Quant Memo
Core

FX Order Flow and Price Discovery

FX prices move less from news itself and more from the pattern of buy and sell orders that news produces — which is why order flow, not headlines, is what market makers actually watch tick by tick.

Prerequisites: FX Quoting Conventions, FX Liquidity Aggregation and Client Tiering

Traditional exchange-rate theory says currency prices should move on macroeconomic fundamentals — interest rate differentials, growth data, trade balances. In practice, over short and even medium horizons, a huge share of exchange-rate movement is better explained by order flow: the net signed volume of buy versus sell orders actually passing through the market, regardless of whether that flow is triggered by news or not.

Order flow is the running net difference between buyer-initiated and seller-initiated trading volume. It moves prices because a market maker absorbing net buying pressure must raise its price to attract offsetting sellers and to protect itself from clients who know something it doesn't — so persistent one-sided flow reveals information even when no headline explains it.

Why flow, not news, moves the price tick by tick

A piece of macro news only moves the exchange rate to the extent that it changes what people actually do — and different market participants interpret the same headline differently, trade different amounts, and trade at different speeds. Order flow aggregates all of that reaction into a single number a market maker can actually observe in real time: net buying or net selling pressure. A market maker facing sustained one-directional flow has to assume some of it comes from participants with better information or faster processing of the news, and defensively moves its quote in the direction of the pressure — which is the mechanism that actually moves the tradable price, distinct from the news event itself.

cumulative signed order flow (buys minus sells) price rises with the imbalance
The price line tracks the running order-flow imbalance more closely, tick by tick, than it tracks any single news release.

Worked example

A macro data release comes in stronger than expected for a currency, but the release itself doesn't move the price by contract or formula — it moves the price through the orders it generates. Suppose in the 15 minutes after the release, a market maker observes $800 million of client buy orders against $500 million of sell orders in that currency pair.

  1. Net order flow. 800 - 500 = \300$ million of net buying pressure.
  2. Facing this imbalance, the market maker must find offsetting sellers somewhere — either other clients or by trading in the interbank market — and the fastest way to attract sellers is to raise its offer price, nudging the whole quoted market higher.
  3. If a second market maker sees comparatively balanced flow (say $650 million bought against $630 million sold) after the same release, that maker's price barely moves — even though both makers received the identical piece of news. The price impact tracked the flow imbalance, not the headline.

What this means in practice

This is why professional FX market makers watch order-flow indicators and their own book's imbalance as closely as they watch the economic calendar, and why some of the largest FX dealing banks sell proprietary order-flow data and analytics to clients specifically because that data has predictive power over near-term price direction that public news alone does not. It's also why two market makers can legitimately quote slightly different prices for the same currency pair at the same instant — each is reacting to the flow that happens to be hitting its own book, not to some single shared "true" price.

Don't assume a currency's price move after a news release is proportional to how "big" the surprise was. A modest surprise that triggers lopsided, persistent order flow can move the price further than a large surprise that different participants interpret in offsetting directions — the flow imbalance, not the surprise size, is what a market maker's price actually responds to.

Related concepts

Practice in interviews

Further reading

  • Evans & Lyons, Order Flow and Exchange Rate Dynamics
  • Lyons, The Microstructure Approach to Exchange Rates
ShareTwitterLinkedIn