Reading the data

News Tone: Aggregated Sentiment Per Currency

How positive or negative the news flow has been for each currency, z-scored against the last 30 days. Tone-vs-price divergence often precedes mean reversion.

N Written by Nick, founder of Markets Mastered · Trading professionally since 1989

Last updated 23 May 2026

Most platforms surface news as headlines. We surface it as a number: how positive or negative coverage has been for each currency relative to the last 30 days. The meter sits alongside currency strength because the two together tell a fuller story than either alone.

The simple version

Every day, the platform pulls news article tone from GDELT, an open dataset that aggregates sentiment across thousands of sources globally. We filter to economic themes (interest rates, central banks, inflation) per country so political noise does not dominate, then compute a daily tone score per currency.

The bar in the dashboard shows the z-score of today's tone against the trailing 30-day baseline for that currency:

  • z-score 0: today's tone is around the 30-day average.
  • z-score +2: today's coverage is unusually positive (top ~5% of the last month).
  • z-score -2: today's coverage is unusually negative.

Bars colour amber for normal, emerald for extreme positive (z >= 1.5), rose for extreme negative (z <= -1.5).

Why news tone matters as a third dimension

Currency strength tells you what the market is doing. CoT tells you who is positioned. News tone tells you what the market is hearing - which often leads price by hours or days.

The most useful read is divergence. When tone is sharply negative for a currency but price is holding up, you have a setup where the market has not yet priced the bad news. The opposite - tone positive, price weak - often precedes mean reversion back toward the news read.

GDELT is what state-of-the-art quant funds have been mining for over a decade. Bringing it into a retail dashboard is the biggest single piece of "intelligence retail does not have" we offer. The catch is that it is noisy at the daily level - which is why we z-score against a 30-day baseline. Absolute tone numbers are nearly meaningless; the deviation from recent normal is the signal.

"Price is the consensus print. News tone is what the consensus is being shaped by. When the two diverge, you have your edge."

— Nick, founder of Markets Mastered

How to use it in practice

As a confirmation when stacked with positioning and strength

The cleanest setups have all three layers point the same way. USD news tone running +1.8 z-score positive, USD currency strength at the top of the meter, USD CoT below the 30th percentile (specs not yet long): that is a setup where the market hasn't caught up to the news yet, positioning is light, and price is starting to move. Stack the trade with confidence.

As a divergence flag

The opposite reads are warning flags. Strong EUR price action plus EUR news tone at -1.5 z-score: the rally is happening into bad news, which historically does not last. Either the market knows something the news flow doesn't, or the rally is about to reverse. Trade with smaller size or wait for the convergence to resolve.

As a sanity check on briefings

When the morning briefing leads with "USD weakness driven by dovish Fed expectations", check whether news tone for USD is actually negative. If tone is neutral or positive, the briefing's thesis is on thinner ice than it reads.

In quiet markets

When price action is choppy and the squeeze meter is high, news tone often resolves the ambiguity first. A sudden swing in tone for a major currency tells you which direction the eventual breakout will favour.

What the meter does not do

It does not tell you what news drove the score. We aggregate sentiment across thousands of sources; the underlying articles are not surfaced here. For specific catalysts, use the economic calendar and the briefings.

It also doesn't predict where price goes tomorrow. Sentiment is one input. Markets price news at their own speed. The meter is most useful in combination with the other intelligence layers.

A note on the data

GDELT publishes daily aggregates with a one-day lag (today's tone for yesterday's news flow). The platform refreshes daily at 23:00 UK. Per-currency queries are filtered to source country plus economic themes - we use sourcecountry:US for USD, German + French sources for EUR, etc. The 30-day z-score window stabilises within a week of fresh ingestion.

Eight currencies are covered today: USD, EUR, GBP, JPY, CHF, CAD, AUD, NZD. Coverage is the same set the briefing team writes about.

Keep reading

This article is general market education, not financial advice. See our risk disclaimer.

Start today

Ready to trade smarter?

Join traders who've stopped watching charts and started making better decisions.

We use cookies to analyze site traffic and improve your experience. Privacy Policy