Support and resistance help organize a chart into areas worth examining. They do not reveal a price that must reverse, an order that must fill or a trade that must be profitable. The useful question is not merely “Where is the line?” but “What information defined this area, what would contradict the idea, and what would acting on it cost?”

This guide develops a repeatable chart-reading exercise rather than a live signal. Every price, quantity, cost and plotted path below is invented. The examples are not current gold levels, a tested strategy or MoneyTopTon trading results. A clear drawing is a starting point for investigation, not evidence that real capital should be committed.

1. What support and resistance describe

Support is an area being considered for a pause or reversal in a downward move; resistance is an area being considered for a pause or reversal in an upward move. The same location can have different roles depending on where price currently stands. CME Group's educational lesson describes these areas as zones rather than exact prices.

Do not confuse a descriptive label with a causal explanation. A chart alone does not tell you how many orders remain there or whether their owners will cancel them. “Buyers defended this area” is an interpretation of observed movement, not a complete view of the order book. Use language such as “candidate support” when recording an untested idea.

2. Mark a zone instead of demanding an exact touch

Suppose completed swings cluster near 2,998, 3,000 and 3,002 dollars per ounce. An exercise could mark 2,998–3,002 as one candidate zone. This interval is a choice made from the selected observations; it is not a universal four-dollar width. Another dataset, volatility regime or timeframe may require a different definition.

Record whether the boundaries come from wicks, closing prices or both. Mixing those rules only after a trade fails makes the drawing difficult to audit. A zone so wide that it includes nearly every nearby move may explain the past beautifully while giving little guidance about when the idea would be wrong.

Invented price path between a lower 2998–3002 zone and upper 3018–3022 zone3018–3022 USD/oz2998–3002 USD/ozt →
Schematic only: the path and zones are invented, the horizontal direction represents sequence rather than dated candles, and no breakout probability is implied.

The shaded areas make uncertainty visible. They are not executable buy and sell prices. A touch of the lower band, a close inside it and a close below it are different observations that should not be silently treated as the same event.

3. Start with consistent data and completed swings

Write down the instrument, feed, quote currency, timeframe and candle timezone before marking levels. Spot quotations, a particular futures contract and a retail derivative are not interchangeable charts. Contract expiry, rollover conventions and session boundaries can change what a historical high looks like. Use the same series for drawing and subsequent evaluation.

Select a visible swing using a rule defined before the next move. A pivot that needs two later candles is only known after those candles finish; a backtest must not act as though it was known at the earlier turning point. Keep the original screenshot and decision time. That simple discipline prevents future information from making an old decision appear easier than it was.

4. Give each timeframe a defined job

One possible study workflow uses a daily chart for broad context, an hourly chart for a candidate zone and a shorter chart for observing execution. These are illustrative roles, not optimal settings. Choose them in advance so that switching charts cannot become a way to search indefinitely for a picture that agrees with a preferred trade.

A short-term upward move can occur beneath a higher-timeframe resistance area. That is a conflict to document, not proof one chart is wrong. Record the distance to the next relevant zone and decide whether the exercise should wait. A higher timeframe does not make a level unbreakable, and a lower timeframe does not automatically provide a safer entry.

5. Choose levels without filling the chart with lines

Begin with a small, declared set of observations, such as the nearest completed swing above and below price. Add a level only when its selection rule can be explained. Several nearly identical horizontal marks can often be represented by one zone, but preserve the underlying observations in the journal rather than discarding inconvenient ones.

Repeated visits do not mechanically certify strength. They can show that an area has mattered, while also occurring during changing participation and conditions. Do not invent a win probability from the number of touches. If you score zones, define the scoring method before looking at later outcomes and test whether the score adds information beyond a simple baseline.

6. Distinguish a touch, a break and a retest

With an upper zone of 3,018–3,022, a print at 3,021 remains inside the area. A completed candle closing at 3,024 is outside its upper edge, but calling it a breakout still depends on your rule: one close, a distance filter or another predefined observation. No definition eliminates failed breaks.

A later return to 3,021 is a revisit of the area, not automatic confirmation that it became support. A subsequent close at 3,016 would fall below the entire marked zone. These prices describe a sequence, not an order recommendation. Some moves never retest; waiting may miss them, while entering earlier introduces different uncertainty. Neither choice can be assessed honestly without including its missed and failed cases.

7. Define failure before seeing the outcome

Specify what counts as a false breakout and the observation window used to label it. For example, a research rule might require a close outside the zone and a return inside within a fixed number of completed candles. The chosen number is a parameter to evaluate, not a magic filter. A wick beyond the boundary and a close beyond it should be recorded separately.

Do not redraw the zone after the fact so every failed break becomes a successful touch. Keep the first version and create a second version with a timestamp if new information justifies revision. During fast markets, a chart may show a brief crossing that could not have been traded at the displayed historical price. A visually correct classification is not the same as an executable result.

8. Separate the chart idea from the exit order

Invalidation states why an interpretation no longer fits your rules. A stop order is an instruction with its own trigger and execution behavior. An idea evaluated at candle close can be invalidated differently from an order triggered intrabar. Record both, so a position is not left unmanaged while waiting for a chart condition that the original plan never required.

Putting a stop outside a zone does not cap every possible loss at that distance. Gaps and execution differences can produce a worse result. If the required distance makes the planned monetary risk too large, examine a smaller quantity or no trade; moving the stop inward solely to fit a preferred size changes the idea being tested.

9. Calculate the cost of being wrong

Assume a hypothetical long fill at 3,006 USD per ounce, exposure of 2 ounces and additional round-trip charges of USD 4. The planned exit for a losing scenario is 2,996, while an upside observation is 3,018, near the lower edge of the upper zone. Use actual executable entry and exit sides in a real record; this exercise already treats fills as bid/ask prices, so do not subtract spread again.

Invented long-position outcomes, USD; entry 3006, exposure 2 ounces, additional charges 4
Exit fillPrice resultNet result
3018(3018 − 3006) × 2 = 2420
3024(3024 − 3006) × 2 = 3632
2996(2996 − 3006) × 2 = −20−24
2988(2988 − 3006) × 2 = −36−40

The planned loss magnitude is 24 and the net gain at 3,018 is 20, giving a reward-to-risk ratio of 20/24, about 0.83. Moving the target to 3,024 makes the arithmetic 32/24, about 1.33, but does not make that farther exit more likely or remove the intervening resistance. The adverse 2,988 fill shows why planned loss is not a guarantee.

10. Let quantity follow a documented budget

For a separate sizing exercise, suppose the loss budget is USD 30, fixed additional charges are USD 4 and the entry-to-planned-exit distance is USD 10 per ounce. The arithmetic ceiling is (30 − 4)/10 = 2.6 ounces. If the permitted step were 0.5 ounce, rounding down gives 2.5 ounces and a planned loss of USD 29.

That step and fixed fee are invented specifications. With a worse exit at 2,988, the same 2.5 ounces would lose 18 × 2.5 + 4 = USD 49. Real fee schedules may depend on size, so solve using the actual schedule rather than reusing this shortcut. If minimum size already exceeds a budget, no valid smaller order exists under those assumptions.

11. Treat news and execution as separate inputs

A zone drawn before a major release does not guarantee unchanged liquidity afterward. Record the event time and timezone, the price feed, spread observations and whether orders were accepted. A chart-only review can miss rejected orders, partial fills and the difference between a quoted price and a price available for the intended quantity.

If important information is missing, pausing the exercise is a valid outcome. Do not describe every break as a “stop hunt” without evidence of intent. Price alone cannot identify who acted or why. Maintain a distinction between observations, plausible interpretations and claims that would require order-level evidence you do not possess.

12. Build a journal that preserves the original map

Suggested fields for a zone-reading exercise
Before the eventAfter the event
Instrument, feed, timeframe, timezone and rule versionSame series and the observation end time
Zone edges, selection reasons and original screenshotTouches, closes and any explicitly timestamped revision
Definition of break, failure and invalidationWhich condition occurred, including no-trade cases
Quantity, planned exits and cost assumptionsActual fills, charges and net result if a trade occurred

Include zones that did not work and occasions on which no order was placed. A correct forecast without a fill is not realized profit. Separate chart-classification accuracy from trading results, and avoid counting several partial exits as independent predictions. Keep a new sample for revised rules rather than repeatedly adjusting one historical period until it looks persuasive.

13. Keep gold quotation and account currency distinct

The examples use USD per troy ounce, not a domestic bullion-shop quote. A chart boundary cannot be copied unchanged into another unit or currency. Account conversion fees, product structure and contract size matter alongside the direction of gold. At a purely invented conversion of ten local currency units per dollar, a USD 24 loss would be 240 local units before conversion charges.

That arithmetic is not a current exchange rate or a statement that a product is available to every reader. Keep the chart study in its original quote unit, calculate the account result using the contract terms, and only then show a clearly labeled household-currency conversion if needed.

14. Checklist and common questions

  • Use a known instrument, timeframe and candle timezone.
  • Save zone edges and selection rules before the next move.
  • Define a break, a failed break and invalidation separately.
  • Check the next zone before interpreting a distant target.
  • Calculate net outcomes and possible worse execution.
  • Retain failures, no-trade cases and rule versions.

Should I buy whenever support is touched?

A touch is an observation, not a complete decision rule. Costs, invalidation, quantity and missing information still need to be addressed.

Is a zone with more touches always better?

No fixed touch count provides a success probability. The definition, sample and changing conditions matter.

Can waiting for a close remove false breakouts?

No. It changes the entry condition and may change the available price; the tradeoff must be evaluated rather than assumed away.

Educational material prepared 3 October 2026. All illustrated paths and trading calculations are hypothetical, not live levels or performance claims.