Skip to content
Oxarus
QUANT · RESEARCH

we
measure
first.

We test widely held but rarely measured market ideas against data: seven years of minute prices, a matched control group for every idea and a repeat on a second pair. We publish the ideas that don't hold up, too.

Institutional researchFindings
  • Minute data, 2020–2026
  • A matched control group in every test
  • Replicated on EURUSD and GBPUSD
level_study.ipynbRepresentative chart
EURUSD · M1 · 2020–2026 · 40,658 touches

Previous day high (PDH)band = 0.05 × ATR

PDHmatched control

  1. [1]Count touches

    › touches = scan(levels, band=0.05 * atr)

    Three touches on one level

    • #14 minREJECT
    • #29 minREJECT
    • #331 minBREAK

    The touch that stayed longest in the band broke the level. One representative day; the rates are in the third cell.

  2. [2]Build control lines
  3. [3]Compare rates

Representative chart; rates are from the published study.

Representative chart. Findings are for research purposes only and are not investment advice.
LevelControlResult
STUDIES

two
published
studies.

Each study states its sample, period and control method. Pick a finding to bring its panel forward in the figure.

EURUSD and GBPUSD · 2020–2026Published

At pivots, previous-day levels, round numbers, swing and Fibonacci levels, we checked whether price broke through, and compared the results with meaningless control lines that share the same distance distribution.

STUDY FACTS

Sample
83,527 touches
Period
2020–2026
Pairs
EURUSD · GBPUSD
Resolution
M1 (minute bars)
Control
Distance-matched lines
Replication
Replicated on GBPUSD
Go to the dwell-time chart

FINDINGS

  1. aEvery level family broke about 41% of the time; matched control lines, 40%. The same held in each of the seven years.

    aBreak rate by level family

    • PIVOT41% · 40%
    • PDHLC41% · 40%
    • ROUND41% · 40%
    • SWING41% · 40%
    • FIB39% · 39%
    levelmatched control2020–2026: same every year
    • 20
    • 21
    • 22
    • 23
    • 24
    • 25
    • 26

  2. bWhich touch it was made no difference: from the 1st touch to the 5th, 41%, 40%, 39%, 42%, 44%.

    bBreak rate by touch number

    41%40%39%42%44%
    #1#2#3#4#5

    matched control 40%touch number

  3. cThe only thing that changed the outcome was how long price stayed near the level on that touch. The same result replicated point for point on GBPUSD.

    cBreak probability by time in the band

From the published study. For research purposes only, not investment advice.

TIME AT THE LEVEL

how
long
at
the
level?

The longer price stays in a level's band, the more likely it is to break the level, until the effect levels off. Pick a duration; the chart and the ranges update together.

EURUSD · GBPUSD · M1 · 2020–2026 · n=83,527

Time already in the bandP(break | dwell ≥ 30 min)

55%break probability
base
41%
vs base
+14 pts
GBPUSD
54%
FIG 4By session: base → 30 minutes30 min
  • ASIA43→55% +12
  • LDN40→56% +16
  • NY40→55% +15

35%45%60%

FIG 3break probabilityEURUSD n=40,658 · GBPUSD n=42,869
plateau 56–57%35%40%45%50%55%60%base 41% · control 40%≥0 min≥10 min≥30 min≥60 min≥30 minEURUSD 55% +14GBPUSD 54% +14

EURUSDGBPUSDbase

Break probability by time spent in the level band. A research finding, not a trade recommendation.
FIG 5Time in the bandmedian

Rejections9–10 min

Breaks28–31 min

0 min10 min20 min30 min40 min

FIG 6Robustness to band width30 min
  • 0.03 × ATR+12…14
  • 0.05 × ATR+12…14
  • 0.08 × ATR+12…14

0+10+20pts

With the band varied between 0.03 and 0.08 × ATR, the 30-minute lift stayed between +12 and +14 points.

WEEKEND GAPS

when
do
they
close?

Across 345 weekend and holiday breaks, how long the opening gap took to get back to Friday's close. Pick a horizon.

n
345
median fill
11 min
60 days
98%
FIG 7Cumulative share of gaps closedEURUSD n=345
Horizon
0%25%50%75%100%opensession1 day5 days60 daysunfilled 2%10%85%65%95%98%

All gapsOver 20 pips

All gaps
95%
Over 20 pips
85%
median fill
11 min
EURUSD, 345 weekend and holiday breaks. A research finding, not investment advice.
FIG 8By gap size

Closed in the same session

All gaps65%
Over 20 pips10%

Within 5 trading days

All gaps95%
Over 20 pips85%
FIG 9Excursion before closing (≥5 pips)pips · log scale
p50 21p90 120max 470

110100470 pips

Before the gap closed, price moved the other way by a median of 21 pips, 120 pips in the worst decile, and 470 pips at most.

The remaining 2%: gaps still open after sixty days. Every one of them coincides with a major geopolitical shock.

METHOD

how
we
avoid
fooling
ourselves.

We logged every interim result that turned out wrong. Switch any entry between how it looked at first and what it actually was.

Correction log4 / 4 resolved
  1. M-01CONTROL BIASresolved

    matched lines / levelsdistance to price →

    schematic

    AT FIRST GLANCE

    Random lines broke less often than levels, so levels looked like 'magnets.'

    ACTUALLY

    The random lines sat farther from price. Once the control lines were built with the same distance distribution, the difference vanished.

  2. M-02MULTIPLE COMPARISONSresolved
    41%40%39%42%44%
    #1#2#3#4#5

    7 years of data

    schematic

    AT FIRST GLANCE

    Three months of data showed a clear dip at the third touch.

    ACTUALLY

    Across seven years of data, it evaporated. Run enough comparisons, and some differences show up by chance.

  3. M-03SIGN BUGresolved
    • bug: sign(dist)100%
    • fixed41%
    • matched control40%

    schematic

    AT FIRST GLANCE

    One measurement came out with a 100% break rate.

    ACTUALLY

    When a rate sticks to 0 or 1, suspect the measuring machinery first. It was a single sign error.

  4. M-04REPLICATIONresolved

    schematic

    AT FIRST GLANCE

    An idea seemed to hold on one pair.

    ACTUALLY

    We rebuilt the same test from scratch on another pair. Nothing that failed to hold on both counted as a finding.

RESEARCH

let
data
answer
your
question.

If you'd like us to run measurements or hypothesis tests on market data for your organization, or build its research infrastructure, get in touch.

info@oxarus.com