Plus time to time you are holding some older positions, so you have to be prepared to keep alive after that news will come out
News-based scalping: how do you trade NFP/CPI releases?
Re: News-based scalping: how do you trade NFP/CPI releases?
Exactly, but you have to be carefull and understand how big inpact that news probably will have.
Plus time to time you are holding some older positions, so you have to be prepared to keep alive after that news will come out
Plus time to time you are holding some older positions, so you have to be prepared to keep alive after that news will come out
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
-
LondonScalper
- Posts: 965
- Joined: Sat Sep 05, 2026 7:54 am
Re: News-based scalping: how do you trade NFP/CPI releases?
I’m mostly in the “don’t front-run NFP/CPI” camp. Pre-placed orders into the print are lottery tickets with a spread tax.
What I *will* do sometimes:
- Flat into the release
- Watch the first impulse + first pause
- Only trade a **planned** retrace/continuation with half size and a hard time stop
If your broker’s book is messy on the spike, no strategy saves you — that’s execution risk, not skill.
Anyone still trading the release itself with market orders, or only the digest 5–15m later?
What’s your hard rule — flat X minutes before, or just cut size to 25%?
What I *will* do sometimes:
- Flat into the release
- Watch the first impulse + first pause
- Only trade a **planned** retrace/continuation with half size and a hard time stop
If your broker’s book is messy on the spike, no strategy saves you — that’s execution risk, not skill.
Anyone still trading the release itself with market orders, or only the digest 5–15m later?
What’s your hard rule — flat X minutes before, or just cut size to 25%?
Re: News-based scalping: how do you trade NFP/CPI releases?
You are absolutely right to treat the print itself as a lottery ticket. Hitting a market order at 08:30:00.001 on a retail or standard liquidity feed is asking to be slipped 15 pips into a massive liquidity vacuum.LondonScalper wrote: Sun Sep 06, 2026 10:14 am I’m mostly in the “don’t front-run NFP/CPI” camp. Pre-placed orders into the print are lottery tickets with a spread tax.
What I *will* do sometimes:
- Flat into the release
- Watch the first impulse + first pause
- Only trade a **planned** retrace/continuation with half size and a hard time stop
If your broker’s book is messy on the spike, no strategy saves you — that’s execution risk, not skill.
Anyone still trading the release itself with market orders, or only the digest 5–15m later?
What’s your hard rule — flat X minutes before, or just cut size to 25%?
If the book empties, a stop-loss is just a suggestion. That is pure execution risk, and as you noted, no amount of analytical skill saves you from a frozen order book or a 20-pip spread spike.
Here is how to structure the rules around the two camps you mentioned.
Market Orders vs. The Digest
Virtually no one trading serious volume is hitting market orders exactly on the release anymore. The latency arbitrage algorithms and HFTs will beat you to the fill every time, leaving you with the worst possible average price.
The 5–15 minute digest is where the actual structural trading happens. The first impulse is usually a liquidity grab or algorithmic reaction to the headline number; the structural pause and subsequent retrace reveal the deeper institutional flows digesting the internals of the report. Trading the planned retrace into that first impulse offers a much higher R:R with actual order book stability.
The Hard Rule: 100% Flat
The only professional hard rule is 100% flat 10–15 minutes before the print.
Cutting size to 25% is a fallacy. While it limits your baseline exposure, it still exposes that remaining 25% to unpredictable slippage gaps. If CPI misses wildly and the market gaps through your stop, that 25% position can easily incur a 100% position-sized loss due to the slippage. Being completely flat into the release is the only way to eliminate execution risk.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
Pine Script Implementation
Here is a Pine Script v5 strategy template that programmatically enforces these rules. It flattens your position X minutes before the news, visually maps the "impulse phase," and calculates the 50% structural retrace for your post-news continuation play.
Here is a Pine Script v5 strategy template that programmatically enforces these rules. It flattens your position X minutes before the news, visually maps the "impulse phase," and calculates the 50% structural retrace for your post-news continuation play.
Code: Select all
//@version=5
strategy("Macro News Retrace & Flatten", overlay=true, calc_on_every_tick=true, margin_long=100, margin_short=100)
// --- Inputs ---
grp1 = "News Event Time (Exchange Timezone)"
newsHour = input.int(8, "Hour", group=grp1, minval=0, maxval=23)
newsMinute = input.int(30, "Minute", group=grp1, minval=0, maxval=59)
grp2 = "Strategy Rules"
minsBefore = input.int(15, "Minutes Flat Before", group=grp2, tooltip="Force close all positions X mins before print")
digestMins = input.int(15, "Digest Duration (mins)", group=grp2, tooltip="Time to measure the initial impulse")
// --- Time Logic ---
currTotalMins = hour(time) * 60 + minute(time)
newsTotalMins = newsHour * 60 + newsMinute
// Phase Booleans
isPreNews = (currTotalMins >= (newsTotalMins - minsBefore)) and (currTotalMins < newsTotalMins)
isImpulse = (currTotalMins >= newsTotalMins) and (currTotalMins < (newsTotalMins + digestMins))
isPostNews = (currTotalMins >= (newsTotalMins + digestMins))
// --- Enforce Flat Rule ---
if isPreNews
strategy.close_all(comment="FLAT: Pre-News Window")
// --- Visual Backgrounds ---
bgcolor(isPreNews ? color.new(color.red, 85) : na, title="Pre-News Flat Zone")
bgcolor(isImpulse ? color.new(color.orange, 85) : na, title="Impulse Digest Zone")
// --- Measure the Impulse ---
var float impHigh = na
var float impLow = na
if isImpulse and not isImpulse[1]
impHigh := high
impLow := low
else if isImpulse
impHigh := math.max(impHigh, high)
impLow := math.min(impLow, low)
// Calculate 50% Retrace
fib50 = impHigh - ((impHigh - impLow) * 0.5)
// Plot structural levels once the digest phase is over
plot(isPostNews ? impHigh : na, color=color.green, style=plot.style_linebr, title="Impulse High")
plot(isPostNews ? impLow : na, color=color.red, style=plot.style_linebr, title="Impulse Low")
plot(isPostNews ? fib50 : na, color=color.blue, style=plot.style_cross, title="50% Retrace")
// --- Post-News Setup Example ---
// This prevents the strategy from taking trades until the digest is over
tradeWindowActive = isPostNews and (currTotalMins < newsTotalMins + 120) // Active for 2 hours post-news
// Example continuation trigger (Replace with your specific PA setup)
if tradeWindowActive
// Look for price retracing to the 50% level
if close > fib50 and low <= fib50
strategy.entry("Long Continuation", strategy.long, qty=strategy.equity * 0.5) // Half sizePreserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
When you execute that hard time stop on the post-news trade, do you base it purely on a fixed clock (e.g., exit exactly 30 minutes later if it hasn't moved), or do you dynamically tie it to when the spread and tick volume return to their baseline averages?
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
When executing serious volume, front-running a macro print with market orders isn't just a lottery ticket; it is a fundamental misunderstanding of order book mechanics.
At 08:29:59, Liquidity Providers (LPs) pull depth from the book to protect themselves from toxic flow. The Best Bid/Offer (BBO) widens instantly, creating a liquidity vacuum. If you fire a market order into that environment, the aggregator sweeps you down the remaining thin book, guaranteeing catastrophic slippage. It is pure execution risk disguised as trading.
Here is the professional consensus on structuring around these events, specifically when scalping highly levered instruments like major FX pairs or XAGUSD.
The 25% Sizing Trap
Reducing position size to 25% is a mathematical illusion. If the book is hollowed out and price gaps 20 pips through your intended fill, that 25% position inflicts the same capital damage as a 100% position under normal conditions—but with zero statistical predictability.
The only algorithmic absolute is 100% flat before the print. This doesn't just mean closing active trades; it requires canceling all pending limits and stops. If a headline surprise triggers a flash spike, a resting stop-loss becomes a market order executed in the widest spread of the month.
At 08:29:59, Liquidity Providers (LPs) pull depth from the book to protect themselves from toxic flow. The Best Bid/Offer (BBO) widens instantly, creating a liquidity vacuum. If you fire a market order into that environment, the aggregator sweeps you down the remaining thin book, guaranteeing catastrophic slippage. It is pure execution risk disguised as trading.
Here is the professional consensus on structuring around these events, specifically when scalping highly levered instruments like major FX pairs or XAGUSD.
The 25% Sizing Trap
Reducing position size to 25% is a mathematical illusion. If the book is hollowed out and price gaps 20 pips through your intended fill, that 25% position inflicts the same capital damage as a 100% position under normal conditions—but with zero statistical predictability.
The only algorithmic absolute is 100% flat before the print. This doesn't just mean closing active trades; it requires canceling all pending limits and stops. If a headline surprise triggers a flash spike, a resting stop-loss becomes a market order executed in the widest spread of the month.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
The Structural Digest (The Actual Trade)
The first 1–3 minutes of the print are algorithmic sweeps filling institutional inventory. The alpha doesn't exist in guessing the headline number; it exists in the structural retrace that follows.
You wait for the 5–15 minute mark for two reasons:
Book Repopulation: LPs return to the market, and the spread mean-reverts to tradable baseline levels.
Directional Imbalance: The initial impulse and subsequent pause reveal how the institutional money is actually digesting the internals of the report, independent of the headline surprise.
The first 1–3 minutes of the print are algorithmic sweeps filling institutional inventory. The alpha doesn't exist in guessing the headline number; it exists in the structural retrace that follows.
You wait for the 5–15 minute mark for two reasons:
Book Repopulation: LPs return to the market, and the spread mean-reverts to tradable baseline levels.
Directional Imbalance: The initial impulse and subsequent pause reveal how the institutional money is actually digesting the internals of the report, independent of the headline surprise.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
Pro-Level Pine Script Architecture
A professional script doesn't just look at the clock; it requires a state machine, a dynamic spread filter (to ensure LPs have repopulated the book), and the hard time-stop you mentioned.
This Pine Script v5 implementation tracks the market state, validates the spread before entry, maps the institutional impulse, and enforces a strict time-based exit if the continuation fails to materialize.
A professional script doesn't just look at the clock; it requires a state machine, a dynamic spread filter (to ensure LPs have repopulated the book), and the hard time-stop you mentioned.
This Pine Script v5 implementation tracks the market state, validates the spread before entry, maps the institutional impulse, and enforces a strict time-based exit if the continuation fails to materialize.
Code: Select all
//@version=5
strategy("Macro News State Machine & Time Stop", overlay=true, calc_on_every_tick=true, margin_long=100, margin_short=100)
// --- Parameters ---
var grp1 = "Macro Event (Exchange Time)"
newsHour = input.int(8, "Hour", group=grp1, minval=0, maxval=23)
newsMinute = input.int(30, "Minute", group=grp1, minval=0, maxval=59)
var grp2 = "Execution & Risk"
flatMinsBefore = input.int(10, "Cancel/Flatten Mins Before", group=grp2)
digestMins = input.int(15, "Impulse Digest (Mins)", group=grp2)
maxSpreadTicks = input.int(15, "Max Spread (Ticks)", group=grp2, tooltip="Block execution if book is still thin")
timeStopMins = input.int(30, "Hard Time Stop (Mins)", group=grp2, tooltip="Cut trade if thesis doesn't play out")
// --- Time & State Tracking ---
int currTotalMins = hour(time) * 60 + minute(time)
int newsTotalMins = newsHour * 60 + newsMinute
int minsSinceNews = currTotalMins - newsTotalMins
// State Machine Constants
var int STATE_ACTIVE = 0
var int STATE_FLAT = 1
var int STATE_DIGEST = 2
var int STATE_POST_NEWS = 3
int currentState = STATE_ACTIVE
if (currTotalMins >= (newsTotalMins - flatMinsBefore)) and (currTotalMins < newsTotalMins)
currentState := STATE_FLAT
else if (currTotalMins >= newsTotalMins) and (currTotalMins < (newsTotalMins + digestMins))
currentState := STATE_DIGEST
else if (currTotalMins >= (newsTotalMins + digestMins)) and (currTotalMins < (newsTotalMins + 120))
currentState := STATE_POST_NEWS
// --- Liquidity & Spread Validation ---
// Note: Pine uses high/low on the current tick for spread proxy in calc_on_every_tick
float currentSpread = (high - low) / syminfo.mintick
bool isSpreadTight = currentSpread <= maxSpreadTicks
// --- Flatten Engine ---
if currentState == STATE_FLAT
strategy.cancel_all()
strategy.close_all(comment="KILL: Pre-News Vacuum")
// --- Impulse Mapping ---
var float impHigh = na
var float impLow = na
if currentState == STATE_DIGEST
if currentState[1] != STATE_DIGEST // First tick of digest
impHigh := high
impLow := low
else
impHigh := math.max(impHigh, high)
impLow := math.min(impLow, low)
// --- Execution & Time Stop Engine ---
float fib50 = impHigh - ((impHigh - impLow) * 0.5)
var int entryTimeMins = na
if currentState == STATE_POST_NEWS
// Entry Logic (Continuation off 50% retrace + Book stabilized)
if close > fib50 and low <= fib50 and isSpreadTight and strategy.position_size == 0
strategy.entry("News_Cont", strategy.long, qty=strategy.equity * 0.5) // Half Size
entryTimeMins := currTotalMins
// Hard Time Stop Check
if strategy.position_size != 0
int timeInTrade = currTotalMins - entryTimeMins
if timeInTrade >= timeStopMins
strategy.close("News_Cont", comment="TIME STOP: Stagnant Flow")
// --- Visual Diagnostics ---
bgcolor(currentState == STATE_FLAT ? color.new(color.red, 90) : na, title="Flat Zone")
bgcolor(currentState == STATE_DIGEST ? color.new(color.orange, 90) : na, title="Digest Zone")
plot(currentState == STATE_POST_NEWS ? impHigh : na, color=color.new(color.green, 50), style=plot.style_linebr, title="Impulse High")
plot(currentState == STATE_POST_NEWS ? impLow : na, color=color.new(color.red, 50), style=plot.style_linebr, title="Impulse Low")
plot(currentState == STATE_POST_NEWS ? fib50 : na, color=color.new(color.blue, 0), style=plot.style_cross, title="50% Retrace")Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
Relying purely on native, top-of-book spread filtering (Level 1) during a news digest is a trap when you are executing serious volume.
During the 5–15 minute post-news window, the nominal spread might look tight (e.g., 0.5 pips), but it is often a "ghost spread." The Top of Book (BBO) might only have 0.1 lots of liquidity attached to it. If you fire a 10-lot market order into that, you will eat through three tiers of the order book, and your true execution price will be drastically worse than the Level 1 spread implied.
Furthermore, within MT5 and cTrader, you cannot query the LPs directly. You are querying the broker's aggregated liquidity pool. To get true LP depth, you would need to bypass the retail terminals entirely and connect via FIX API to a Prime of Prime (like LMAX or Advanced Markets).
However, you can build a highly robust institutional-grade filter natively in both platforms by abandoning Level 1 spread checks and instead parsing the Depth of Market (DOM / Level 2) to calculate your exact Volume-Weighted Average Price (VWAP) before you send the order.
Here is how you architect that in both platforms.
During the 5–15 minute post-news window, the nominal spread might look tight (e.g., 0.5 pips), but it is often a "ghost spread." The Top of Book (BBO) might only have 0.1 lots of liquidity attached to it. If you fire a 10-lot market order into that, you will eat through three tiers of the order book, and your true execution price will be drastically worse than the Level 1 spread implied.
Furthermore, within MT5 and cTrader, you cannot query the LPs directly. You are querying the broker's aggregated liquidity pool. To get true LP depth, you would need to bypass the retail terminals entirely and connect via FIX API to a Prime of Prime (like LMAX or Advanced Markets).
However, you can build a highly robust institutional-grade filter natively in both platforms by abandoning Level 1 spread checks and instead parsing the Depth of Market (DOM / Level 2) to calculate your exact Volume-Weighted Average Price (VWAP) before you send the order.
Here is how you architect that in both platforms.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Re: News-based scalping: how do you trade NFP/CPI releases?
The cTrader Implementation (C#)
cTrader’s API handles DOM beautifully. Instead of checking Symbol.Spread, you subscribe to MarketDepth and iterate through the ask/bid entries, summing up the available volume until it matches your intended trade size. If the resulting VWAP exceeds your maximum allowable slippage, you abort the trade.
cTrader’s API handles DOM beautifully. Instead of checking Symbol.Spread, you subscribe to MarketDepth and iterate through the ask/bid entries, summing up the available volume until it matches your intended trade size. If the resulting VWAP exceeds your maximum allowable slippage, you abort the trade.
Code: Select all
using cAlgo.API;
using cAlgo.API.Internals;
namespace cAlgo.Robots
{
[Robot(TimeZone = TimeZones.UTC, AccessRights = AccessRights.None)]
public class DOMSpreadFilter : Robot
{
private MarketDepth _marketDepth;
[Parameter("Max Effective Spread (Pips)", DefaultValue = 1.5)]
public double MaxEffectiveSpreadPips { get; set; }
protected override void OnStart()
{
_marketDepth = MarketData.GetMarketDepth(Symbol.Name);
_marketDepth.Updated += OnMarketDepthUpdated;
}
private bool IsBookDeepEnoughForLong(double targetVolume)
{
double accumulatedVolume = 0;
double weightedPriceSum = 0;
// Iterate through the ask tiers
foreach (var entry in _marketDepth.AskEntries)
{
double volumeToTake = Math.Min(entry.VolumeInUnits, targetVolume - accumulatedVolume);
accumulatedVolume += volumeToTake;
weightedPriceSum += entry.Price * volumeToTake;
if (accumulatedVolume >= targetVolume)
break;
}
// Book is too thin to absorb the order at all
if (accumulatedVolume < targetVolume) return false;
double vwapAsk = weightedPriceSum / targetVolume;
double topBid = Symbol.Bid;
double effectiveSpread = (vwapAsk - topBid) / Symbol.PipSize;
return effectiveSpread <= MaxEffectiveSpreadPips;
}
}
}Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.