‍A market-structure perspective

Large-cap tokens are often assumed to trade efficiently due to scale, volume, and global participation. In theory, high liquidity and broad exchange coverage should compress spreads and eliminate persistent price differences.

In practice, that assumption does not hold.

Even the largest crypto assets trade with measurable and recurring inefficiencies across centralized exchanges.

Market makers do not run a single global book

Each MM runs separate risk books per exchange, sometimes even per trading pair. Inventory on Binance is not fungible with inventory on Coinbase or OKX in real time.

That means pricing decisions are local.

If an MM is long inventory on Exchange A and short or flat on Exchange B, they will quote differently even if the global reference price is the same. This is intentional risk management, not inefficiency.

So yes, part of the discrepancy is internal and deliberate.

Market Makers Do Not Operate a Single Global Book

Professional market makers do not run one unified inventory across all exchanges.

While exposure and performance are managed at a global portfolio level, capital, margin, and liquidation risk are enforced per exchange. In practice, each venue behaves as a risk silo during periods of stress, as inventory and capital are not fully fungible in real time

This means price alignment across venues is expected as a result from this structural design.

When volatility increases or funding conditions tighten in different cases, these silos become even more pronounced, allowing discrepancies to persist longer than classical arbitrage models would predict.

Exchange-Affiliated Market Making Shapes Local Prices

On many major centralized exchanges, liquidity support is bundled into the listing process.

This can include:

• Exchange-appointed market makers
• Incentivized quoting programs
• Fee rebates tied to spread and depth requirements
• Coordinated marketing and launch liquidity

This is one reason projects pursue top-tier listings beyond brand exposure. The value lies not only in visibility, but in the infrastructure, liquidity support, and market access that come with those platforms.

However, exchange-linked market makers are primarily designed to support local market health. Their mandate is to keep listed tokens active, liquid, and orderly on their own platform, rather than enforcing price alignment across venues, as inactive or illiquid markets reflect directly on the exchange’s public reputation.

They are not incentivized to aggressively arbitrage price differences across other venues.

As a result, a token can trade cleanly within one exchange while remaining misaligned elsewhere.

Project-Hired Market Makers Operate Under Different Constraints

When a token lists on a centralized exchange, liquidity does not automatically become global or efficient. Listing creates access, but how liquidity behaves afterward depends on deliberate structural choices made by the project.

In practice, teams usually choose between three approaches:
• Relying on exchange-provided liquidity programs
• Hiring independent market makers
• Using a hybrid of both

Exchange liquidity programs are often the default starting point. Many top-tier exchanges bundle in-house market making into the listing package, designed to ensure continuous quoting, visible depth, and acceptable spreads on their own venue. This avoids thin books or inactive markets immediately after listing, which can negatively affect both the project’s credibility and the exchange’s reputation.

Independent market makers operate differently. They deploy capital across multiple exchanges simultaneously, arbitraging discrepancies and tightening global pricing over time. In theory, this is what drives convergence across venues. In practice, these strategies are capital-intensive and operationally complex.

Independent desks face higher infrastructure costs, exposure to withdrawal delays, counterparty risk across exchanges, and balance sheet constraints that become binding during volatility spikes. When conditions deteriorate, capital is reduced, transfers slow, or deployment pauses entirely, allowing price differences to persist.

This is why even large-cap tokens can trade at sustained premiums or discounts across exchanges after listing. During periods of stress, the risk of capital being trapped often outweighs the marginal profit of closing small gaps.

As a result, many projects follow a phased approach. They begin with exchange-provided liquidity to ensure immediate activity and visible depth, then gradually onboard independent market makers as volume, volatility, and treasury capacity allow. Cross-exchange alignment typically improves over time, often months after the initial listing, as operational frictions ease and external desks scale exposure.

This dynamic reflects how crypto markets function today. Liquidity is built locally first, then stitched together globally as capital, infrastructure, and risk tolerance mature.

Arbitrage is gated by operational risk

In theory, cross-exchange price differences should close quickly.

In practice, arbitrage is limited by operational risk rather than pricing logic.

Executing an arbitrage trade requires more than identifying a spread. It depends on reliable inventory movement and balance sheet safety. This includes predictable withdrawal times, functional deposit rails, stable margin requirements, and confidence that capital will not be trapped during transfer.

When asset transfers take minutes or hours instead of seconds, market makers must account for tail risk. A trade that appears profitable at entry can turn negative if withdrawals pause, queues form, or collateral rules change while inventory is in transit.

Because of this, arbitrage desks operate with tolerance bands instead of enforcing strict price parity. Small premiums or discounts are allowed to persist when the risk of being stranded on one venue outweighs the marginal gain from closing the gap.

These frictions surface most clearly during volatility spikes, exchange maintenance periods, regional outages, or regulatory uncertainty. Even well-capitalized firms reduce activity under these conditions, favoring capital preservation over tight cross-venue alignment.

As a result, large-cap tokens can show persistent price differences across exchanges even when multiple market makers are active. The limitation is not competition or intent. It is transfer risk, settlement uncertainty, and prudent risk management.

Participant flow differs by exchange

Even when the same market makers operate across multiple exchanges, order flow is rarely balanced.

Liquidity naturally concentrates where users are most active. Exchanges with larger and more established user bases attract steadier volume, deeper participation, and higher turnover. Geographic access also plays a role. Regional platforms often capture local demand because they integrate more easily with domestic banking systems or are more familiar to users in those markets.

This leads to persistent differences in flow. Some venues skew retail-driven, others attract leverage-heavy positioning, while certain exchanges see activity cluster around specific regional trading hours. Funding dynamics and liquidation events further reinforce these patterns.

Market makers adjust quotes based on observed order flow rather than theoretical fair value. When sustained buying pressure appears on one exchange, prices move higher there instead of being immediately flattened across venues.

This behavior reflects adaptive pricing in response to user activity, regional demand, and exchange-specific participation.

Risk systems override price convergence

During periods of market stress, price alignment across exchanges is no longer the primary objective for market makers.

Risk systems take priority.

When volatility accelerates, internal controls adjust automatically. This typically includes:

• Tighter inventory limits
• Stricter VAR thresholds
• Exposure caps by exchange
• Correlation controls across related assets

Once these constraints are triggered, closing cross-exchange gaps becomes secondary to protecting capital.

This is also why market makers may sell during sharp declines. That behavior is often misunderstood as active distribution, but it is usually inventory risk being reduced to stay within system limits, not a discretionary view on the asset.

As risk thresholds bind, pricing tolerance widens by design. Spreads are allowed to persist, premiums are not immediately arbitraged away, and convergence is deferred until volatility subsides and risk budgets reopen.

This is why the largest pricing discrepancies tend to appear precisely when volume and volatility peak.

‍