Skip to content
Financial Context Wiki

Options Bid-Ask Spread: Quotes, Execution Cost, and Market Quality

Analyze option bid-ask spreads through quote provenance, displayed depth, signed effective and realized spreads, price impact, executable package prices, and liquidation marks.

Updated

For educational purposes only; not investment advice. Investing may result in loss.

Direct answer

For one exact option series at one timestamp, the bid is the highest eligible displayed price to buy and the ask or offer is the lowest eligible displayed price to sell. The quoted spread is Q = A − B, and the arithmetic midpoint is M = (A + B) ÷ 2. A relative spread such as Q ÷ M is meaningful only when the quote is two-sided, synchronized, uncrossed and has a positive midpoint.

The quote must be identified as a consolidated options NBBO, a venue BBO, a proprietary-feed quote or a broker display. It describes the displayed best prices and associated size that survived the relevant feed rules at that moment, not the full order book, hidden interest, auction responses, complex-order liquidity or a guaranteed fill. A locked, crossed, one-sided, stale, delayed or closed-market display requires separate treatment rather than blind use of a midpoint.

The spread is not a separate fee and is not automatically the cost of one trade. A marketable buyer commonly pays above the arrival midpoint and a marketable seller below it, but price improvement, resting-limit execution, partial fills, routing, depth, market movement and fees change the result. Best execution is a broader order-handling duty; observing or beating one displayed quote does not by itself prove that duty was satisfied.

Seven-step quote, execution, and measurement process

  1. Lock the exact contract and economic unit. Match underlying or index, call or put, expiration, strike, exercise and settlement terms, session, regular or adjusted deliverable, quote multiplier, contract count and currency. Premium quotations, exercise cash, deliverable units and account P&L can use different scales.
  2. Identify quote provenance and state. Record OPRA-consolidated NBBO, local venue BBO, direct or proprietary feed, vendor transformation, timestamp, timezone, latency, condition code, session, opening or reopening state, halt and whether the market is two-sided, locked, crossed or one-sided. Preserve price, size and venue attribution together.
  3. Compute descriptive spread measures. For valid bid B and ask A, calculate M = (A + B) ÷ 2, Q = A − B, relative spread Q ÷ M, and dollar width Q × quote multiplier × contracts. Treat the last calculation as full quoted width, not a promised round-trip loss or a separate charge.
  4. Reconstruct the actual order lifecycle. Capture decision time, broker receipt, route, venue receipt, marketability, limit, time-in-force, auction or complex-book handling, each fill, cancellation, reroute, correction and report. Compute size-weighted execution price and preserve unfilled quantity; top-of-book price without depth is incomplete.
  5. Measure cost with explicit sign and benchmark. Let d = +1 for a buyer-initiated trade and d = −1 for a seller-initiated trade. With fill price P, arrival midpoint M_0 and a synchronized later midpoint M_Δ, define full-spread-equivalent measures effective spread = 2d(P − M_0), realized spread = 2d(P − M_Δ), and price impact = 2d(M_Δ − M_0), so effective spread equals realized spread plus price impact. Also calculate one-way midpoint shortfall, commissions, exchange fees and an appropriate decision-price implementation shortfall.
  6. Separate marks from executable value. Distinguish last sale, midpoint, broker mark, model value, official settlement, accounting fair value and an executable liquidation bid or ask. For a long, immediate liquidation references an executable bid; for a short, it references an executable ask. For multiple legs, use the signed package bid, ask, ratios and quantity from a complex book or auction when available rather than assuming all leg midpoints are jointly executable.
  7. Compare distributions and validate controls. Segment by exact series, price, moneyness, tenor, side, size, marketability, session, volatility, event, venue eligibility and order instructions. Compare fill rate, partial fills, speed, price improvement, effective and realized spread, impact and total shortfall; retain data and code versions, inspect outliers, and stress wider spreads, thinner depth and failed fills.

Worked examples

  • NBBO, execution, and liquidation mark. A standard option is quoted B = $2.40 × 8 and A = $2.60 × 3, so M = ($2.40 + $2.60) ÷ 2 = $2.50, Q = $2.60 − $2.40 = $0.20, and Q ÷ M = 8.0000%. One marketable buy fills at $2.55, giving price improvement of ($2.60 − $2.55) × 100 = $5 and buyer effective spread 2 × ($2.55 − $2.50) = $0.10, or $10 per contract. Later the quote is $3.00 bid / $3.30 ask, with midpoint $3.15. Midpoint-marked gain is ($3.15 − $2.55) × 100 = $60, but executable sale at the bid gives ($3.00 − $2.55) × 100 = $45; the liquidation haircut to the midpoint mark is $60 − $45 = $15.
  • Displayed depth and partial-price levels. At arrival the quote is $1.00 bid / $1.10 ask, so midpoint is $1.05. Assume static displayed offers of 2 contracts at $1.10, 3 at $1.12, and 5 at $1.18. A marketable buy of eight that fills two, three and three contracts has VWAP = [(2 × $1.10) + (3 × $1.12) + (3 × $1.18)] ÷ 8 = $1.1375. Midpoint shortfall is ($1.1375 − $1.05) × 100 × 8 = $70; disimprovement from the initial best ask is ($1.1375 − $1.10) × 100 × 8 = $30. This static ladder is an arithmetic illustration, not a routing or fill prediction.
  • Signed effective spread, realized spread, and price impact. A buyer-initiated trade has d = +1, arrival quote $4.90 bid / $5.10 ask, M_0 = $5.00, and fill P = $5.04. Five minutes later, the synchronized midpoint is M_Δ = $5.12. Effective spread is 2 × ($5.04 − $5.00) = $0.08, realized spread is 2 × ($5.04 − $5.12) = −$0.16, and price impact is 2 × ($5.12 − $5.00) = $0.24; the identity is $0.08 = −$0.16 + $0.24. With multiplier 100, those full-spread-equivalent amounts are $8, −$16, and $24. The later midpoint movement is benchmark-dependent evidence of selection or information, not proof of permanent impact, dealer profit or causation.
  • Executable complex-order price versus leg midpoints. For a one-by-one call vertical, the bought leg is $1.40 bid / $1.50 ask and the sold leg is $0.55 bid / $0.65 ask. The single-leg natural debit is $1.50 − $0.55 = $0.95, while the algebraic leg-midpoint debit is $1.45 − $0.60 = $0.85. A complex-order fill at a $0.88 net debit improves on natural by ($0.95 − $0.88) × 100 = $7 but costs ($0.88 − $0.85) × 100 = $3 relative to the non-guaranteed midpoint. If the trader instead buys at $1.50 and the second leg’s bid falls to $0.45, the completed debit is $1.50 − $0.45 = $1.05, or ($1.05 − $0.88) × 100 = $17 worse than the package fill. Per-leg prices allocated to a package execution are not necessarily independent executable markets.

Risks and validation controls

  • Match the exact option series; nearby strikes, expirations, call/put types or adjusted roots are different markets.
  • Verify quote multiplier, exercise multiplier, deliverable, contract count, premium scale and currency before converting to cash.
  • Label consolidated NBBO, venue BBO, direct feed, vendor feed and broker display rather than treating them as interchangeable.
  • Preserve timestamps, timezone, clock synchronization, feed latency, display refresh and order-arrival time.
  • Record regular, extended, opening, reopening, halt, closing and post-event states separately.
  • Exclude or explicitly handle locked, crossed, one-sided, zero-midpoint, invalid and stale quotes.
  • Carry displayed quantity and deeper levels; a best price without capacity cannot price a larger order.
  • Distinguish displayed, hidden, reserve, auction, floor and complex-order interest.
  • Model quote cancellation, routing latency, venue rejection, rerouting and price changes before execution.
  • Preserve marketability, limit price, time-in-force, directed status and special handling; a limit controls price, not execution.
  • Calculate fill-level and order-level VWAP, partial-fill rate, residual quantity and opportunity cost.
  • Validate trade-side classification; a wrong d reverses effective spread, realized spread and impact signs.
  • Use synchronized arrival, execution and post-trade benchmarks and disclose the chosen post-trade horizon.
  • Keep full-spread-equivalent measures distinct from one-way midpoint cost, quoted width and dollar P&L.
  • Do not infer permanent market impact, adverse-selection causality or dealer profit from one later midpoint.
  • Add commissions, exchange and regulatory fees, price impact, financing, borrow, exercise and assignment costs where relevant.
  • Reconcile midpoint, last, model, broker, settlement, accounting and executable liquidation marks.
  • Recheck corporate actions, nonstandard deliverables, expiration, stale markets and erroneous or corrected trades.
  • For complex orders, preserve leg ratios, signed net debit or credit, package quantity, auction state, legging exposure and allocation.
  • Assess best execution, routing conflicts, payment arrangements and regular review across comparable orders; NBBO alone is not a safe harbor.

Common misconceptions

  • “The spread is a fixed fee charged separately.” It is a quoted price interval; actual cost depends on side, fill, size, timing, market movement and explicit fees.
  • “The midpoint is the true value and can always be traded.” It is an arithmetic benchmark with no guaranteed opposing quantity.
  • “NBBO shows all liquidity and proves best execution.” It shows eligible displayed best prices and size, not full depth or the complete order-handling outcome.
  • “Quoted, effective, and realized spread are the same.” They use different observations; realized spread also depends on trade sign, horizon and later midpoint.
  • “Adding leg midpoints gives an executable strategy price.” A package needs opposing interest at the correct net price and ratios; separate execution introduces legging risk.

Authoritative sources

Navigation

Search the wiki...