Page 1 of 1
EURUSD at Friday afternoon: mid-session mean reversion
Posted: Sat Sep 19, 2026 8:55 pm
by LondonScalper
EURUSD Friday afternoon: mid-session mean reversion
Friday afternoon mean reversion on EURUSD is seductive and often expensive. Liquidity thins, spreads wake up, and “fade the move into the weekend” becomes a hobby with a stop loss. I have paid that tuition more than once.
My rules if I even allow the idea:
- Only if London already left a clear, extended move and I have a pre-written magnet — not a vibe.
- Half size versus midweek London baselines.
- Hard time-stop — Fridays do not get overnight hope dressed as patience.
- If spread exceeds filter, the idea is cancelled, not widened.
Most Fridays I simply pass. Passing is a position, and it shows up as missing damage in the monthly review.
Do you trade EURUSD mean reversion into Friday afternoon, or is it a forced flat for your desk?
If Monday’s review shows Friday mean reversion as a repeated cost sink, the playbook answer is removal for a month, not a smarter fade. Some edges are seasonal hobbies pretending to be systems.
Re: EURUSD at Friday afternoon: mid-session mean reversion
Posted: Mon Oct 12, 2026 4:13 pm
by Shadow Trader
The hard time stop is where I'd add a number, because "Fridays don't get hope" still leaves room to argue at 15:45. Mine is 15:30 London whatever the trade is doing, since that's roughly when London desks start handing over and the tape gets thinner by the minute.
I sketched a typical case in the attached illustration. London sold from 1.1690 at the open to 1.1645 by about 12:10, so the magnet was the halfway point at 1.16675, and I'd only consider a long once the 13:00 hour stopped making new lows.

- img_12560_r1.png (84.64 KiB) Viewed 687 times
Entry at 1.1650, stop 10 pips under at 1.1640, target the magnet. Price drifted up into the low 1.1660s and stalled, and the 15:30 time stop took me out at 1.1659 for 9 pips instead of 17. That partial reversion, with a fade back after the time stop, is the honest version of this trade on most Fridays.
One edge case for your rules is the first Friday of the month. With payrolls at 13:30 London, the "mid-session" window is really a news window, and I don't run the idea at all on those days.
Do you treat the 50% retrace as the magnet, or something like London VWAP? In my log VWAP got tagged more often but paid less.
Re: EURUSD at Friday afternoon: mid-session mean reversion
Posted: Tue Oct 13, 2026 8:53 am
by Fairman
LondonScalper wrote: Only if London already left a clear, extended move and I have a pre-written magnet
What usually counts as the magnet on a Friday afternoon? The obvious candidates are the London open price, the midpoint of the London range, or the day's VWAP, and they sit at quite different distances after an extended move. That choice probably decides more of the result than the entry does. Same with the time stop: by 15:00 London the main US data is usually out, and liquidity drops off quickly after the London close at 16:00, so a reversion that hasn't happened by then may not have the participants left to happen. Is your time stop tied to the clock, like flat by 16:30, or to a number of bars after entry?
Re: EURUSD at Friday afternoon: mid-session mean reversion
Posted: Thu Oct 15, 2026 8:54 pm
by Shadow Trader
Fairman, magnet choice is the whole trade on a Friday, so I write it before lunch rather than pick it after the move. On a typical extended London sell from 1.1690 to 1.1645, the open sits 45 pips away, the midpoint at 1.16675 is 22.5, and London VWAP that day sat near 1.1672, about 27 pips from the low.
The chart below is an illustration of those three magnets against the same afternoon path. VWAP got tagged more often in my log, but the midpoint paid when it hit, so I only allow the closer of midpoint or VWAP as the written target.

- img_12560_r2.png (77.89 KiB) Viewed 631 times
My time stop is clock based, flat by 15:30 London, not a bar count after entry. Bar counts stretch when the tape slows, and Fridays already slow after 15:00.
Half size stays non negotiable once the magnet is more than 25 pips from entry. Past that distance the reversion rarely finishes cleanly before the clock kills it.
The London open is almost always too far for me on a Friday fade. Do you ever allow it after a full session trend, or is that target reserved for midweek only?
Re: EURUSD at Friday afternoon: mid-session mean reversion
Posted: Mon Oct 26, 2026 1:37 pm
by Shadow Trader
At tick level, quoted prices alternate between the bid and the ask even when the underlying fair value has not changed. This bid-ask bounce creates negative serial correlation in successive price changes that has nothing to do with genuine reversal in market sentiment.
A chart built from the bid alone shows a different microstructure than one built from the ask, and a mid-price series smooths part of the effect. Statistics computed on trade-side prices, such as the first-order autocorrelation of tick changes, tend to come out negative largely because of this mechanical alternation.
The danger lies in backtesting. A system that appears to profit from tiny reversals on tick or second data may simply be harvesting a pattern that the spread would consume in live execution, since entries fill at the worse side and exits at the other.
Sensible practice includes testing on bid and ask separately, modelling the actual spread at the time of each trade, and being skeptical of any edge whose size is comparable to the typical spread. As the horizon lengthens to one minute and beyond, the bounce becomes small relative to genuine price movement, which is why statistics at different sampling frequencies can disagree so sharply.