Courses / Commodity & FX Hedging

Execution and market microstructure

The order book, the order types, and the chain of parties between the desk's click and a fill at the exchange.

Module 5 of 143 lessonsAbout 40 minQuiz: 6 questions
Hedge

Lesson 1 · free to read

The order book and order types

Modern futures prices form in a central limit order book (CLOB): an electronic ledger of every resting buy order (bids) and sell order (offers/asks), ranked by price, then by time of arrival at each price. The highest bid and lowest offer are the top of book; the gap between them is the bid–ask spread; the quantity resting at each level is depth. A trade happens when an incoming order crosses the spread and matches a resting one — the exchange's matching algorithm decides exactly who.

Do not assume that algorithm is price-time. Pure price-time (FIFO) — first in at a price, first filled — is the intuitive model and is what many markets use, but it is not universal. CBOT grains run a blended algorithm: a share of each incoming order is allocated FIFO and the remainder pro-rata, in proportion to the size each resting order is showing, with an extra allocation to whoever set the price first. Two consequences matter to a hedger. First, under pro-rata your fill share depends on displayed size, not arrival time, so showing a larger order gets you more of each trade — which is why some participants over-show and why partial fills are the norm rather than the exception. Second, you must look up the algorithm per product rather than assuming; it is published by the exchange and it changes.

An amendment is a queue position, not a row update. This is the single most misunderstood mechanic in order handling. Engineers picture an order as a record you can edit; the exchange treats it as a ticket in a queue. Amend some attributes and you keep your place; amend others and you go to the back. On CME exactly six changes lose priority — price, a quantity increase, order type, stop trigger price, account, and minimum fill quantity — and everything else, critically a quantity decrease, keeps it. ICE's rule is equivalent. The practical rule for a hedging system: if the intent is trade less, shrink the resting order rather than cancelling and replacing it, because the shrink keeps your queue position and the cancel/replace throws it away.

Cancels are not free. Exchanges score messaging: CME weights a cancel far more heavily than a new order, and a system that re-quotes too eagerly runs up a real monthly bill and, more seriously, attracts supervisory attention as an inadequately controlled algorithm. "Just cancel and resend whenever anything changes" is a design smell with a price tag.

The order types a hedging desk actually uses:

  • Market order — "fill me now at whatever the book offers." Certainty of execution, zero certainty of price: it takes the best level, then the next, walking the book. In a liquid front month (the front month = the nearest delivery month, where trading concentrates; later ones are "back months"), 20 lots at market fills at the touch (the touch = the current best bid and offer); the same order in an illiquid back month can move the price several ticks against you — that cost is slippage.
  • Limit order — "fill me at this price or better." Certainty of price, none of execution: the market can trade near your limit all day and never reach it, leaving the exposure unhedged while you wait. For a hedger this is the key trade-off — a limit order is a price opinion, and price opinions are what the hedge was supposed to remove. Desk practice: limits at or near the touch to control slippage, not deep in the book to chase a better level.
  • Stop order — dormant until a trigger price trades, then becomes a market (or limit) order. Used to cap the damage of an unattended position; can fill badly in fast markets (a stop-limit can also not fill at all — choose which failure you prefer).
  • Time-in-force: day orders die at the close; GTC (good-till-cancelled) rest until filled or cancelled — and a forgotten GTC that fills weeks later against a hedge you already lifted is a classic operational own-goal. Systems should surface every working order next to the exposure it belongs to.

Liquidity has a clock. Volume concentrates in the front months and in particular hours — the exchange's main session, the settlement window, the overlap of regional trading days. The same 50-lot order costs materially more slippage at 3 a.m. than during the liquid window. Hedgers who can choose their moment execute when the book is thick; hedges triggered automatically (an offer accepted overnight) accept the book as it is — a real cost that belongs in the auto-hedge design conversation, not swept under it.

Reading a fill honestly. Your execution quality is measured against the market at the moment you arrived — arrival price vs achieved price, including spread crossed and book walked. Per lot it is small; across a year of hedging flow it is a real budget line, and it is the number that decides whether an execution setup is good.

The rest of this module

  1. From the desk to the exchange — brokers, FCMs, gateways, clearing
    Who the parties are, what each one does, and where your orders and your money actually sit.
  2. Fills, statements and reconciliation — and why an acknowledgement is not a fill
    Partial fills, average pricing, order states — and proving your book against the broker's every day.

Carry on with the course

The other lessons and the quiz are in the course itself, where your progress is saved and counts towards the certificate: Commodity & FX Hedging Operations.

Next module: Margin, brokers and funding the hedge