Launching soon — TradeOnyx is still in preparation; public sign-up isn't open yet.

Onyx Patterns

Friday drift

Some traders systematically lose money on Fridays while making it the rest of the week. The detector compares Friday-only performance against the Monday-Thursday baseline and surfaces the gap when it's material.

What it is

Friday drift is the pattern of materially-worse trading on Fridays compared to the rest of the week. Several converging causes typically produce it:

  • Cognitive fatigue — by Friday the trader has spent four days in front of charts; decision-quality drops measurably.
  • Weekend-squaring — closing positions early to avoid weekend gap-risk leads to suboptimal exits.
  • Lower liquidity / wider spreads in many markets late on Friday afternoons.
  • End-of-week boredom — the desire to 'wrap up the week with a green close' produces forced entries.

The detector compares Friday-only trades against the Monday-Thursday baseline ON THE SAME USER. The card emits only when Friday performance is materially worse — a Friday that simply tracked the rest of the week or outperformed isn't a behavioural problem and the card stays silent.

Weekend trades (Saturday / Sunday) are excluded from the baseline. They're a different behavioural mode (24/7 crypto, news-driven re-entries) that would muddy the Mon-Thu reference.

Formula
Approach: group all closed trades by close-day weekday. Take Friday on one side, Mon-Thu on the other (weekends excluded). Compute avg net P&L per trade and win-rate per side. The card emits when Friday is materially worse than Mon-Thu — either a sign-flip (Friday loses while Mon-Thu wins) or a substantial proportional gap.
 
TradeOnyx-internal: the materiality thresholds (sign-flip cutoff and the proportional ratio that separates 'meaningfully worse' from noise) plus the minimum sample sizes are calibrated empirically and not published.
Example

Over the analysis window: 12 Friday trades averaging −€8 per trade vs 48 Mon-Thu trades averaging +€10 per trade. Friday win-rate 0%; Mon-Thu win-rate 100%.

ResultCard emits with verdict **Severe**: classic sign-flip. Mon-Thu makes consistent money; Fridays consistently lose. The actionable read is mechanical — either halve the Friday risk or cut Friday entirely until the rate normalises.
How to read it

Severity bands on the Friday vs Mon-Thu gap: - Mild — Friday is meaningfully worse but still in the same direction (e.g. Friday +3 vs Mon-Thu +10). Note in the journal; one slow week per month is human. - Moderate — Friday is well below Mon-Thu (well under half the per-trade P&L). The pattern is established. Halve the Friday risk per trade. - Severe — sign-flip: Friday loses while Mon-Thu wins. The cleanest behavioural signature. Stop new entries on Friday until the rate normalises over a fresh month.

The Onyx-Engine assigns the gap to one of these bands; the cutoff thresholds are TradeOnyx-internal calibration.

The fix is mechanical, not motivational. Two interventions reliably help: 1. Friday risk-half — set risk-per-trade at 50% of normal on Fridays only. Lets you keep trading without the magnified loss footprint. 2. Friday cutoff time — define a clock-time after which no new entries open on Fridays. Most users find the early afternoon is the right cutoff (the post-lunch + thin-liquidity zone is when the drift typically lands).

Pick whichever fits your routine and write it into the journal as a hard rule.

Tier: Pro. Behavioural-discipline axis. Co-fires with Overtrading (TRA-220) when the user trades MORE on Fridays AND worse — both signals point to the same root state.

Where TradeOnyx uses it

How to read the card:

1. Hero (left) — avg P&L per Friday trade with severity verdict. The single number to commit to memory. 2. Two MicroStats — Avg P&L Mon-Thu (the baseline you want Friday to match) and Friday win-rate (the secondary signal — a Friday with positive avg P&L but very low win-rate is a different problem worth flagging). 3. Hint line — the two mechanical interventions (risk-half OR cutoff time).

Re-look frequency: weekly. The Friday signature stabilises after ~6-8 trading weeks. Don't chase a single bad Friday — the detector waits for the pattern.

Tier: Pro.

Related reading