---
title: "Volume breakout vs trend-confirmed breakout: the broader shortlist led"
description: "A 31-session IDX replay compares two five-name shortlists. The broader Screener had stronger reported outcomes; trend confirmation reduced selections without improving the median."
author: "Trade Inc Research"
published_at: "2026-09-17T08:28:15Z"
updated_at: "2026-09-17T08:28:36Z"
canonical: "https://tradeinc.app/en/blog/volume-breakout-vs-trend-confirmation-aug-sep-2026"
tickers: []
citations: ["https://www.schwab.com/learn/story/ways-volume-can-help-confirm-price-trends","https://www.schwab.com/learn/story/how-to-read-stock-charts-and-trading-patterns","https://www.fidelity.com/learning-center/trading-investing/technical-analysis/technical-indicator-guide/sma"]
---

# Volume breakout vs trend-confirmed breakout: the broader shortlist led

A 31-session IDX replay compares two five-name shortlists. The broader Screener had stronger reported outcomes; trend confirmation reduced selections without improving the median.

Last updated: September 17, 2026. Screening window: August 1–September 16, 2026, inclusive. Report snapshot: September 17, 2026, 14:28 WIB. Outcomes are provisional stored-bar observations, not a frozen September 16 closing-price dataset.

## Summary

The broader volume-breakout Screener produced a reported mean 10-observation open-to-close return of **4.57%**, versus **1.83%** for the trend-confirmed Screener. The median was much less flattering: **0.35%** for the broader shortlist and **−1.52%** for the trend-confirmed version. These figures describe completed historical signal observations, not returns on an investable portfolio.

Both Screeners look for liquid Indonesian stocks breaking out with elevated volume. The second adds a strong-uptrend condition. Over 31 covered screening sessions, it generated 112 shortlist records versus 134 for the broader Screener—a reduction of 16.4%—but did not show a better typical outcome in this snapshot.

For research, the broader shortlist is the stronger candidate for further validation on these reported measures. The trend-confirmed version offers fewer names to review, not demonstrated downside protection. Because each Screener retains at most five ranked names per session, this is a comparison of two selection processes, not an isolated test of the trend filter's causal effect.

## Key takeaways

- **The broader shortlist led on mean and median outcomes.** Its mean advantage was 2.74 percentage points; its median advantage was 1.87 points.
- **A positive average did not describe the typical result well.** Only 54 of 105 completed observations in Screener A, and 37 of 88 in Screener B, finished above their entry open.
- **Touching a gain was not the same as retaining it.** A 15% high was observed in 36 completed A samples, but only 23 ended at least 15% above entry. For B, the counts were 23 and 13.
- **The sample remains incomplete.** A had 29 pending observations and B had 24. Repeated tickers and overlapping holding windows also limit independence.
- **This is exploratory evidence.** There is no cost-adjusted portfolio, verified fill model, market-index comparison, or untouched out-of-sample validation in this report.

## On this page

- [What each screener looks for](#what-each-screener-looks-for)
- [Results at a glance](#results-at-a-glance)
- [Methodology: what was actually compared](#methodology-what-was-actually-compared)
- [Where the broader shortlist helped](#where-the-broader-shortlist-helped)
- [What trend confirmation did—and did not—deliver](#what-trend-confirmation-didand-did-notdeliver)
- [Why A led](#why-a-led-better-outcome-breadth-not-just-more-signals)
- [Strategy mechanics](#strategy-mechanics)
- [How to use the comparison](#how-to-use-the-comparison)
- [Limitations and the next test](#limitations-and-the-next-test)
- [Conclusion](#conclusion)
- [FAQ](#faq)
- [Appendix: five best and five worst stocks](#appendix-five-best-and-five-worst-stocks)

## What each screener looks for

**Screener A: volume breakout.** It looks for a new 20-session price high with volume at least 1.5 times the reference level. The tested rules also require a nonnegative daily price change, a price of at least Rp50, and 20-session average daily traded value of at least Rp10 billion. Price supplies the breakout condition; volume tests whether activity is elevated alongside it.

**Screener B: trend-confirmed breakout.** It starts with exactly those conditions, then requires the platform to classify the stock as being in a strong uptrend. The general idea is to look for an established upward direction as well as a breakout. This is an additional selection rule—not a guarantee of a better outcome, and not a moving-average crossover we have independently specified.

![Conceptual comparison of a volume breakout and the same setup with an additional strong-uptrend condition; not actual market data](https://cdn.sanity.io/images/eybaj6wm/production/26b85da2c6b55a5bd83dce1f0ccbe03abe188855-1200x756.png)

*Conceptual illustration—not actual market data. The two panels intentionally show the same schematic setup; only B adds the trend-classification condition. Neither chart depicts the classifier's formula or a predicted price path.*

**Both screeners retain up to five ranked names per session, so adding a filter can change which stocks enter the shortlist.** The full test assumptions appear below; public references for volume and trend concepts are in the strategy-mechanics section.

## Results at a glance

Screener A is the volume-breakout baseline. Screener B adds trend confirmation. Percentages below use completed samples unless the row explicitly describes total selections.

| Measure | Screener A: volume breakout | Screener B: trend-confirmed |
| --- | ---: | ---: |
| Covered screening sessions | 31 / 31 | 31 / 31 |
| Selected ticker-session records | 134 | 112 |
| Completed holding samples | 105 | 88 |
| Pending holding samples | 29 | 24 |
| Mean open-to-final-close change | +4.57% | +1.83% |
| Median open-to-final-close change | +0.35% | −1.52% |
| Positive final-close observations | 54 / 105 (51.43%) | 37 / 88 (42.05%) |
| Negative / flat final-close observations | 47 / 4 | 49 / 2 |
| High reached at least +15% from entry open | 36 / 105 (34.29%) | 23 / 88 (26.14%) |
| Final close reached at least +15% from entry open | 23 / 105 (21.90%) | 13 / 88 (14.77%) |

Source: Trade Inc comparison report, captured at 14:28 WIB on September 17. The mean and median are per-signal price changes with no fees or slippage. They are neither compounded returns nor returns earned by holding five equally weighted positions. The reported 15% threshold is a grading convention, not an executed take-profit order.

## Methodology: what was actually compared

The research question is narrow: **how did two breakout shortlists differ when one required trend confirmation?** The audience is investors evaluating a screening process, not readers seeking a current stock recommendation.

The two manual-only definitions were fixed before their results were retrieved. No alternative thresholds were swept to select the winner. However, the test period is historical and was already over when the Screeners were created; it is not a prospective track record.

| Design element | Shared rule or distinction |
| --- | --- |
| Historical universe | The replay reports 991 symbols; eligibility is then determined by the rules below. Historical constituent completeness is not independently certified. |
| Breakout condition | The platform's 20-session new-high condition, nonnegative daily price change, and volume ratio of at least 1.5×. |
| Liquidity condition | Price of at least Rp50 and 20-session average daily traded value of at least Rp10 billion. |
| Additional rule in B | The platform's uptrend regime with strength classified as strong. This is not represented as a specific moving-average crossover. |
| Output | Up to five ranked, distinct names per session for each Screener. A ticker may recur on different dates. |
| Price convention | Entry-session open to the final available close in a 10-bar holding window, including the entry bar. |
| Completion | The report labels an observation completed when it has the required 10 stored daily bars. Pending samples do not enter the completed mean or median. |
| Capital and execution | No cash budget, position sizing, compounding, fill rejection, or overlapping-position netting is simulated. |
| Costs | No transaction fees or slippage are applied; these are not after-cost returns. |

The five-name cap matters. Filtering changes which names can reach the shortlist; B need not be a subset of A's retained records. For example, the report contains three completed MDIA observations for B and two for A. The 22-record difference therefore means **22 fewer selected records overall**, not 22 identified A trades that the filter successfully avoided.

The report confirms 31 covered screening sessions and no unsupported rule nodes. Its separate outcome-window counter is 32; those counters are not treated as interchangeable trading-calendar evidence. Outcomes may use stored bars after the September 16 screening endpoint. Repeated retrievals changed some marks, so this article fixes one retrieval snapshot rather than claiming a settled end-of-day cutoff.

## Where the broader shortlist helped

### The advantage was not confined to the average

A's 4.57% mean exceeded B's 1.83%, but the median is the more useful counterweight. At 0.35%, A's middle observation was only slightly positive before trading costs. B's middle observation lost 1.52% on the same price convention.

How B compares: it had a positive mean despite 49 negative and only 37 positive completed observations. This is consistent with a distribution in which larger gains offset more frequent losses. A positive mean alone is insufficient evidence of a dependable trading experience.

### Weekly cohorts expose a real losing stretch

These are entry-week groups, not calendar-week portfolio returns. Each mean follows the group's completed observations through their own holding windows.

| Entry week starting | A: completed samples | A: mean change | B: completed samples | B: mean change |
| --- | ---: | ---: | ---: | ---: |
| August 3 | 25 | +9.35% | 17 | +4.82% |
| August 10 | 16 | −3.43% | 11 | −5.44% |
| August 17 | 19 | +9.98% | 15 | +11.49% |
| August 24 | 20 | +5.63% | 20 | +2.78% |
| August 31 | 25 | −0.03% | 25 | −3.56% |

A led in four of these five groups, but both lost in the August 10 cohort. B led in the August 17 cohort. Neither observation identifies a market regime or explains its cause; these are calendar groupings, not predeclared bull-, bear-, or sideways-market tests.

The September 7 and September 14 cohorts had no completed samples in this snapshot. Their 29 A records and 24 B records remain visible as pending, not silently excluded as if no signals occurred.

## What trend confirmation did—and did not—deliver

### Fewer names, without a demonstrated improvement in outcomes

B produced 112 records, or 16.4% fewer than A's 134. That is a smaller review queue. It is not measured turnover reduction: repeated names may overlap, and no portfolio was built.

How A compares: the broader process required reviewing more records and achieved stronger completed-sample mean, median, and positive-close frequency. This does not establish that all of its extra names were useful, because shortlist membership also changed.

### Reaching a threshold did not lock in the gain

B's high reached +15% in 23 of 88 completed observations, but only 13 finished at or above that threshold. For A, the corresponding counts were 36 and 23 out of 105. The gap illustrates the difference between an observed intrahold opportunity and an end-of-window outcome.

It does not prove that an investor could have sold at the high or filled a target order. The report's separate recorded-target hit rate is unavailable because neither Screener's observations have recorded target prices. That missing metric must not be confused with a 0% success rate.

## Why A led: better outcome breadth, not just more signals

Our reading is that A retained a more productive mix of names and entry dates in this window. Merely having more records cannot explain a higher per-record average. More importantly, A's advantage survived two checks that a few spectacular gains cannot settle alone: its median was higher, and 51.43% of completed observations finished positive versus 42.05% for B.

### B's gains were more concentrated

The two largest contributing stocks accounted for 40.84% of A's summed positive return percentages, versus 53.73% for B. This measure adds only positive completed price changes, groups them by stock, and measures the largest two groups' share. It is not a share of net portfolio profit, and repeated observations still count.

That distinction sharpens the editorial conclusion. A was not simply a larger list rescued by one impressive average; B actually depended more heavily on a narrow group of positive contributors under this measure. Nevertheless, A's +0.35% median leaves little room to assume an attractive after-cost result.

The stock-level rankings also resist a simplistic story. PACK's completed-observation average was stronger in B (+41.44%) than A (+38.69%), yet B's overall average was lower. Retaining a strong winner is not sufficient when the remaining selection mix is less productive. These are different observation sets, so this is an illustration of composition—not a matched-trade attribution.

### Confirmation may trade earlier access for stronger evidence

A breakout and an established trend ask different questions: has price crossed a recent boundary, and has directional persistence already developed? Requiring both can plausibly miss an early move or admit a stock at a later stage. Familiar trend tools illustrate that timing trade-off: Fidelity explains that longer simple moving averages smooth more but react with greater delay. That is background, not evidence that B uses a moving average. [Fidelity: Simple Moving Average](https://www.fidelity.com/learning-center/trading-investing/technical-analysis/technical-indicator-guide/sma).

For B, later selection and less remaining upside are **hypotheses**, not measured findings. Establishing them would require matching the same stocks across both processes and comparing first eligibility, entry prices, prior advances, and later outcomes. The available aggregate reports cannot do that.

### The strongest alternative is a short-window composition effect

B outperformed in the August 17 entry cohort, even though A led overall. That is a reason not to conclude that confirmation is intrinsically harmful. A different mix of market conditions, names, or mature outcomes could change the ranking.

The defensible conclusion is therefore process-specific: B's additional condition and resulting five-name selection did not improve these reported outcomes. The evidence does not identify whether the decisive factor was timing, ranking displacement, market conditions, or a small number of extreme observations.

## Strategy mechanics

A **breakout** is a move above a prior price boundary; a downward breach is commonly called a breakdown. Volume supplies context for the move. Schwab describes above-average activity as possible confirmation, while warning that high volume can also accompany exhaustion. This is a rationale for testing volume—not a guarantee of continuation. [Schwab: Volume and price trends](https://www.schwab.com/learn/story/ways-volume-can-help-confirm-price-trends).

An **uptrend** conventionally involves successively higher peaks and troughs. Schwab's chart guide explains that structure; the platform's “strong uptrend” classifier is a separate implementation, not certified by that guide. [Schwab: Trends and chart patterns](https://www.schwab.com/learn/story/how-to-read-stock-charts-and-trading-patterns).

A uses the price-and-volume setup plus liquidity requirements. B asks for the same setup with additional trend confirmation. The distinction is between a breakout condition and a breakout that also passes a trend classification—not between a technical strategy and a risk-free version of it.

These public references explain the concepts only. Our thresholds, five-name selection, historical outcomes, and interpretation come from Trade Inc Research's own replay, not from the cited publishers.

## How to use the comparison

| Research objective | What this evidence supports |
| --- | --- |
| Choose a baseline for further validation | Retain A as the reference: it led on the reported mean and median. |
| Reduce the number of shortlist records to inspect | B did that in this window, but its typical reported outcome was weaker. |
| Establish lower portfolio risk | Neither report answers this; there is no portfolio drawdown or exposure path. |
| Decide whether the filter itself adds value | Compare full same-date candidate sets before the five-name cap, then separately test ranked shortlists. |

These are research choices, not allocation instructions. Combining both outputs would also require matching ticker and date to avoid counting shared exposure twice.

## Limitations and the next test

**The observations are not independent trades.** The same stock can appear repeatedly, with overlapping 10-bar windows. Neither a conventional independent-sample significance claim nor a portfolio return can be inferred from these counts. Statistical uncertainty has not been quantified.

**The outcome snapshot is provisional.** It was taken during the September 17 session, and the endpoint supplied no immutable outcome-data version or settled-bar certification. The report's word “completed” refers to bar count; it is not an independent audit of final closing prices. Pending samples may change the comparison as they mature.

**Coverage is not a complete bias audit.** The 31/31 replay coverage establishes availability for the reported rules. It does not certify point-in-time universe membership, delisted-stock coverage, corporate-action adjustment, dividend inclusion, or the exact historical availability of corrected price data. Returns are reported price changes, not verified total returns.

**Large individual observations deserve scrutiny.** RGAS appears among the worst completed outcomes at −57.86% in both reports. That is a reported price change, not an independently verified executable loss. It remains in the evidence; no company-event explanation or adjustment is invented to explain it away.

**There is no external investment benchmark.** A is the comparator for B, not an IDX index or buy-and-hold portfolio. This study cannot establish market outperformance or risk-adjusted superiority.

The next useful test would preserve both definitions, freeze settled outcome data, and inspect complete candidate sets and ranked selections separately. It would then evaluate an untouched later period with explicit position sizing, costs, liquidity constraints, corporate-action treatment, and a matched passive benchmark. It should report all pending and rejected observations, not just resolved winners.

## Conclusion

For the reported August–September sample, the broader breakout shortlist led the trend-confirmed process on both mean and median outcomes. The practical lesson is narrower than “trend filters fail”: **an additional condition did not earn a presumption of better selections in this comparison**.

The evidence would strengthen if A retained its advantage on settled, cost-aware outcomes in a later untouched period. It would weaken or reverse if pending outcomes, corrected data, or a properly specified portfolio test favored B. Until then, A is a research baseline—not a validated investment strategy—and B's demonstrated benefit is fewer shortlist records.

## FAQ

### Is the reported 4.57% a portfolio return?

No. It is the arithmetic mean of 105 completed signal-level open-to-final-close changes. It does not incorporate capital allocation, overlapping holdings, or costs.

### Does “win” mean any positive return?

Not in every report field. This article separates positive final closes from the default +15% high and final-close thresholds. The denominators are shown beside each measure.

### Were these picks published before their historical dates?

No such claim is made. Both reports identify all selected records as reconstructed replay picks and report zero live picks. This is retrospective research, not an audited historical recommendation record.

### Can the appendix be used as a current buy list?

No. It documents dated examples from the replay, including losing and flat observations. It does not assess today's prices, valuation, or suitability.

## Appendix: five best and five worst stocks

These rankings use each stock's **mean open-to-final-close change across its completed observations within that Screener**, not its single best trade. They reproduce the five highest and five lowest stock cohorts supplied by the report. “Completed / selected” shows the denominator and any still-pending selections.

This appendix limit is separate from the replay's five-name daily shortlist cap. Rankings are hindsight descriptions, not current recommendations or evidence that a stock repeatedly offers the same opportunity. One-observation rows are especially fragile; overlapping observations are not independent trades. No minimum sample-size filter was imposed.

### Screener A: five best stocks

| Rank | Ticker | Completed / selected | Mean close change |
| --- | --- | ---: | ---: |
| 1 | MDIA | 2 / 2 | +41.17% |
| 2 | PACK | 7 / 7 | +38.69% |
| 3 | SINI | 5 / 5 | +19.98% |
| 4 | PIPA | 3 / 4 | +19.22% |
| 5 | ARCI | 1 / 1 | +18.22% |

### Screener A: five worst stocks

| Rank | Ticker | Completed / selected | Mean close change |
| --- | --- | ---: | ---: |
| 1 | RGAS | 1 / 1 | −57.86% |
| 2 | GMFI | 1 / 1 | −20.00% |
| 3 | VKTR | 1 / 1 | −16.75% |
| 4 | KOTA | 2 / 2 | −13.63% |
| 5 | SQMI | 1 / 1 | −13.00% |

### Screener B: five best stocks

| Rank | Ticker | Completed / selected | Mean close change |
| --- | --- | ---: | ---: |
| 1 | PACK | 6 / 6 | +41.44% |
| 2 | MDIA | 3 / 3 | +30.01% |
| 3 | PIPA | 3 / 4 | +19.22% |
| 4 | KETR | 1 / 1 | +11.05% |
| 5 | AADI | 2 / 4 | +10.86% |

### Screener B: five worst stocks

| Rank | Ticker | Completed / selected | Mean close change |
| --- | --- | ---: | ---: |
| 1 | RGAS | 1 / 1 | −57.86% |
| 2 | TMPO | 1 / 1 | −33.33% |
| 3 | GMFI | 1 / 1 | −20.00% |
| 4 | SQMI | 1 / 1 | −13.00% |
| 5 | GIAA | 1 / 1 | −12.00% |

### Dated examples from the report

The following six rows are illustrations selected from the report's limited dated sample to include gains, losses, and a flat observation. They are not the individual trades underlying every ranked stock above, and are not a random sample. Dates are entry sessions—not proof of historical publication. Figures use the same provisional, before-cost 10-bar price convention.

| Screener | Entry session, 2026 | Ticker | Reported close change |
| --- | --- | --- | ---: |
| A | August 3 | EMAS | +14.39% |
| A | August 19 | DOOH | 0.00% |
| A | August 27 | RGAS | −57.86% |
| B | August 3 | MDIA | +45.08% |
| B | August 31 | AUTO | −4.90% |
| B | September 4 | PTBA | +5.82% |

Do not recompute the headline mean from either the extreme-stock rankings or these illustrative dated rows. The headline uses all completed observations, including names outside this appendix.

## Sources and research attribution

**Original research:** Trade Inc Research, historical Screener replay for August 1–September 16, 2026; numerical snapshot captured September 17, 2026 at 14:28 WIB. All performance tables and dated examples are our own reported replay results. The underlying reports require private access and are not presented as public citations. Readers can inspect the disclosed design and tables here, but cannot independently reproduce the full result from the conceptual references below.

**Public conceptual references:**

- [Charles Schwab — 3 Ways Volume Can Help Confirm Price Trends](https://www.schwab.com/learn/story/ways-volume-can-help-confirm-price-trends), June 5, 2026. Background on volume confirmation and its limitations.
- [Charles Schwab — How to Read Stock Charts and Trading Patterns](https://www.schwab.com/learn/story/how-to-read-stock-charts-and-trading-patterns), May 29, 2025. Background on trends, support, resistance, and breakouts.
- [Fidelity — Simple Moving Average](https://www.fidelity.com/learning-center/trading-investing/technical-analysis/technical-indicator-guide/sma), accessed September 17, 2026. Background on smoothing and signal lag; not the specification of Screener B.

These references do not validate our replay, endorse either Screener, or establish why one outperformed the other.


## Methodology

Exploratory comparison of two retrospectively constructed, manual-only IDX Screener shortlists over August 1–September 16, 2026 inclusive. Source: paired Trade Inc rule-replay comparison retrieved September 17, 2026 at 14:28 WIB; all published figures use that saved response, not a subsequently refreshed report. This is a retrieval cutoff, not a certified market-close cutoff. Definitions were fixed before outcome retrieval, with no threshold sweep; there is no untouched validation period. Both require the platform's 20-session new high, nonnegative daily change, volume ratio >=1.5, price >=Rp50, and average daily traded value over 20 sessions >=Rp10 billion. B additionally requires platform-classified uptrend/strong strength. Both retain up to five distinct ranked names per session; changes in candidate eligibility can change shortlist membership, so B's records need not be a subset of A's. The report universe is 991 symbols and covered screening sessions are 31/31; a separate outcome-window counter is 32 and has not been independently reconciled as an exchange calendar. All 134 A and 112 B records are replay-generated, with zero live picks. Completed/pending counts are 105/29 and 88/24. Grading measures entry-session open to final stored close over 10 daily bars including entry; pending observations are excluded from completed means, medians and threshold frequencies. The default +15% high/close tests are not executed exits. Recorded target-price hit rates are unavailable. Repeated ticker-session observations receive equal weight; capital, compounding, position sizing, overlapping-position netting, fees, slippage and rejected fills are not simulated. A is B's selection-process baseline, not a passive market benchmark. Corporate-action adjustment, dividends, historical constituent completeness, delisted coverage, settled-bar finality and original as-published data availability are not independently certified. Weekly rows are entry cohorts, not portfolio weekly returns or predeclared market regimes. No statistical significance or causal filter effect is claimed. Appendix extremes rank stocks by mean completed price change with completed/selected counts and no minimum sample filter; dated examples are explicitly illustrative, not random or a complete ledger. Both screeners and all unfavorable groups are retained.

[1]: https://www.schwab.com/learn/story/ways-volume-can-help-confirm-price-trends "Schwab — Volume confirmation: conceptual background"
[2]: https://www.schwab.com/learn/story/how-to-read-stock-charts-and-trading-patterns "Schwab — Trends and chart patterns: conceptual background"
[3]: https://www.fidelity.com/learning-center/trading-investing/technical-analysis/technical-indicator-guide/sma "Fidelity — Simple Moving Average: smoothing and lag"