The bid-ask spread is the gap between the highest price a buyer is currently willing to pay (the bid) and the lowest price a seller is willing to accept (the ask/offer); slippage is the broader difference between the price a trader expected when deciding to trade and the price they actually got when filled. The spread is the visible, quoted toll on every round trip; slippage is the full, often larger, realized cost that also captures price movement and market impact between decision and execution. Together they are the dominant "hidden" transaction cost of trading — the single biggest reason a strategy that looks profitable on closing-price backtests can lose money live. Their core tension: tighter execution (limit orders, patience) reduces cost but risks not filling at all, while certain execution (market orders, urgency) guarantees a fill but pays the spread and accepts adverse slippage.

How it's calculated / formed

Quoted spread = best ask − best bid, often expressed in percentage terms as (ask − bid) / midpoint. The midpoint (ask + bid) / 2 is the conventional "fair value" benchmark.

Effective spread measures the cost actually paid: 2 × |execution price − midpoint at order arrival|. It is usually narrower than the quoted spread because of price improvement — dealers and internalizers frequently fill inside the quote (Wikipedia; Berkeley DeFi survey). Effective spread is the honest number; quoted spread overstates cost.

The spread exists to compensate liquidity providers, and microstructure theory decomposes it into three components (de Jong & Rindi; Glosten-Milgrom 1985):

• Order-processing cost — the operational cost of making markets (systems, capital, clearing).
• Inventory-holding cost — compensation for the risk of holding an unwanted position; rises with volatility and falls with volume.
• Adverse-selection cost — the Glosten-Milgrom insight: market makers lose to better-informed counterparties, so they widen the spread to break even on average. This component produces a positive spread even for a risk-neutral, zero-profit dealer, purely from information asymmetry.

Slippage is typically measured as implementation shortfall (Perold 1988, The Implementation Shortfall: Paper vs. Reality): the difference between the "decision price" (the price when the trade was decided) and the volume-weighted average execution price, including commissions, market impact, timing-delay cost, and the opportunity cost of unfilled shares. For a large parent order ("metaorder"), the dominant slippage term is market impact — the price drifting away as your own order consumes liquidity. Arrival slippage on a buy is expected to be adverse (negative) on average, because execution itself pushes price up (Wikipedia, Implementation shortfall).

How it's used in practice

Spread and slippage are inputs to two decisions: whether the edge survives costs, and how to execute.

• Cost vetting. Subtract realistic round-trip cost (spread + expected slippage + commissions) from a strategy's gross edge per trade. A setup that nets 0.4% on average is destroyed by a 0.5% effective spread; one that nets 8% over weeks barely notices it. This is why high-frequency and scalping strategies live or die on execution while position trades are nearly cost-insensitive.
• Order-type choice. Market orders cross the spread and pay it immediately; limit orders post liquidity and can earn price improvement but carry fill risk and adverse selection (your resting bid fills exactly when the stock is falling). Marketable limits and midpoint pegs are common compromises.
• Order slicing. Because market impact is concave in size, large orders are split over time (VWAP, TWAP, implementation-shortfall algos) to reduce footprint. The square-root law of market impact — impact ∝ √(order size / volume) — is one of the most robust empirical regularities in microstructure (Bouchaud; arXiv surveys): doubling size raises impact by roughly √2 (~1.41×), so slicing always helps relative to a single block, though it trades impact cost for timing/opportunity risk.
• Liquidity filtering. Traders avoid or size down in names where the spread is wide relative to expected move, screening on average spread, dollar volume, and depth at the inside.
• Timing. Spreads are widest at the open, near the close, around news, and in thin overnight/pre-market sessions; they are tightest mid-session in liquid names. Patient traders concentrate execution in liquid windows.

Standing & evidence

Spread and slippage are not contested — they are measured market facts and the foundation of Transaction Cost Analysis (TCA), which institutions are effectively required to perform for best-execution compliance. The square-root impact law is well-documented across many markets and asset classes, with the exponent empirically near 0.5 (Bouchaud substack; arXiv:2411.13965, Tokyo Stock Exchange survey). What varies enormously is magnitude: large-cap US equities (e.g., mega-caps) routinely quote a one-cent spread, a tiny fraction of a percent, while illiquid micro-caps, options far from the money, and thin overnight markets can show spreads of several percent. Any single "typical spread" number is misleading without naming the instrument and time — treat published averages as order-of-magnitude, not precise.

The honest caution is the opposite of folklore: retail traders systematically underestimate total trading cost because they see commissions (now often zero) and ignore the spread and slippage, which are the real bill. Zero-commission does not mean zero-cost — order flow is frequently monetized via payment for order flow, and execution quality is the thing to scrutinize.

Strengths & limitations

• Where the concept "works": as a budgeting and execution discipline it is always valid — every trade pays it, and modeling it correctly is what separates a realistic backtest from a fantasy one. The single most valuable use is forcing a cost-adjusted edge calculation before trading.
• When estimates fail: historical/average spread badly understates real cost in stress — during volatility spikes, gaps, and liquidity vacuums, spreads blow out and slippage on stops can be catastrophic (a stop "at" a level can fill far below it). Costs are state-dependent, not constant.
• #1 misuse: backtesting on closing or midpoint prices with no spread/slippage assumption, producing an equity curve that is unachievable live. The second most common: trusting quoted spread as the cost when effective spread (with price improvement) or, for size, market impact is the relevant figure. A close third: using market orders in thin names and treating the resulting slippage as bad luck rather than a predictable tax.
• Regime dependence: highly liquid, calm, regular-session conditions → costs negligible for most discretionary timeframes. Illiquid, volatile, off-hours, or large-relative-to-volume conditions → costs can exceed the entire expected edge.

System relevance

This node sits in Market Structure & Mechanics and underpins any Delvantic system that scores or simulates trade outcomes. For the Augustus trade-setup agent the operative caveat is liquidity gating: a technically clean setup on a name with a wide spread relative to its expected move is a worse trade than the chart suggests, and a stop placed in a thin name will slip on the days it matters most. Augustus should treat spread/impact as a cost subtracted from any candidate setup's expected edge and as a hard filter on universe selection, not as a footnote. It pairs with the sibling Liquidity & Volume and Order Types nodes (which cover depth and execution tactics in detail) — cross-link rather than duplicate.

Sources

• Wikipedia — Bid–ask spread; Slippage (finance); Implementation shortfall
• A. Perold (1988), The Implementation Shortfall: Paper vs. Reality, J. Portfolio Management — canonical definition of slippage as decision-to-execution shortfall
• Glosten & Milgrom (1985), via ScienceDirect / Glosten & Harris (1988, JFE) — spread component decomposition (order-processing, inventory, adverse-selection)
• Berkeley DeFi course material — Bid-Ask Spreads: Measuring Trade Execution Costs (quoted vs effective spread, price improvement)
• Optiver explainer — Bid-ask spread (liquidity / market-maker compensation)
• J-P. Bouchaud, The Square-Root Law of Market Impact (substack); arXiv:2411.13965 (Tokyo Stock Exchange empirical survey) — concave impact, exponent ≈ 0.5
• Markets4you / TCA references — measuring spread, slippage, and market impact; mitigation via limit orders, slicing, liquid-hours timing

Flag: spread/slippage magnitudes are highly instrument- and regime-specific; figures here are qualified order-of-magnitude statements, not precise constants.