Dollar rebounds as US payrolls beat forecasts: what a data surprise is

The US dollar rebounded on 6 November as US jobs data came in far stronger than expected. According to investingLive, US non-farm payrolls rose by 165,000 in October against a forecast of 65,000, while Canada lost 4,700 jobs when economists expected a gain.

Since our overnight lesson on the yen and the tech-stock sell-off, the focus has shifted from stocks to jobs. The ECB reference rates for 6 November show a firmer dollar. The USD/CAD fixing rose 0.72% to 1.4066, from 1.3966 on 5 November. AUD/USD fell 1.85% to 0.6871 and EUR/USD fell 0.53% to 1.1165. Investing.com reports that the tech rout and strong US jobs data lifted safe-haven demand for the dollar. For a wider market wrap, see fxcash's report on the US payrolls beat and Canada's jobs miss.

What is a data surprise?

Before a major economic report, economists publish their forecasts. The typical forecast is called the consensus (often shown as the "expected" figure). Traders set prices partly around that number before the release.

A data surprise is the gap between the actual number and the consensus:

• Surprise = actual minus consensus

Markets tend to react to the surprise, not to the number on its own. A strong figure that everyone expected may barely move a currency. A modest figure that nobody expected can move it a lot.

The spread of forecasts matters too. investingLive explains that when data deviates from expectations it creates a surprise effect, and that how forecasts cluster inside the range also shapes the reaction.

A worked example

US payrolls. The consensus was +65K, within a range of +50K to +125K, according to investingLive's preview.

• Surprise: 165K minus 65K = +100K
• Distance above the highest forecast: 165K minus 125K = 40K

Other parts of the report also beat expectations. Unemployment was 4.1% against 4.2% expected, 0.1 point lower. Average hourly earnings rose 3.1% from a year earlier against 2.9% expected, 0.2 point higher. The prior month was also revised up, from +29K to +60K.

Canada jobs. The consensus was +13.6K.

• Surprise: minus 4.7K minus 13.6K = minus 18.3K
• Unemployment: 6.7% against 6.5% expected, 0.2 point worse

The currency move. USD/CAD puts both surprises in one price, because a strong US number and a weak Canadian number point the same way.

• Change: 1.4066 minus 1.3966 = 0.0100, or 100 pips (a pip is 0.0001 here)
• Percentage: 0.0100 divided by 1.3966, times 100 = about 0.72%

One caution: the ECB fixing is a single official rate per day, not a live quote. From the fixings alone we cannot tell how much of this 100-pip move happened after the jobs releases.

A smaller surprise. investingLive reports US consumer sentiment at 51.0 against 48.0 expected, a +3.0 surprise. The same source notes that this survey has lost some of its weight in markets, because its readings often do not match actual spending.

Why it matters for new traders

1. The gap drives the reaction. "165K jobs" means little without knowing that 65K was expected.
2. Size matters. A result outside the whole forecast range, as with US payrolls, is a bigger shock than one inside it.
3. Reports have many parts. Jobs, unemployment, wages and revisions can point in different directions. Today most pointed the same way.
4. Not all data is equal. Markets weight payrolls more heavily than some surveys.
5. Pairs have two sides. USD/CAD reacted to surprises from both countries on the same day.

Key terms

• Consensus forecast: the typical economist estimate published before a data release.
• Data surprise: the actual figure minus the consensus.
• Forecast range: the lowest to highest estimates submitted.
• Revision: a later correction to a previously published figure.
• Non-farm payrolls (NFP): the monthly count of US jobs added outside farming.
• Fixing: the ECB's official daily reference rate, not a live market price.