Crypto OTC vs Order Book 2027: At What Trade Size Does OTC Become Cheaper?

DN research object: DN OTC Break-Even Index 2027
Methodology: DN-OTCBE v1.0
Reviewed: 15 September 2026

What Matters

There is no universal dollar amount at which OTC automatically becomes cheaper than an exchange order book.

The crossover depends on:

order-book depth + spread + trading fee + market impact + volatility + information leakage + OTC quote spread + settlement costs + execution urgency.

A $1 million BTC trade can be cheaper on a deep public order book during liquid conditions while a $250,000 trade in a thinner asset may already obtain better economics through RFQ.

The DN OTC Break-Even Notional is therefore defined as:

The smallest trade size at which the all-in expected cost of an executable OTC/RFQ quote becomes lower than the all-in cost of executing the equivalent order through the public order book.

Rather than assuming OTC is better because the trade is “large,” DN measures where the crossover actually occurs.

Key Findings

• OTC does not mean free execution. Kraken currently charges no separate OTC trading fee, for example, but states that its quoted bid or offer is the all-inclusive price. Binance similarly says its OTC service does not levy a separate trading fee. The economically relevant comparison is therefore the quoted price versus a realistic exchange VWAP, not “0% fee” versus an exchange fee.
• OTC access thresholds vary dramatically. Kraken currently lists a $50,000 minimum for OTC, while Binance says its current minimum generally starts at $200,000.
• RFQ is increasingly blurring the line between exchange and OTC execution. Coinbase Prime’s current RFQ system conducts a competitive LP auction off the central limit order book, returns an all-inclusive fixed quote and holds that quote for three seconds in its UI or one second through the API.
• Block trading does not always require multimillion-dollar tickets. OKX currently lists a $1,000 minimum block notional, although its separate onboarding guidance says users need an account balance equivalent to at least $100,000 to access Liquid Marketplace block trading.
• Academic evidence supports treating off-book execution as genuinely different market microstructure. A 2026 study using Gemini Bitcoin trades found that large trades executed away from the continuous order book had lower price impact than comparable large trades executed directly on the book in much of the sample, although the study also found important nuances around the exchange’s block-size rule change.
• OTC is not automatically superior. A deep BTC order book, low fee tier and patient algorithmic execution can beat a mediocre OTC quote.
• The correct comparison is therefore not fee versus fee. It is final executable value versus final executable value.

The Wrong Question Is “How Big Does a Trade Need to Be for OTC?”

You will often hear numbers such as:

$50,000

$100,000

$250,000

$1 million

presented as though they represent some universal OTC threshold.

They do not.

Those numbers may represent:

• a desk’s minimum ticket;
• an exchange’s eligibility requirement;
• a salesperson’s preferred client profile;
• an arbitrary industry convention.

They do not tell us where OTC becomes economically superior.

Consider two BTC markets.

Market A

$5 million of genuine executable liquidity sits within 5 basis points of midpoint.

A $500,000 marketable order may barely move the book.

Market B

Only $150,000 sits within 5 basis points.

The same $500,000 order sweeps several levels.

The appropriate OTC crossover can therefore be radically different despite identical transaction size.

DN Introduces the OTC Break-Even Notional

Let:

CLOB(N) = all-in public order-book cost at notional N.

OTC(N) = all-in private/RFQ cost at notional N.

Then:

DN OTC Break-Even Notional

N* = smallest N where OTC(N) ≤ CLOB(N)

Everything before N* favors the order-book model under the assumptions tested.

Everything beyond it favors the OTC quote until the underlying cost curves change again.

This is not necessarily one permanent number.

It can move minute by minute.

The Public Order Book Cost Stack

Most comparisons make the mistake of looking only at the trading fee.

DN separates the public execution cost into:

1. Half-Spread

If BTC is:

Bid: $69,990

Ask: $70,010

then midpoint is:

$70,000

An immediate buyer starts approximately:

1.43 basis points

above midpoint.

That cost exists before meaningful book impact begins.

2. Exchange Trading Fee

The effective taker rate depends on:

• venue;
• user tier;
• rolling volume;
• asset;
• maker/taker classification;
• promotional pricing;
• institutional programme.

DN therefore does not hard-code a supposedly universal exchange fee into the benchmark.

The calculator lets users enter their actual fee tier.

3. Order-Book Sweep Cost

Suppose a trader wants to buy $1 million of BTC.

The displayed best ask might contain only:

$80,000

The next level another:

$100,000

Then:

$200,000

Then:

$300,000

and progressively worse prices.

The relevant execution price is not the best ask.

It is:

Volume-Weighted Average Execution Price

or VWAP across every price level consumed.

DN Order-Book Sweep Basis

For a buy:

Sweep Cost = (VWAP − Reference Mid) ÷ Reference Mid × 10,000

For a sell:

Sweep Cost = (Reference Mid − VWAP) ÷ Reference Mid × 10,000

If BTC midpoint is:

$70,000

and a $2 million purchase executes at an average:

$70,105

then market-impact/slippage cost is approximately:

15 bps

before trading fees.

4. Timing Cost

An institution does not necessarily have to sweep the entire order book immediately.

It can execute using:

TWAP

VWAP

iceberg

passive limit orders

smart order routing

adaptive execution

Coinbase Prime, for example, currently provides algorithms including TWAP, VWAP and adaptive execution and aggregates liquidity across multiple venues rather than requiring every institutional order to hit one visible book.

That can reduce market impact.

But it introduces another cost:

Market Risk While Waiting

If execution takes 30 minutes, the asset can move against the trader during those 30 minutes.

There is no free execution technique.

You are often exchanging:

impact risk

for:

timing risk.

5. Information Leakage

Large visible orders can reveal demand.

Imagine a trader repeatedly buying BTC in large clips.

Other market participants may detect:

• persistent aggressive flow;
• order replenishment;
• repeated iceberg activity;
• declining ask depth.

Liquidity providers can respond.

Spreads may widen.

Offers may be pulled.

Other participants can trade ahead of expected future demand.

This effect is harder to quantify than explicit fees, but it is part of execution economics.

Why OTC Exists

An OTC/RFQ execution changes the problem.

Instead of announcing:

“I want to consume $5 million from this visible order book.”

the client asks one or more liquidity providers:

“What firm price will you give me for this entire block?”

The market maker internalizes:

• inventory risk;
• hedge cost;
• expected market impact;
• volatility;
• funding;
• liquidity;
• capital usage;
• profit margin.

and returns a price.

The user then decides whether that price is better than alternative execution.

OTC Quote Cost

For a buy:

OTC Quote Cost = (OTC Price − Reference Mid) ÷ Reference Mid × 10,000

For a sell:

OTC Quote Cost = (Reference Mid − OTC Price) ÷ Reference Mid × 10,000

Then add any:

• explicit commission;
• settlement fee;
• financing cost;
• withdrawal cost;
• operational cost.

The result is:

DN OTC All-In Cost

“Zero OTC Fee” Does Not Mean Zero Cost

This distinction is especially important.

Kraken currently states that its OTC desk charges no separate fee because the price quoted is all-inclusive.

Binance similarly says its OTC spot service has no separate trading fee.

That does not mean:

OTC cost = 0 bps

The market maker still needs an economically acceptable price for assuming and hedging the block.

The proper question is:

How far is the firm OTC quote from a fair contemporaneous reference price?

That is the number we compare with the public book.

A Simple Example

Assume BTC midpoint:

$70,000

You want to buy:

$1 million

Order Book

Average sweep price:

$70,077

Market impact:

11 bps

Trading fee:

5 bps

All-in direct cost:

approximately:

16 bps

or:

$1,600

OTC

Firm quote:

$70,070

Quote basis:

10 bps

No separate trading commission.

All-in cost:

approximately:

10 bps

or:

$1,000

Under these simplified assumptions:

OTC saves approximately $600.

But change the book.

If $1 million can execute with only:

2 bps

of impact,

then:

Order book:

2 + 5 = 7 bps

OTC:

10 bps

Now the public market wins.

The transaction amount has not changed.

Liquidity has.

DN Block Trade Savings

DN therefore defines:

Block Trade Savings

BTS = Order Book All-In Cost − OTC All-In Cost

If:

CLOB cost:

16 bps

OTC:

10 bps

then:

BTS = +6 bps

OTC is cheaper by six basis points.

If:

CLOB:

7 bps

OTC:

10 bps

then:

BTS = −3 bps

The public book is cheaper.

At $5 million:

6 bps = $3,000

Small differences matter quickly.

DN Visible Depth Consumption Ratio

Another useful metric is:

Visible Depth Consumption Ratio

VDCR = Order Notional ÷ Visible Depth Within Selected Impact Band

Suppose:

Trade:

$1 million

Visible BTC liquidity within 10 bps:

$2.5 million

VDCR:

40%

Now suppose available depth is only:

$300,000

VDCR:

333%

The latter order cannot execute inside that band using the observed book.

That is a much stronger signal that private liquidity or slower execution may deserve consideration.

There Are Now More Than Two Execution Choices

A modern institutional trader does not simply choose:

exchange

versus:

guy on Telegram claiming to run an OTC desk.

The genuine execution spectrum is more sophisticated.

This is why the article should not frame OTC as universally sophisticated and public markets as universally unsophisticated.

Kraken OTC

Operational status: LIVE

Kraken currently lists an OTC minimum equivalent to $50,000 and says there is no maximum size for chat-based trades.

The desk supports both high-touch chat execution and self-service RFQ.

Kraken says its OTC desk does not levy a separate trading fee because the bid or offer supplied is the all-inclusive price.

Settlement can occur:

• through a Kraken account;
• by external crypto wallet;
• and through supported fiat settlement routes, depending on eligibility.

That gives DN a valuable benchmark venue because it exposes a meaningful transition point around the $50,000 level.

Binance OTC & Execution Services

Operational status: LIVE

Binance’s current Execution Services documentation lists a typical minimum starting at:

$200,000 equivalent

although smaller transactions may occasionally be accommodated.

Its current OTC FAQ says no separate OTC trading fee is charged.

Binance also reported a notable institutional execution example in May 2026:

$33.5 million BTC/USDC

executed across:

three clips

in under:

20 minutes

with Binance claiming the result achieved tighter pricing than a single order-book execution would have delivered. That is useful evidence of the type of execution service offered, but DN classifies the performance statement as PLATFORM-REPORTED, not independent benchmark evidence.

Coinbase Prime RFQ

Operational status: LIVE

Coinbase Prime provides an interesting middle ground between traditional OTC and electronic exchange execution.

Its RFQ system runs a short competitive auction among liquidity providers.

The winning quote is:

• fixed;
• all-inclusive of fees;
• off the central limit order book;
• fill-or-kill.

Current quote holds are:

3 seconds through the UI

and:

1 second through the API.

If the client does not accept during that period, the quote expires.

This creates a fascinating future DN benchmark:

RFQ Quote Decay vs CLOB Quote Decay

The user has price certainty during the RFQ hold.

But the public order book continues moving.

The relative value of those seconds can itself be measured.

OKX Liquid Marketplace

Operational status: LIVE

OKX’s Liquid Marketplace provides RFQ and block-trading infrastructure.

Current documentation defines block trades as privately negotiated transactions intended to reduce the slippage risk of putting a large position directly through an exchange book.

Its block-trading documentation currently states a minimum block notional equivalent to:

$1,000

while a separate updated onboarding guide says users require an account balance equivalent to at least:

$100,000

to access Liquid Marketplace block trading.

These numbers describe different things:

ticket minimum

versus:

access requirement.

That distinction is exactly the kind of detail ordinary “best OTC desk” articles tend to collapse.

Bybit RFQ

Operational status: LIVE

Bybit currently operates an RFQ system for block trading across:

• spot;
• futures;
• options;
• multi-leg strategies.

It describes the system as private execution directly with liquidity providers rather than public order-book matching.

Bybit markets the design as avoiding public-book market impact.

DN would classify that as an architectural benefit claim until independent execution comparisons measure the actual all-in quote against a synchronized public-book alternative.

The Most Important Point: Compare Against VWAP, Not Best Bid/Ask

Suppose the displayed ask is:

$70,010

and the OTC desk quotes:

$70,080

It is tempting to conclude:

OTC is 10 bps worse.

But perhaps the book contains only:

$25,000 at $70,010

The $5 million client cannot buy $5 million at that price.

Maybe the actual CLOB sweep VWAP is:

$70,150

Now the supposedly expensive OTC quote becomes superior.

DN therefore prohibits this comparison:

OTC quote vs best ask

for block execution.

The correct comparison is:

OTC firm quote vs simulated or realized CLOB VWAP

for the same notional and timestamp.

A New DN Metric: Phantom Top-of-Book Liquidity

This creates another proprietary research concept.

Phantom Top-of-Book Advantage

The difference between:

cost inferred from best bid/ask

and:

cost of actually completing the full requested size.

A venue can advertise:

1 bp spread

while a $2 million order experiences:

30 bps of VWAP impact.

The 1 bp number is real.

It is simply irrelevant to the $2 million trader.

$100K, $250K, $500K, $1M, $5M and $10M Must Be Separate Cohorts

I would standardize DN testing around:

$100,000

Entry institutional / HNW block.

$250,000

OTC becomes accessible across more venue structures.

$500,000

Meaningful market-impact test.

$1 million

Institutional execution.

$5 million

Deep-liquidity stress.

$10 million

True large-block execution.

Where a venue cannot produce a valid executable quote:

NO QUOTE

is recorded.

That is information.

BTC and ETH Should Be Tested First

The initial live study should use:

BTC/USD or BTC/USDT

and:

ETH/USD or ETH/USDT

depending on venue comparability.

Why start with deep majors?

Because if OTC beats the public market even where order books are deepest, that is important.

Then expand into:

• SOL;
• XRP;
• HYPE;
• selected liquid altcoins;
• selected long-tail assets.

The crossover should generally become more interesting as liquidity thins.

DN OTC Break-Even Curve

The final visualization should look like this conceptually: