How to measure news slippage honestly
Talking about slippage around news is easy. Measuring it takes some care.
A rough method that works: for every order placed within five minutes of a major release, log the requested price, the filled price, the time of the order, the time of the fill, the spread just before, and the type of order. Include rejected and cancelled orders too, since leaving them out hides the worst cases.
After enough events, group the data by release type, pair, and order type. Typical findings are that market orders suffer much more than limit orders, stops are the worst, and some releases hurt far more than others.
Comparing the numbers with the broker's claims is revealing. A broker that advertises fast execution may look different during news, and the difference can be the reason to stay or switch.
The process is tedious but valuable. Many scalpers drop news trading entirely after seeing the numbers, and that is a legitimate and profitable decision.
Some traders also compare the slippage against a time-matched benchmark from quiet periods, which shows how much of the cost is specifically news-related versus general execution quality. The difference is often a sharper argument than the raw numbers.
What does the real cost of news trading look like when slippage and rejections are measured honestly?
How to measure news slippage honestly
How to measure news slippage honestly
It’s Fairman 
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: How to measure news slippage honestly
Once I forced myself to log the rejects as well as the fills, news trading looked a lot more expensive than the screenshots suggested. Over four months I recorded forty-seven orders placed within five minutes of CPI, payrolls, or the ECB decision. Market orders came back about one point nine pips worse than requested, with six outright rejections. Stops averaged two point seven pips, and one payrolls stop filled more than six pips through. Limits only slipped about three tenths, but two fifths of them never filled at all.
I then compared those fills with orders of the same size in quiet hours. The quiet sample averaged twenty-two hundredths of a pip, so most of that pain was the news itself, not my usual execution.
On a four-pip news scalp a nearly two-pip market-order hit is about half the target, which fails any sensible cost rule. I stopped sending market orders in the release minute and I now sit out the first three minutes more often than I trade them.
I then compared those fills with orders of the same size in quiet hours. The quiet sample averaged twenty-two hundredths of a pip, so most of that pain was the news itself, not my usual execution.
On a four-pip news scalp a nearly two-pip market-order hit is about half the target, which fails any sensible cost rule. I stopped sending market orders in the release minute and I now sit out the first three minutes more often than I trade them.