Cryptocurrency Markets and Quant Trading › Market Structure

Cryptocurrency market structure differs fundamentally from the traditional exchange models we discussed in earlier chapters. Understanding these differences is essential for developing effective trading strategies.

Centralized exchanges function similarly to traditional exchanges, maintaining order books and matching buyers with sellers. However, there are critical differences:

• Custodial model: When you deposit assets on a centralized exchange, you transfer ownership of those assets to the exchange. The exchange maintains an internal ledger of your balance, and you must trust them to honor withdrawals. This creates counterparty risk that simply does not exist when trading equities through a regulated broker.
• Global fragmentation: Unlike equities, where a single exchange typically dominates trading in each stock, cryptocurrency trading is fragmented across hundreds of exchanges worldwide. This fragmentation creates arbitrage opportunities but also complicates execution.
• Variable fee structures: Exchange fees typically follow a maker-taker model, with fees ranging from 0.01% to 0.5% depending on the exchange and trading volume. This is significantly higher than equity markets, where institutional traders might pay fractions of a basis point.

Major centralized exchanges include Binance (the largest by volume), Coinbase (the most prominent US-regulated exchange), and Kraken. Each exchange has different trading pairs, fee structures, and liquidity profiles.

Decentralized exchanges represent a fundamentally different approach to trading. Instead of a central order book, most DEXs use automated market makers (AMMs) that provide liquidity through smart contracts.

The concept of an automated market maker emerges from a simple but powerful question: what if we could replace human market makers with mathematical formulas? In traditional markets, market makers are sophisticated firms that continuously quote bid and ask prices, managing their inventory and adjusting quotes based on market conditions. An AMM encodes this behavior into an immutable smart contract, removing the need for human intermediaries entirely. The mathematical formula determines prices automatically based on the current state of the liquidity pool, and anyone can trade against this formula at any time.

The AMM pricing mechanism creates unique dynamics derived from the constant product formula $x \cdot y = k$. To understand how prices emerge from this simple equation, let us walk through the mechanics step by step. When you want to swap token X for token Y, you add $\Delta x$ units of X to the pool. The pool's X reserve increases to $x + \Delta x$, but the constant product constraint demands that $k$ remain unchanged. This means the Y reserve must decrease to compensate. Solving algebraically, the new Y reserve must equal $y' = k/(x+\Delta x)$, and you receive the difference between the old and new Y reserves.

The price impact measures the percentage deviation of the effective execution price $P_{\text{eff}}$ from the spot price $P_{\text{spot}}$. We derive this by first calculating the amount of tokens received $\Delta y$ (based on the constant product $k = xy$), then comparing the effective price to the spot price. This derivation reveals why large trades face increasingly unfavorable prices, a phenomenon that will shape our understanding of execution costs in decentralized markets:

$$\begin{aligned} \Delta y &= y - \frac{xy}{x+\Delta x} = y \left(1 - \frac{x}{x+\Delta x}\right) = y \frac{\Delta x}{x+\Delta x} \\ P_{\text{eff}} &= \frac{\Delta y}{\Delta x} = \frac{y}{x+\Delta x} \\ \text{Price Impact} &= \frac{P_{\text{spot}} - P_{\text{eff}}}{P_{\text{spot}}} \\ &= \frac{y/x - y/(x+\Delta x)}{y/x} \\ &= 1 - \frac{x}{x+\Delta x} \\ &= \frac{\Delta x}{x + \Delta x} \end{aligned}$$

where:

• $\Delta x$: the trade size (amount of token being sold)
• $x$: the current reserve of the token being sold in the pool
• $y$: the reserve of the other token in the pool
• $P_{\text{spot}}$: the theoretical market price ($y/x$) before the trade
• $P_{\text{eff}}$: the realized average price of the trade

The final expression $\frac{\Delta x}{x + \Delta x}$ provides a remarkably intuitive result: the price impact equals your trade size as a fraction of the post-trade reserve. A trade of 10 units into a pool with 100 units would face approximately 9% price impact (10/110), while a trade of 1 unit would face only about 1% impact (1/101). This formula demonstrates that execution price worsens non-linearly as trade size increases relative to liquidity. The relationship is concave, meaning the first units traded suffer relatively little impact, but marginal impact accelerates as the trade size grows. This means large trades face significant price impact, and the market has no concept of a "best bid" or "best offer" in the traditional sense. The following function constant_product_amm implements this pricing logic, calculating the new reserves resulting from a trade and deriving the price impact by comparing the effective execution price to the initial spot price:

Key ParametersLink Copied

The key parameters for the Constant Product AMM are:

• x: Reserve of the first token (e.g., ETH). High reserves reduce price impact because the same trade size represents a smaller percentage of the pool.
• y: Reserve of the second token (e.g., USDC). The ratio y/x determines the spot price at which infinitesimally small trades would execute.
• k: The constant product invariant ($x \cdot y$). This remains fixed during a trade (ignoring fees), serving as the fundamental constraint that generates the pricing curve.
• $\Delta x$: The trade size. Larger trade sizes relative to reserves cause exponential price impact because each marginal unit must be purchased at an increasingly unfavorable price along the hyperbolic curve.

Perhaps the most operationally significant difference for you is that cryptocurrency markets never close. Trading occurs 24 hours a day, 7 days a week, 365 days a year. This creates challenges across multiple dimensions:

• Infrastructure requirements: Systems must run continuously without maintenance windows. Unlike equity trading where you can perform updates over weekends, crypto systems require redundancy and hot-swappable components.
• Staffing and monitoring: Human oversight cannot follow the markets continuously. This places greater emphasis on automated risk controls and alerting systems.
• Volatility timing: Some of the largest price moves occur during what would be "off-hours" in traditional finance, often during weekend periods when liquidity is thinner.
• Funding and margin: Positions must be monitored for margin requirements at all times, not just during market hours.