

Two more survey findings deserve attention, because they describe the conditions for the next leg rather than the last one.

First, the regulatory paradox: 65% cite improving regulatory clarity as the top reason to allocate more, while 66% simultaneously name regulatory uncertainty as their primary concern, and 78% point at market structure as the area most needing clear guardrails. The CLARITY Act, which would settle the SEC and CFTC jurisdiction question, passed the House a year ago and now sits on the Senate calendar after a bumpy markup season. Meanwhile the CFTC approved the first US-regulated perpetual futures contract in May 2026 and said it will evaluate further perp products case by case. The direction is unmistakable even where the timetable is not; each step of clarity converts some fraction of that 78% from waiting to allocating.

Second, the custody criteria shift, which I find the most revealing numbers in the whole survey. Year over year, the share of institutions naming regulatory compliance as a key custodian-selection factor rose from 25% to 66%, and the share naming security and key-signing protocols rose from 8% to 66%. Cost fell to the bottom. Institutions are no longer asking what crypto exposure costs. They are asking who holds the keys, under what controls, and what happens when a counterparty fails. Those are market structure questions, and they decide where the derivatives flow goes next.

The void: why the flow stops at the last step

If institutions want derivatives exposure, and increasingly want it on-chain, why does 90% of perpetuals volume still sit on centralized venues, and why is most institutional flow still routed through custodial intermediaries?

Because the on-chain derivatives market, as built, asks institutions to accept four things they are structurally unable to accept.

Pre-funding and fragmentation. Crypto liquidity is spread across more venues than any market in financial history: centralized exchanges, on-chain venues across multiple chains, layer-2 networks, all trading the same economic exposure. Each venue demands its own pre-funded collateral. A desk trading three venues holds three stranded pools of capital, bears three counterparties' credit risk, and still finds that the depth it fragmented itself to reach can vanish in stressed conditions. Practitioner analyses are blunt on this point: institutions will not commit flow at scale until execution is deterministic and pricing is consistent across venues.

Custodial counterparty risk. The post-FTX lesson was not subtle: $8 billion of customer assets can disappear inside an opaque venue. The survey's custody numbers show the lesson institutionalized into policy. Yet the dominant derivatives venues still require depositing assets with the exchange as a condition of trading. Every custodial deposit onto a venue is a risk decision that now has to clear a compliance bar two-thirds of institutions now weight explicitly that did not exist two years ago.

Margin that does not travel. In traditional markets, a portfolio nets: the hedge on one venue offsets the position on another, and capital requirements reflect the whole book. On-chain derivatives venues offer no cross-venue netting; capital stranded on venue A cannot defend a position on venue B. It is telling that the loudest recent innovations in crypto prime brokerage, at Coinbase and at Ripple Prime, are both cross-margining products. The market is paying for exactly this, wherever it can find it.

Execution that leaks. Public order books broadcast intent. A large order is information before it is a fill, and on thin, fragmented books the market moves against size before size is done executing. This is why institutional-scale crypto execution has been migrating to OTC and bilateral channels, and why events like the October 2025 deleveraging cascade, which liquidated roughly $19 billion in a day, make risk committees treat public-book perp venues as retail infrastructure. Auto-deleveraging and socialized losses, inherited from the earliest perp designs, are the opposite of the defined loss hierarchies institutions require.

Put the four together and the shape of the void is precise. Institutions need multi-venue access without multi-venue fragmentation, depth without custodial deposits, one collateral pool across venues, and execution that does not advertise their intent. In traditional markets, no allocator solves these problems alone, and none is expected to. The broker solves them: access to every venue through one relationship, competed pricing, margin efficiency across the book, and discretion before the fill.

DeFi built the venues first, hundreds of them, because venues are what crypto knew how to build. It never built the layer institutions actually trade through. That is the void the flows keep stopping in front of.

Where Gryps fits

Gryps is built to be that layer for on-chain perpetuals: the brokerage layer above venues, not another venue beside them.

Concretely, that means five things, each mapped to a friction above.

One account, multiple venues. Gryps brokers access to deep perp liquidity across integrated venues, Binance today with Hyperliquid integration next, so a desk reaches multi-venue depth through a single integration instead of fragmenting itself across exchange accounts.

Cross-venue unified margin. Positions hedged across venues share one collateral pool. Capital defends the whole book, not one venue's silo. This is the capability institutional desks are currently paying custodial prime brokers to approximate, delivered without the custodial part.

Private intents, firm quotes. A trader submits intent privately; competing solvers return firm, executable quotes priced off full venue depth; the trade fills at the quoted price with zero slippage on acceptance. Size is never public pre-trade information. And because solvers hedge into the integrated venues' depth, coverage scales with the venues rather than with any single balance sheet.

Non-custodial bilateral settlement. Settlement is bilateral and on-chain; the trader's collateral never becomes a deposit on an exchange. Counterparty risk is isolated per trade rather than pooled in a venue, which is the only architecture that clears the custody bar institutions have now set. It also produces something no off-chain intermediary can offer: execution and settlement records that live on a public chain, best-execution evidence built into the settlement layer itself.

A programmatic, always-on surface. Intent-based execution is native to how protocols, treasuries, and increasingly agents route risk, and it runs on crypto's clock, the one CME just adopted.

A broker earns its seat by doing three things: providing access, forcing price competition, and sitting on the client's side of the trade. Everything above is one of those three, rebuilt for on-chain settlement.

I want to be direct about what this is not. Gryps is not an exchange competing with the venues it connects to; the venues are the liquidity, and they benefit from flow they could not otherwise serve. It is also not an order-spraying router hoping public books absorb the damage; quotes are competed privately and firm before commitment. The role is older and simpler than either: the intermediary every maturing market eventually builds, arriving in the one market that has conspicuously not built it.

The sequence, and the timing

Step back and the last thirty months form a sequence. Regulated wrappers brought TradFi capital into crypto exposure. Regulated cash rails, stablecoins and tokenized funds, brought TradFi money on-chain. Regulated derivatives brought institutional risk transfer onshore, then around the clock. Committed capital is now visibly migrating to on-chain venues, with open interest share quadrupling in a year even as volume share lags.

Each step moved institutions one layer closer to trading where settlement is native. The remaining distance is not conviction, and after this year it is mostly not regulation either. It is execution infrastructure: the brokerage layer that lets an institution reach every venue, keep its keys, unify its margin, and move size without being seen.

That layer is what we are building. The flows described in this article are arriving on their own schedule, and markets reward execution, not promises. Our work is to make sure that when the money takes its last step, the infrastructure under it holds.