Round numbers — why prices are pulled like a magnet

Anyone who has watched EUR/USD for a week has seen the same scene. Price drifts toward 1.1000, slows, tries to break through, retreats, comes back and tries again. Sometimes for hours, sometimes for weeks. The number with three zeros has no fundamental power — arithmetically it is exactly the same as 1.0997 or 1.1003. And yet the market does not treat it that way. This article explains why, how the magnet behaves on the three biggest pairs and how to use it in live trading without falling into the traps that institutions set on retail orders.

Where the pull of round numbers comes from

The mechanism is psychological. A trader who decides "I'll buy euro when it drops to 1.10" will, in the vast majority of cases, place a limit order at exactly 1.10000, not at 1.10027. A bank client phoning the dealer with "give me a take profit at 1.32 when we get there" does not specify whether they mean 1.32000 or 1.32048. The same effect explains 9.99 in a shop rather than 10.00 — the difference is one cent, but the brain treats the two numbers as qualitatively different.

The second factor is institutions. FX option strikes are set at round levels; bank algorithms are calibrated to react there; analyst reports round their tables to two or three zeros. When large barrier options expire on a Friday at 1.1000, dealers really do defend that number to the close. This is not mysticism — it is accounting. The third element is the self-fulfilling prophecy of collective attention at known support and resistance: because thousands of traders know that 1.10 matters, all of them watch what price does there. That activity alone, regardless of any macro rationale, produces extra volatility exactly where we see it.

Order clustering: evidence from FX research

Carol Osler, an economist working for the Federal Reserve Bank of New York, published a paper in 2003 that has become a classic of the FX order-flow literature. She analysed millions of orders collected by a major investment bank and showed something far from accidental: stop-loss orders cluster at prices ending in 00, while take-profit orders cluster at prices ending in 00 and 50. The concentration was large enough to predict short-term price moves with accuracy meaningfully above random.

EUR/USD and GBP/USD: reading the magnets on majors

For most of the 2022–2024 cycle EUR/USD stalled repeatedly at 1.1000. Every test was an opportunity for range traders: buying near 1.0820, selling near 1.1080, with 1.1000 treated as the axis. It worked well during consolidations and badly when a strong directional trend was running. The conclusion is not that the level lost its pull; it is that in a trend a magnet becomes a waypoint rather than a wall. More on the pair is in our profile of EUR/USD characteristics.

On GBP/USD an extra effect shows up: quarter-figures act disproportionately strongly. The 1.2500 level matters more in practice than 1.2400, even though both have two zeros. The reason is the same as the 9.99 effect in a shop: the brain likes natural divisions, and a quarter of the way between 1.20 and 1.30 is intuitive. The same pattern appears at 1.5000 and 1.7500. The simpler and rounder the value, the more eyes are on it.

Stop hunts: how the market harvests liquidity below the round level

A stop hunt is one of the most frequently observed patterns in the currency market and one of the most expensive for an inexperienced trader. The mechanic is simple: an institution knows that a cluster of stop-loss orders sits below a round level (say 1.1000), left by long positions opened around 1.1020 to 1.1050. To offload a larger position, the institution pushes price a few pips below 1.1000, to 1.0993 or 1.0987. Stops trigger in a cascade and the buying side scoops up cheap euro from the traders just stopped out. Ten or fifteen minutes later price drifts back near 1.1030. For someone whose stop sat at 1.0997 it looks like malice; for someone who understands the pattern, it is a predictable feature of order flow trading around round numbers.

The operational takeaway is unambiguous: never place a stop loss exactly at a round level, nor two or three pips beyond it. That is an invitation to be stopped out on the next liquidity sweep. A safe distance depends on the pair's average daily range, from the teens of pips on EUR/USD to twenty-something on USD/JPY. The mechanics of the hunt itself are worth knowing in detail before you trade near magnets, and they are covered in our note on how stop hunts work from the broker side.

"The most obvious psychological price levels are those that occur in round numbers. These levels matter to the foreign exchange market because both retail traders and institutions tend to place their orders around them." — John J. Murphy, Technical Analysis of the Financial Markets, New York Institute of Finance, 1999.

Confluence: why a round level alone is not enough

An uncomfortable point has to be made. The win rate of the naive setup "price touched 1.1000, buy" is only marginally better than random — not enough once spread, commission and overnight financing are paid. Round levels are not a standalone signal — they are a context filter. Their real strength shows up only in confluence with other technical clues: a Fibonacci retracement level, a long-period moving average, a clear candlestick pattern or an RSI divergence. When a round number and the 61.8% retracement point at the same price, the probability of a reaction rises noticeably. Add a hammer or an engulfing candle on top and you have a hypothesis worth working with.

That is also why a trading plan should never rest on "because it's a round number" alone. The magnets are real but indiscriminate — they work both ways and give no warning when their pull is about to run out. A round level on its own tells you only "something can happen here". What actually happens is decided by the other signals on the chart.

What to do tomorrow

1. Draw horizontal lines at every full figure and every half-figure on the daily chart of one pair. For EUR/USD use 1.0500, 1.1000, 1.1500 and so on; for USD/JPY 145, 150, 155. Count how many times each was tested over the last two years and compare with random levels in between — the difference builds intuition for where reactions cluster.
2. Write down your own rule for stop-loss distance from a round number. Start with something simple: at least one fifth of the pair's daily range beyond the round level, never closer. Writing the rule down blocks the temptation to park the stop five pips under 1.10 because statistics say it bounces — exactly the behaviour stop hunts feed on.
3. Require two independent confirmations before you trade a reaction at a round level. A touch of the magnet on its own is not enough. Only the confluence with a Fibonacci retracement, a moving average or a readable candle gives a setup that holds up on the maths of the trade. Without that discipline, round levels turn into a pretext for impulsive entries.
4. Practise the whole approach on a demo account for at least twenty trades. Pick one pair, restrict yourself to a single setup type — a round number in confluence with a second signal — and log every decision with its result. Only repeatability on demo earns the right to move to live capital with one percent risk per trade.

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