📅 Originally Published: · Last Updated:
Support and resistance often attract attention near round prices, and transaction data confirms that trading clusters there. That does not mean a level such as $100 is more likely to hold, reverse, or produce a predictable return. Treat roundness as a clue about where other traders may concentrate, then base the actual trade on valuation, liquidity, risk, and the price you are genuinely willing to accept.
In support and resistance analysis, a clean price such as $50 or $100 feels important before anyone opens a chart. Traders remember it easily, financial media repeats it, and order-entry screens make it effortless to type. That shared attention can create a visible concentration of transactions.
Round-number support and resistance can therefore look more precise than the underlying order flow. The harder question is what the concentration means. A crowded price can become a temporary liquidity point, but crowding alone does not tell you which side will win. The same level may attract buyers, sellers, market makers, stop orders, and traders trying to step ahead of everyone else.
This distinction matters because support and resistance analysis often blends two different claims. The first is descriptive: people trade disproportionately at round prices. The second is predictive: those prices reliably stop or reverse a move. The strongest available account-level evidence supports the first claim. It does not establish the second.
What 134 Million Transactions Actually Show
Bloomfield, Chin, and Craig studied 134,066,741 transactions from 20,798,516 U.S. securities accounts between July 2019 and June 2020. Their April 2025 revision found that transactions clustered at integer prices far more often than a uniform distribution of cent endings would predict.
| Price pattern | Observed share | Uniform baseline | Observed vs. baseline |
|---|---|---|---|
| Integer prices, such as $100.00 | 3.73% | 1.00% | 3.73 times |
| Whole-dollar or 50-cent prices | 5.64% | 2.00% | 2.82 times |
| Prices ending in 0 or 5 cents | 21.34% | 20.00% | 6.7% above baseline |
The 50-cent category includes both whole-dollar and half-dollar prices, so two of the 100 possible cent endings form the uniform 2% baseline. The 6.7% figure is the relative excess: 21.34% divided by 20%, minus one.
Primary source: Bloomfield, Chin, and Craig, The Allure of Round Number Prices for Individual Investors, July 2024, revised April 2025. Read the working paper.
For support and resistance, the measurement boundary is crucial. The headline 3.73-times result applies to integer-ending transaction prices across the sample. It does not mean that trades specifically at $100 were 3.73 times more common, and it does not say that a $100 support level was 3.73 times more likely to hold.
The paper also found that round-price clustering increased with the stock’s price, especially for integer and 50-cent prices. Individual investors showed more integer clustering than institutions, although institutional trades also appeared at round values more often than chance alone would predict.
What Round-Number Data Does Not Prove
Support and resistance claims require more than a count of transactions. A transaction can occur at a round price for many reasons. A buyer may lift an offer, a seller may hit a bid, two patient limit orders may meet, or a market maker may provide liquidity. The transaction record identifies the price, but it does not automatically reveal whether the level acted as support, resistance, or a profitable signal.
That creates four limits a trader should keep visible:
- No directional forecast: clustering does not tell you whether price will rise or fall after reaching the level.
- No measured hold rate: the account study did not compare how often round levels held versus broke.
- No universal execution penalty: it did not estimate a fixed basis-point cost for placing an order at an integer.
- No household dollar-loss estimate: it did not convert the pattern into a standard ten-year wealth gap for an individual investor.
Those missing links are important. A technically correct statistic can become misleading when it is spent on a different question. The 3.73-times concentration is strong evidence of attention and order clustering. It is not evidence that a round-number line has predictive power.
The same caution applies to the widely cited $813 million figure from Bhattacharya, Holden, and Jacobsen. Their study examined more than 100 million transactions in 100 randomly selected U.S. stocks from 2001 through 2006. The authors estimated aggregate wealth transfers around round-number trades, then scaled the result to the wider market.
They explicitly described the market-wide estimate as rough. Their trade-price method produced about $1.02 billion per year, while a midpoint-based method produced about $606 million. Averaging the two yielded approximately $813 million. That is a historical, model-based aggregate estimate, not a current annual bill and not a per-investor loss.
Primary source: Bhattacharya, Holden, and Jacobsen, “Penny Wise, Dollar Foolish: Buy-Sell Imbalances On and Around Round Numbers,” Management Science, 2012. Read the paper.
The practical lesson is narrower than the dramatic number suggests. Obvious price clusters can create predictable order placement. Other traders may respond by supplying liquidity near those clusters or by stepping ahead of them. Whether that helps a particular order depends on the spread, queue position, volatility, urgency, and what happens if the order never fills.
Why Trading Clusters Near Round Prices
The psychology behind support and resistance starts with a simple convenience: round prices reduce mental and operational effort. They are easier to remember, communicate, and enter. A trader deciding that a stock looks attractive “around $50” can convert that thought into a $50.00 limit order without making a separate decision about the final cents.
That convenience is one plausible mechanism, but the evidence does not support a single universal explanation. Bloomfield, Chin, and Craig found patterns consistent with facility, or ease of use, rather than a simple claim that investors lacked the ability to process precise prices. Institutions also clustered around round values, which suggests that market convention and strategic behavior matter alongside psychology.
Strategic responses can reinforce the pattern. Bhattacharya, Holden, and Jacobsen found excess buying immediately below round prices and excess selling immediately above them. Their interpretation was that some informed or liquidity-supplying traders position around the obvious clusters created by other market participants.
International evidence points in the same broad direction, but it is not universal. Chen studied 41 equity markets and reported statistically robust round-number effects in 30 of them. The study also found buy pressure just below round values and sell pressure just above them in many markets.
Supporting source: Chen, “The Influence of Round Number Bias on Trading Behavior,” Journal of Business Research, 2018. View the publisher record.
The mechanism therefore looks less like a magic line and more like a coordination point. Many traders can independently choose the same easy reference, and once the cluster is visible, other participants may trade around it. A chart records the result after those choices reach the market.
Can Non-Round Limit Orders Improve Execution?
For support and resistance traders, the honest answer is sometimes, but there is no universal “move the order by two or three cents” rule.
Bhattacharya, Holden, and Jacobsen concluded that traders demanding liquidity might consider non-round action thresholds, while limit-order traders supplying liquidity might benefit from placing orders around the predictable clusters. That finding supports testing less obvious prices. It does not guarantee that $99.97 is better than $100.00 in every stock or market condition.
A one-cent improvement is useless if the order never fills and the market moves away. An odd-cent order can also be inferior when it causes you to accept a price outside your valuation or risk plan. Price improvement, fill probability, urgency, and opportunity cost belong in the same decision.
For a purchase, start with the highest price you are willing to pay. For a sale, start with the lowest price you are willing to accept. Roundness may help you anticipate where competing orders accumulate, but it should not override the economic boundary that made the trade sensible.
This is also where backtesting matters. A visual pattern should not be trusted because it looks clean in hindsight. Before treating a level as a signal, define the entry, exit, costs, failed fills, and out-of-sample test. The process in our technical analysis backtest guide is more demanding than drawing a line after the move.
A Practical Decision Router
The right use of support and resistance evidence depends on the decision in front of you. This router keeps the descriptive finding from turning into a stronger claim than the data can carry.
| Your question | What the evidence can tell you | What to do |
|---|---|---|
| “Is $100 likely to attract activity?” | Yes. Integer prices receive disproportionate transaction attention in the cited U.S. account-level data. | Expect possible crowding, then inspect the spread, liquidity, and current order conditions. |
| “Is $100 likely to hold as support?” | The cited studies do not provide a hold rate or directional forecast. | Require separate evidence, a defined invalidation point, and a tested rule. |
| “Should my limit order avoid .00?” | Non-round thresholds may avoid the most obvious cluster, but the benefit is context-dependent. | Use an acceptable-price boundary first. Change the cents only when the revised order still fits that boundary. |
| “Should I buy because price touched a round level?” | Roundness alone does not establish value, momentum, or reversal probability. | Combine it with your investment thesis, position size, and risk plan. |
| “How should I draw the zone?” | A single precise line may hide the uncertainty around actual orders and fills. | Use a range and test the rule. Our trendline survivorship-bias guide explains why hindsight can make chart levels look cleaner than they were in real time. |
Long-term investors should keep the scale of the decision in view. For a liquid security, the final cent of a single order may matter less than fees, taxes, asset allocation, or abandoning a plan during volatility. An investment policy statement can prevent a visually appealing level from quietly replacing the rules that govern the portfolio.
Shorter-horizon traders face a different trade-off. Queue position and non-fill risk can matter more, so a detailed execution log is useful. Even then, the evidence supports a hypothesis to test, not a fixed cent offset that works everywhere.
Frequently Asked Questions
Why do round numbers attract more trading?
They are easy to remember, communicate, and enter. Shared attention can create order clusters, and other traders may then respond strategically around those clusters.
Does 3.73 times more trading mean a round level is 3.73 times more likely to hold?
No. The figure compares the observed share of integer-price transactions with a uniform 1% baseline for cent endings. It does not measure support accuracy, resistance accuracy, or post-touch returns.
Should I place a buy limit at $99.99 instead of $100.00?
There is no universal answer. A non-round order may avoid an obvious cluster, but a one-cent difference can reduce fill probability or conflict with the price you are willing to pay. Start with valuation and urgency, then consider the queue.
Is the $813 million estimate a current annual cost?
No. It is a rough historical market-wide estimate based on a 2001 to 2006 sample and two scaling methods. It should not be presented as a current annual figure or divided into a standard loss for each investor.
Are round-number support and resistance levels useless?
No. Support and resistance can identify places where attention and orders may concentrate. Their useful role is contextual. They need separate evidence before they become a directional trading rule.
Support and Resistance at Round Numbers: The Bottom Line
Round-number clustering is real. In a large U.S. securities-account dataset, integer-price transactions occurred 3.73 times as often as a uniform cent-ending baseline would predict. That tells you where attention gathers, not where price must reverse.
Use the round level as a map marker. Decide the trade with an acceptable-price boundary, a risk rule, and evidence that can be tested without hindsight. When those elements disagree with the clean line on the chart, the line should lose.
AI-assisted tools were used to organize source material and support editorial review. Danny Hwang reviewed the primary evidence, calculations, limitations, and final wording.
Update History
- July 21, 2026: Rebuilt the article around what the cited studies directly measure. Removed the unsupported $211 execution-drag model, its fixed 10-basis-point assumption, the universal two-to-three-cent order prescription, and the mismatched calculator. Corrected the 50-cent uniform baseline to 2%, narrowed the 3.73-times claim to integer-ending transactions, and labeled the $813 million figure as a rough historical aggregate estimate.
- May 4, 2026: Original publication.
Educational quantitative analysis based on published data. Not investment, tax, or legal advice. Consult a licensed professional before acting on any calculation. About TheFinSense.