forex-scalping.com

5 June 2020 payrolls: +2.509m vs minus 7.5m expected, yet EURUSD closed only 46 pips down

Friday 5 June 2020, 13:30 London time. The US jobs report for May showed nonfarm payrolls up 2.509 million, when the consensus expected a fall of about 7.5 million. The unemployment rate was 13.3%, against an expected 19.0% and 14.7% the month before. Average hourly earnings fell 1.0% on the month.

That is a ten million job miss, one of the largest surprises on record. How big a move should that make in EURUSD? On the close, the euro was reported at about 1.1290, down 0.41%. Working back, the previous close was near 1.1336, so the day's change was 46 pips. GBPUSD went the other way, up 0.59% to about 1.2668 according to a Korean wire report, and USDJPY finished near 109.59.

Two things follow. First, size of surprise and size of move aren't proportional. A normal payrolls miss of 100,000 is 1% of this one, and it doesn't produce a move 1% as big. A 46 pip close to close change is an unremarkable day for the pair. Second, EURUSD down 0.41% and GBPUSD up 0.59% imply EURGBP fell roughly 1% in one day.

For sizing, 0.5 lot with a 20 pip stop risks $100 on EURUSD. The release doesn't change that sum. What it changes is the chance that a stop gets hit and filled badly in the first minute.

The report had a footnote that made the headline harder to read. The Bureau of Labor Statistics said some workers absent from work had been misclassified as employed instead of unemployed, and that the true unemployment rate would have been about 3 percentage points higher than 13.3%, near 16.3%. So at 13:30 the headline suggested one thing, and the details a few minutes later suggested something less clear. Mixed releases are the ones where the first candle and the second candle can disagree. Waiting for the first five minutes to close costs you the first move and avoids some of the reversals.

The earnings number is another cross check. Average hourly earnings fell 1.0% in the month, because the jobs that returned were lower paid, mostly in leisure and hospitality. On a normal day, a payroll beat with weaker earnings would be a mixed signal for the dollar. That was the situation on 5 June: strong job counts, weak earnings, and risk assets that rallied. The dollar index closed up about 0.30% at 96.97, and a currency pair like AUDUSD, which responds to risk appetite, rose 0.39%. So the usual rule, strong data equals strong dollar, held only partly.

To keep a record, write down three prices for a release: the level at 13:29, the extreme in the first five minutes, and the close of the day. For 5 June 2020 on EURUSD that would be your own numbers from the chart, and then the first five minutes extreme minus the 13:29 level, in pips, is the realised surprise move. Do it for every payrolls release over a year, 12 entries, and you have a sheet that tells you how large the first move usually is. A stop sized from that figure is better than one picked from a calm hour.

Cross pairs deserve a look after a release like this. With EURUSD down 0.41% and GBPUSD up 0.59%, EURGBP dropped about 1.0%. A trader long both EURUSD and GBPUSD, thinking of them as two separate ideas, lost 46 pips on one and gained roughly 74 on the other. With 0.5 lot in each, EURUSD lost about $230 and GBPUSD gained about $370, so the pair of trades finished near plus $140 even though one of them was wrong. It's a reminder that two positions in dollar pairs can be one position in a cross, and the cross can move for different reasons.

Had something like this a few years back, on a payrolls Friday when I was on the boat with one bar of signal outside Leighton Buzzard. I was long EURUSD and long GBPUSD, and I honestly thought of them as two separate trades... one on the ECB, one on some Bank of England waffle. Teh euro leg went against me and hit the stop, but the cable leg ran alot further, so I came out a few quid up and couldnt work out why. It was only when I opened EURGBP that evening, over a pint, that it made any sense. Mind you, I definately didnt plan it that way.

G.