News Trading

News trading is the family of strategies that organise entries and exits around macroeconomic events instead of around chart structure alone. Rather than waiting for a setup to appear at an unknown time, the trader knows in advance — from an economic calendar — that at a fixed timestamp new information will reach the market: an inflation print, an employment report, a central-bank rate decision and its press conference, PMIs, or meeting minutes. The strategy's job is to define exactly what happens in the minutes or hours around that timestamp.

The mechanism being exploited is repricing under thin liquidity. Ahead of a scheduled release, market makers widen quotes and cut size because they cannot price an unknown number, so depth drops and price often compresses into a narrow pre-event range. At the release, the actual figure is compared against consensus and against existing positioning, and the gap between them is priced in seconds. That produces the two features every news strategy has to either exploit or survive: a volatility expansion far larger than the instrument's normal bar range, and an execution regime — spreads, slippage, partial fills — that behaves nothing like the rest of the session.

## Main variants

**Breakout or straddle.** Pending buy-stop and sell-stop orders are placed on either side of the pre-release range, on the assumption that the first directional impulse continues far enough to pay for the second order being stopped out.

**Fade or reversion.** The initial spike is treated as an overreaction driven by stop cascades and thin books. Entry comes against the first move once it stalls, usually with a defined exhaustion condition rather than a fixed clock.

**Post-event continuation.** The first minutes are skipped entirely. The system waits for a settling window, then trades the direction the market establishes once the full report — not just the headline number — has been digested.

**Event-structure approaches.** Some versions target the internal shape of a single recurring event, such as the interval between a policy statement and the press conference that follows it, where two separate information shocks arrive minutes apart.

**News as a filter, not a signal.** A large share of systematic traders use the calendar only to flatten or block entries around releases. This is still news trading in the sense that the calendar drives behaviour, but the edge is defended rather than sought.

## What typically differentiates implementations

Two versions can share the same one-line description and behave completely differently. The variables that matter: which events are traded and which are ignored; which instruments (index futures, FX majors, metals, crypto) and why; how the reference range is defined and how long before the release it is measured; whether orders are resting before the print or sent after it; the holding window and whether the exit is time-based, target-based, or trailing; how the stop is sized, fixed or volatility-scaled; and what guards exist for spread, slippage, and maximum adverse excursion. Whether the rules use the surprise itself — actual versus consensus, plus revisions to prior data — or only price action is often the single biggest structural difference between two implementations.

## Common mistakes

The most frequent is backtesting on OHLC bars with a fixed spread. Around releases, spread and slippage are the dominant cost term, and a model that ignores them can turn a losing rule into an apparently profitable one. Related errors: assuming a resting stop order fills at its trigger price; counting trades instead of events, which hides how small the real sample is (eight scheduled FOMC meetings a year, twelve of most monthly reports); tuning the entry delay until the equity curve looks good, which fits noise in a small sample; assuming the surprise direction and the price direction agree; using normal position sizing in an abnormal-volatility window; and mis-stamping event times across timezones and daylight-saving changes. Prop-firm and broker restrictions on trading around news are also frequently discovered after the rules are already built.

## How to evaluate and backtest a version

Start by rebuilding the event dataset: exact release timestamps, consensus, actual, and revisions, aligned to the price series. Test on tick data with recorded spread where possible, and model slippage explicitly — then stress it, because the honest question is how much slippage the rules can absorb before the edge disappears. Report the distribution of per-event outcomes, not the average; in news trading a few events usually dominate the result. Split out-of-sample chronologically and check regime dependence, since behaviour during a hiking cycle differs from a cutting or holding one. Finally, run a control: apply the same rules at random non-event timestamps. If the results are similar, the edge belongs to the time of day or the volatility filter, not to the news.

Two decoded versions are catalogued under this concept — *FOMC Trading Strategy* (ULTRA ACADEMY) and *Trading con Noticias (News Trading)* (Proyecto Algo). Each has its own page with the extracted rules and parameters.

Strategies in this concept (5)

Frequently asked questions

Is news trading the same as fundamental analysis?

No. Fundamental analysis estimates what an asset should be worth over weeks or months. News trading is a short-horizon, event-timed method: it cares about how price reacts in the minutes or hours around a scheduled release, and many versions never form a view on the underlying value at all — they only react to the volatility and liquidity conditions the release creates.

Which events are typically used?

Mostly scheduled, high-attention macro releases: central-bank rate decisions and press conferences, inflation prints, employment reports, GDP, PMIs, and meeting minutes. Some versions add instrument-specific events such as inventory reports for energy. Unscheduled news — geopolitical headlines, emergency statements — cannot be traded systematically in the same way, because there is no calendar entry to build rules around.

Should news be used as a signal or as a filter?

Both are legitimate and they answer different questions. As a signal, the release is the reason to enter. As a filter, the calendar only says when not to trade or when to flatten existing positions. Filter use is far more common among systematic traders, and it is worth testing separately: a rule set may improve simply by avoiding releases, without any event-driven entry.

Can news strategies be backtested reliably?

They can be tested, but with more care than most setups. Two things drive the result: execution modelling (real spread, realistic slippage, no assumption that stop orders fill at their trigger) and sample size, since the number of qualifying events per year is small. A backtest that reports hundreds of trades on a monthly release is usually counting something other than events.

Why do news strategies often behave worse live than in testing?

Because the cost model is wrong. In the seconds around a release, spreads widen, depth thins, and orders fill away from their trigger — exactly the conditions most historical data and default backtest settings do not reproduce. A version that only works with a tight fixed spread is not a strategy with an execution problem; it is a strategy whose edge was the unrealistic spread.

What should be compared between two versions of this concept?

Event list and instrument, timing window and how it is measured, order placement before versus after the print, exit logic, stop sizing, and any slippage or spread guard. Then compare them on the same events with the same execution assumptions. Differences in reported results are frequently differences in test conditions rather than differences in the rules themselves.

All strategy concepts · Explore all strategies