Round price levels: order research, quote units and risk

A round price such as EUR/USD 1.1000 is an easy reference point, not a physical magnet. Price may reverse, cross it or never reach it in the observation period. To investigate its usefulness, define the quote convention, grid, contact rule and outcome horizon before looking for reactions.

Specify the grid and price unit

Roundness depends on notation and units. 1.10 and 1.10000 are the same price; extra displayed zeros do not strengthen it. In the examples here, EUR/USD uses a pip of 0.0001 and USD/JPY 0.01. A pip is not necessarily the smallest displayed or tradable increment.

Choosing 50- or 100-pip spacing defines which reference prices you will observe. It does not rank their effectiveness. A grid can be studied as support and resistance, but chart prices alone do not identify bank algorithms, an option barrier being defended or participants’ intent.

Order clustering: evidence from FX research

Carol Osler’s March 2001 NY Fed Staff Report 125 studied 9667 conditional stop-loss and take-profit orders placed with one large dealing bank from 1 September 1999 to 11 April 2000, covering dollar-yen, dollar-pound and euro-dollar. Take profits clustered at round prices, while buy stops tended to cluster just above and sell stops just below. The proposed mechanism allows both reversal near a level and continuation after crossing it. The author could not prove representativeness across all dealers. This is a historical sample, not a current global order map or a prescribed retail stop distance.

Quote inversion changes roundness

For invented prices quoted in USD per EUR, 1.10 and 1.11 invert to about 0.909091 and 0.900901 EUR per USD, a gap of 0.008190. Likewise 1.50 and 1.51 invert to 0.666667 and 0.662252, a gap of 0.004415. Equal 0.01 steps in the first convention therefore do not form equal inverse steps: 1/p − 1/(p + 0.01) = 0.01/[p(p + 0.01)]. Preserve the study’s original quote convention before comparing round-number grids. Our EUR/USD profile explains the pair context.

In the numerical interval 1.20–1.30, 1.2500 is the midpoint; its .25 fractional part is one quarter of a whole price unit. That arithmetic does not establish that 1.2500 is stronger than 1.2400 as a trading level. Order clustering near round prices is a hypothesis to examine with an identified sample; define the zone width and independent contacts before comparing its behavior.

A chart touch and an executable order differ

With hypothetical Bid 1.1000 and Ask 1.1002, a Bid chart reaches the round number, but a purchase is not available at Bid 1.1000. Typical MT5 forex buys use Ask and sells Bid; buy-limit conditions concern Ask, long SL checks Bid and short SL checks Ask, with different rules possible for exchange symbols. A move below a line and back does not identify an intentional stop hunt or a bank buyer. Keep quote-side evidence separate from an order-flow interpretation.

No pip distance or ATR multiple guarantees avoiding a stop. A wider stop at unchanged volume raises planned loss; an ordinary stop can fill at a worse price. IG’s UK documentation distinguishes it from a separate guaranteed stop product, whose terms must be checked. Before inferring a cause from a stopped-out trade, inspect quotes, trigger rules and the execution report; see stop-hunt mechanics.

Cost model: assume a USD account, EUR/USD purchase at Ask 1.1002 and total opening/closing commission USD 1 per 10,000 EUR, proportional to volume. With 10,000 EUR, Bid exits at 1.0995 or 1.0985 lose USD 7 or 17 gross, hence USD 8 or 18 after commission. With 4,000 EUR and the further stop, the assumed Bid fill 1.0985 loses USD 6.80 gross and USD 7.20 net. Spread is already included through Ask entry and Bid exit; financing, taxes and conversion are omitted.

For a USD 8 planned-loss budget at the further stop, theoretical volume is 8/(0.0017 + 0.0001) ≈ 4444 EUR. With a chosen 1000-EUR volume step, round down to 4000 EUR. An adverse Bid fill at 1.0980 instead loses USD 8.80 gross and USD 9.20 net. Thus planned risk is not a guaranteed loss ceiling. Actual minimum volume and fee schedules require the provider’s specification.

Additional conditions need their own test

A Fibonacci retracement, moving average or candle can be another recorded condition. If they use the same prices, they are not automatically independent confirmations. Their overlap does not establish a win rate, increased reaction probability or net expectancy. Define each condition and compare the complete rule after costs.

Separate the question “did price reach the defined zone?” from “did an executable trade make money?”. The diagram’s overlapping bands show why contact assignment also matters: one quote can satisfy two zones without representing two independent market events.

What to do tomorrow

1. Fix units and a grid before counting. The proposed EUR/USD step from 1.0500 to 1.1000 is 500 pips, as is USD/JPY 150 to 155 under the stated pip conventions. Neither is automatically a 100-pip figure or a strength ranking.
2. Calculate risk under each proposed stop. Keep entry/exit quote sides and costs explicit, reduce volume if required by the chosen budget and assess adverse fills. No one-fifth-ATR distance is a safe universal rule.
3. Record additional conditions before outcomes. Do not call indicators independent merely because they have different names. Test a fixed context rule rather than choosing whatever fits a successful chart.
4. Use a separate evaluation period. Log the contact rule, allowed orders, nonfills, losses and costs. Twenty demo cases do not prove profitability or automatically justify moving to live capital with a fixed risk percentage.

For a broader reference on price structure, the technical analysis section on ForexMechanics.com is a sensible place to keep reading.