Market Microstructure: Bid-Offer Spreads and Price Formation

Market microstructure is the branch of financial economics concerned with the mechanics of how transactions are carried out in financial markets — how prices are formed, how orders are matched, how liquidity is provided, and how information is incorporated into prices. It sits at the intersection of economics, finance, and market operations, and is directly relevant to everyone who executes, prices, or risks financial transactions.

The bid-offer spread (also called the bid-ask spread) is the difference between the price at which a market maker will buy a security (the bid) and the price at which it will sell (the offer or ask). A market maker quoting a bid of 99.95 and an offer of 100.05 for a government bond has a spread of 10 cents (or 10 basis points of price, or approximately 10 ticks depending on the instrument's tick size).

The spread represents the market maker's gross revenue from facilitating the transaction — the difference between what it pays for the asset and what it receives when it sells it. But the spread is not pure profit. It must cover three categories of cost:

1. Order processing costs. The fixed costs of maintaining a market-making operation — technology, people, compliance, regulatory capital — must be recouped across the flow of transactions. Even if every trade is perfectly hedged, the market maker must earn the spread to cover its operating costs.

2. Inventory risk. A market maker that buys from a seller does not immediately find a buyer. In the interim, it holds inventory that is exposed to market price movements. If it buys 10 million government bonds and the market moves against it before it can hedge or offset the position, it suffers a loss. The spread must compensate for this inventory risk — which is why spreads widen in volatile markets (higher inventory risk) and narrow in calm, liquid markets (lower inventory risk).

3. Adverse selection (information asymmetry). Not all traders who approach a market maker are equally informed. Some are liquidity traders — institutions that need to buy or sell for reasons unrelated to their view on the asset's value (a pension fund rebalancing its portfolio, a company hedging an FX exposure). Others are information traders — participants who trade because they have information suggesting the asset is mispriced. When a market maker trades with an information trader, it is systematically on the wrong side of the trade. The spread must be wide enough to compensate for the losses on information-driven trades by profits on the larger volume of liquidity-driven trades.

Bid-offer spreads are not fixed. They vary across instruments, market conditions, and counterparties based on several factors:

• Liquidity: the most liquid instruments — on-the-run US Treasuries, EUR/USD FX, benchmark equity futures — have the narrowest spreads because high transaction volumes allow market makers to offset positions quickly and the inventory risk is low. Illiquid instruments — an off-the-run corporate bond, an exotic OTC option — carry much wider spreads because the market maker may hold inventory for extended periods.
• Volatility: higher volatility increases inventory risk and adverse selection risk simultaneously. During periods of market stress, spreads widen significantly across virtually all instruments. During the March 2020 COVID shock, bid-offer spreads in the investment-grade credit market widened by a factor of five to ten relative to pre-crisis levels.
• Counterparty identity: in voice markets, market makers often quote tighter spreads to counterparties they know are liquidity traders and wider spreads (or refuse to quote) to counterparties they suspect may be information traders. This practice — known as selective quoting — is less possible in anonymous electronic markets.
• Trade size: market impact increases with size. A £1 million order in a liquid government bond may execute at or very close to the mid-price; a £1 billion order in the same bond will move the market as the market maker is forced to hedge across multiple venues at progressively worse prices.

Market structure — how buyers and sellers connect and how prices are formed — varies significantly across asset classes.

Voice markets operate through direct communication between market participants, typically by phone or electronic chat. The client calls a dealer and asks for a price; the dealer quotes a two-way price (bid and offer); the client decides whether to deal. Voice markets dominate in OTC derivatives, structured products, and less liquid credit markets. They allow market makers to assess the nature of the client's inquiry — size, urgency, counterparty quality — before committing a price. This selectivity reduces adverse selection risk.

Electronic markets operate through automated systems where orders or quotes are published and matched without bilateral negotiation. There are two dominant electronic trading models:

• Order book (Central Limit Order Book / CLOB): all participants can see all outstanding bids and offers. When an incoming order matches an outstanding order, they are automatically executed. Used in equity markets, listed futures, and some FX markets. Provides full transparency of market depth but exposes all participants — including market makers — to the full order book.
• Request for Quote (RFQ): a client sends a request to multiple dealers simultaneously asking for a two-way quote in a specific size. The dealers respond with prices (which may or may not be visible to the other quoting dealers); the client selects the best price. Used in electronic credit and rates markets (Tradeweb, MarketAxess). Preserves some of the information protection of voice markets while automating the execution process.

In any market, prices serve as information aggregators: they incorporate the collective knowledge and expectations of all participants. The process by which new information is incorporated into prices — price discovery — is faster in more transparent, more liquid markets.

Information asymmetry arises when some participants have better information than others. The classic model of market microstructure (Glosten-Milgrom, 1985) formalises this: a market maker trades with a mix of informed traders (who know the true value of the asset) and uninformed traders. The spread must be wide enough to cover the expected losses to informed traders from the profits on uninformed trades. If the proportion of informed traders increases, the spread must widen — which may eventually cause the market to break down if the spread becomes prohibitively wide.

In electronic order books, the matching algorithm — the rule that determines which orders are executed when a new order arrives — matters enormously to market participants, particularly electronic market makers.

FIFO (First-In-First-Out) / Price-Time Priority: orders at the same price are filled in the order in which they were received. The first order in the queue at the best price gets filled first. FIFO incentivises speed — participants race to get their orders into the queue early. It rewards fast, consistent market makers and discourages gaming. Used in equity markets and most futures markets.

Pro-Rata matching: orders at the same price are filled proportionally to their size. A large order at the best bid gets a larger proportion of the incoming sell order than a small order at the same price. Pro-rata incentivises posting large orders, which improves apparent liquidity depth but can attract gaming strategies (posting very large orders with no intention of being fully filled). Used in some fixed income and interest rate futures markets.

The choice of matching algorithm affects market maker behaviour, effective spreads, and the distribution of market-making revenue. Understanding which algorithm applies on a given venue is essential for anyone building electronic market-making systems or evaluating execution quality.