Part 4: Strategies, system and reference · Chapter 12 of 14 · 13 min

Trading around the news

At 08:30 the five-minute candle exceeds 10 $ on 37% of days, 20 $ on 12% and 40 $ on 4% (on strong days it has a median range of 42.2 $, up to 109.9 $). Two orders placed before the release (buy above, sell below the 08:00-08:25 range) give −0.03 ± 0.11 R over 197 days with no slippage and −0.29 with 5 $ of slippage; on the 5 strong days out of 15 where the first candle touches both sides, the setup is a double stop-out. The only defensible choices are staying out, waiting, or trading with small size.

Three ways to handle a news event

In front of an important release the trader has three choices:

  1. Stay out. Close or reduce positions beforehand, open no new orders, go back to trading when the market has normalised.
  2. Wait. Let the first move pass and trade afterwards, when the price has given a direction and spreads have come back.
  3. Trade the event. Place orders before the release or enter at market the moment it comes out.

The first two are the choices of most traders who last. The third is the one everyone imagines and almost nobody can do. The chapter explains why with the numbers from our data.

What happens to orders in the minutes of the release

In the minutes around an important release the market changes in three ways that do not show on the candlestick chart but show in the account:

  • The spread widens. Those who make the market (liquidity providers) raise the difference between buying and selling price, because they risk being left with a wrong position. A move that looks like "0.30 $" on the chart costs far more.
  • Liquidity withdraws. Offers near the price vanish for a few seconds. A market order is filled further away than you see.
  • Stops trigger in a chain. A 20 $ move activates many traders' stops: their execution pushes the price further.

We do not have bid and ask data, only the five-minute price, so we cannot measure the spread. We can measure how far the price moves in a five-minute candle at 08:30, which is a measure of how wide the "hole" is in which your stop can be filled. Over the 210 days:

08:30 five-minute candleMedian75th90th95thMaximum
All days (210)8.0 $13.1 $25.7 $37.4 $109.9 $
Normal days (195)7.5 $11.8 $17.9 $21.5 $38.4 $
Strong days, ≥ 5× (15)42.2 $65.8 $74.8 $109.9 $109.9 $
First Fridays of the month (9)19.7 $42.2 $109.9 $109.9 $109.9 $

The candle exceeds 10 $ on 37% of days (77), 20 $ on 12% (26) and 40 $ on 4% (8). A normal day has a median range of 7.5 $ in the data candle, a strong day 42.2 $: six times more. A 10 $ stop sits inside a single candle on more than one day in three, a 20 $ stop on one in eight.

The double order before the release

The classic news strategy is called a straddle: before the release you place two orders, a buy above the high and a sell below the low of the last range, and one of the two triggers when the price starts. You do not have to predict the direction: you follow it. It looks like the solution.

We tested the exact rule, with the chapter 11 rules:

  • The range is the five minutes between 08:00 and 08:25 New York time (median 15.4 $; 90th percentile 31.3 $).
  • Buy order on the high, sell order on the low; the first to trigger is the entry and the other level is the stop.
  • Target at 1 R; if nothing happens, exit at 09:25.
  • Cost of 0.3 $, and a slippage loss of 0, 2, 5 or 10 $ (on the entry price and on the stop): we do not have the real measure, so we show how far the strategy holds as it grows.
Set of daysNo slippage2 $5 $10 $
All days (197)−0.03 ± 0.11−0.16 ± 0.11−0.29 ± 0.11−0.46 ± 0.10
Normal days (182)−0.05 ± 0.11−0.20 ± 0.11−0.33 ± 0.11−0.49 ± 0.10
First Fridays (8)+0.51 ± 0.51+0.42 ± 0.52+0.37 ± 0.55+0.30 ± 0.59
Strong days, known only afterwards (15)+0.31 ± 0.49+0.24 ± 0.52+0.19 ± 0.56−0.07 ± 0.57

The result is more instructive than it seems. Over all days the order loses only the cost, −0.03 R, and every dollar of slippage increases the loss: with 5 $ you reach −0.29 R per trade. Over 197 days that means −5 R without slippage and −57 R with 5 $: risking 1% per trade, more than half the account. Normal days are almost all days, and on those the straddle only picks up noise.

On the days that matter (first Fridays, strong days) the result is positive, +0.51 and +0.31 R, and holds even with 5 $ of slippage. But they are 8 and 15 days: the error (± 0.51 and ± 0.49) is as large as the result. And strong days are known only afterwards: the rule cannot pick them beforehand, and the calendar predicts first Fridays well but with the exceptions seen in chapter 8.

Figure 1. The cumulative R of the orders before the release over 197 days, with no slippage (bars, −5.0 R) and with 5 $ (line, −57.4 R).Data: gold, XAUUSDT perpetual future (real candles) · daily (D1) · from 15 Dec 2025 to 5 Oct 2026.
Read the chart in words

Daily gold chart from December 2025 to October 2026 with, below, the cumulative R of the orders before the release with and without slippage.

When the data arrives

Strong days are the ones where the double order hurts in another way. On the 15 strong days the 08:30 candle touches at least one extreme of the range in all 15, and in 5 out of 15 (33%) it touches both extremes in the same candle. In that case buy order and sell order are both activated: one enters, the other is its stop, and the loss is already 1 R (plus slippage) after five minutes. On normal days it happened 3 times out of 195.

On 4 September the 08:00-08:25 range goes from 4,468.37 to 4,482.23 (13.9 $): the 08:30 candle has a high of 4,479.9 and a low of 4,370.0, that is it falls almost 100 $ below the sell order level. On the chart the sell order shows as filled at its level; in the account you are filled at the price of a market without liquidity, far lower.

Figure 2. Friday 4 September 2026 (UTC hours): the two lines are the high and low of the 08:00-08:25 New York range (13.9 $). The 08:30 five-minute candle falls almost 100 $ below the sell line.Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 4 Sept 2026 00:00 UTC to 4 Sept 2026 23:00 UTC.
Read the chart in words

H1 gold chart of 4 September 2026 with the buy and sell order lines and the moment of the 08:30 data.

WhenWhatPrice
4 Sept 2026 12:00 UTCbuy order 4482.24,482.23
4 Sept 2026 12:00 UTCsell order 4468.44,468.37
4 Sept 2026 12:00 UTC08:30 data (New York)4,482.23

Times in UTC.

Figure 3. Friday 11 September 2026 (UTC hours): the 08:30 candle (range 68.0 $) touches both the buy line and the sell line: the two orders trigger in the same candle and the order that was the entry is stopped by the other.Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 11 Sept 2026 00:00 UTC to 11 Sept 2026 23:00 UTC.
Read the chart in words

H1 gold chart of 11 September 2026 with the buy and sell order lines and the moment of the 08:30 data.

WhenWhatPrice
11 Sept 2026 12:00 UTCbuy order 4348.64,348.65
11 Sept 2026 12:00 UTCsell order 4330.04,330.01
11 Sept 2026 12:00 UTC08:30 data (New York)4,394.62

Times in UTC.

The classic "false first move", the price touching one extreme and then the other, is less frequent: after the first candle that touches a single extreme, the price also touches the other by 09:25 on 22% of days (43 out of 197), never on strong days (0 out of 15) and on one first Friday out of 8. With 15 days it is only an indication.

Why it is not worth it (for almost everyone)

To summarise the numbers of chapters 8, 9 and this one:

  • The price moves more in the minutes of the release (a candle about 6 times wider), but the direction is not predictable and the first move can be partly given back (chapter 8).
  • The real costs (spread, slippage) are almost always higher than the ones used in backtests. Here 5 $ is enough to lose 0.29 R per trade.
  • The positive result, if there is one, comes from a few days; the loss comes from many small stops on normal days. The sample of important days is small and does not let you know which side you are on.

You can use the news, as long as you know what it costs and do not mistake 8 days of profit for a method.

What to do with it

  1. By default, stay out. Ten minutes before the release close or reduce positions and cancel pending orders near the price. The cost of waiting is small, the cost of being filled 20-100 $ from your stop is large.
  2. If you hold a position, widen the stop and reduce the size. The stop must withstand the data candle: a normal day at the 95th percentile means 22 $, a first Friday can mean 110 $. If you widen the stop four times, the position shrinks four times, to lose the same in dollars.
  3. Wait until 08:45 before trading. The peak is the first candle; after a quarter of an hour the market is at 2 times normal, not 9 (chapter 8). Enter with wide stops and reduced size.
  4. Do not use market orders in the first minutes. The difference between the price you see and the one you get is the most expensive part. Better limit orders, even if you risk not being filled.
  5. Log every news trade. Note the time, expected price, price obtained and result in R. After 30 trades you know how much slippage you really pay, and that number decides whether the rule makes sense.
  6. Do not confuse a day with a method. Eight days of profit, in our data, are worth +0.51 R with an error of ± 0.51.

Procedure in five steps

  1. Check the calendar and note the time and release.
  2. Ten minutes before: close or reduce, cancel nearby pending orders, check margin.
  3. At 08:30 do not trade and do not touch stops.
  4. After 08:45 judge direction and spread; if you enter, with wide stops, reduced size and limit orders.
  5. Log time, expected price, price obtained, R.
Figure 4. Exercise: Tuesday 14 July 2026 (UTC hours) with the buy order line (high of 08:00-08:25 New York) and the sell line (low); four candles marked with the letters A-D.Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 14 Jul 2026 00:00 UTC to 14 Jul 2026 23:00 UTC.
Read the chart in words

H1 gold chart of 14 July 2026 with the buy and sell order lines and four candles marked with the letters A, B, C and D.

WhenWhatPrice
14 Jul 2026 00:00 UTCbuy order4,038.35
14 Jul 2026 00:00 UTCsell order4,026.01
14 Jul 2026 10:00 UTCPoint A4,025.74
14 Jul 2026 12:00 UTCPoint B4,106.24
14 Jul 2026 13:00 UTCPoint C4,090.00
14 Jul 2026 16:00 UTCPoint D4,069.52

Times in UTC.

Exercises

Chapter exercises

1. On the chart the lines are the high (buy order) and the low (sell order) of the 08:00-08:25 New York range on 14 July 2026. Which letter is the first candle that goes above the buy line, the one where the order would trigger?

Chart exercise. Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 14 Jul 2026 00:00 UTC to 14 Jul 2026 23:00 UTC.
Read the chart in words
WhenWhatPrice
14 Jul 2026 00:00 UTCbuy order4,038.35
14 Jul 2026 00:00 UTCsell order4,026.01
14 Jul 2026 10:00 UTCPoint A4,025.74
14 Jul 2026 12:00 UTCPoint B4,106.24
14 Jul 2026 13:00 UTCPoint C4,090.00
14 Jul 2026 16:00 UTCPoint D4,069.52

Times in UTC.

2. With buy and sell orders before the release over 197 days the mean R is −0.03 without slippage and −0.29 with 5 $. What does it mean?

3. On the 15 strong days the 08:30 candle touches both sides of the range in 5 cases. What happens to a double order?

4. On 8 first Fridays the orders before the release give +0.51 R with ± 0.51. How do you treat it?

Have a question or want to share your exercise?

Post in the community, or join the free signals room on Telegram.

Educational content, not financial advice. Trading involves risk.