8 to 9 November 2016: USDJPY fell 430 pips overnight, then finished the day higher

[Fairman]

8 to 9 November 2016: USDJPY fell 430 pips overnight, then finished the day higher

US election nights are rare enough that the scrapbook is short. The best documented one for currency traders is 8 to 9 November 2016.

For most of the Tuesday session the market priced a Clinton win. USDJPY rose to about 105.48 during Asia on the Wednesday. As results arrived and Trump took Florida and Ohio, the dollar reversed. Reuters had USDJPY down more than 3% at 101.89 in Tokyo, and the day's low was about 101.16. From 105.48 to 101.16 is 432 pips.

EURUSD ran the other way first: up 202 pips to 1.1231 in the Asian session. Then the move reversed in the US session, with the pair bottoming at 1.0928, about 350 pips below the Asian high. USDJPY ended the day around 105.6, above the 105.1 open.

So the market moved hard one way for hours and then hard the other way before the day was over. A trader who sold dollars in Tokyo at 102 was right for the first leg and underwater by the afternoon.

For a 1 lot short from 102.00 with a 30 pip stop: 30 pips is 30,000 yen, about $290 at that rate. The price went through 102.30 on its way to 105, a 300 pip adverse move, or about 10 times the planned loss.

The 2026 midterms are on Tuesday 3 November. Polls start closing around 18:00 Eastern, which is 23:00 London, since US clocks go back on 1 November.

[FXS HongKong Scalper]

That night shows how much a market can reverse within a single day. The Tokyo session took the brunt of the first move, and the recovery came in London and New York once the speech tone was clear. For traders in Asia, election nights in the US fall in our working day, which is both an opportunity and a risk. Sizing down heavily and avoiding tight stops seems the only sensible approach when outcomes are binary.

[Fairman]

There's no scheduled number on election night, so the usual tools fail. No consensus, no release time, nothing to wait for. Results arrive state by state over hours, and each projection can move prices. Spreads usually widen at times like that and stay wide while the picture is unclear. The Mexican peso hit a record low the same night, which tells you the pressure wasn't confined to majors. For a scalper the problem is that the usual stop assumptions assume continuous prices, and in a night like this prices can jump from one level to another with little in between.

[Fairman]

The two sided nature of the move is the lesson. The first leg sold the dollar for 430 pips on USDJPY, and the second took back almost all of it, with the day finishing at 105.64 against an open of 105.11. Whoever entered mid-way into either leg faced a stop at a bad time. A rule of thumb used by some traders is to wait for the first hour of reaction to complete before taking a view. It costs the first leg, but it avoids a position taken when the information is half complete and the spread is a multiple of normal.

[Fairman]

Position size and stop width interact on a night like this. Take 0.1 lot of USDJPY on a $10,000 account. A 50 pip stop risks about $48 at 105, which is half a percent. If the gap carries 300 pips before a fill, the loss is about $285, nearly 3%. A 20 pip stop on the same lot costs $19 planned, and the same gap costs the same $285. The stop width doesn't change the gap loss at all. The only things that do are the lot size and whether you're in the market. A gap day is decided by the size you carry into it.

[Fairman]

Three honest options for a night like 3 November: stay flat, trade tiny size with a stop you accept might not fill, or wait until the Wednesday London session when the major results are known. Each has a cost. Staying flat gives up the opportunity, tiny size gives up the profit, and waiting gives up the first move. Writing the choice down in advance, and the time you'll revisit it, takes the decision out of the hour when the screen is moving. The date is simple to put in the calendar: 3 November, with polls closing from 23:00 London.