Breakouts led the week while sweet spot setups stayed selective

THE EDGE THIS WEEK

Breakout volume was high, but quality stayed narrow

This week showed a familiar split between broad scanner activity and selective tradable edge. The system found 366 setups, yet only 56 converted into breakouts, which points to a market where participation existed but follow-through was concentrated in a relatively small group of symbols and sectors.

The strongest pattern was not raw setup frequency. It was symbol-specific persistence. The best names delivered repeated clean outcomes across multiple trades, while weaker areas produced frequent failures and negative expectancy. That suggests traders were rewarded more for discrimination than for aggression.

BY THE NUMBERS

Weekly performance snapshot

The broad backtest remained constructive with a 64.7% average win rate across 2,439 trades, but the live scanner data was much less forgiving. That gap matters. It suggests the environment still rewarded the right pattern in the right name, but the average real-time setup had lower odds of continuation than the historical aggregate would imply.

In practical terms, this was a week for selectivity and confirmation. Traders looking for every setup likely encountered more churn, while traders willing to narrow focus toward stronger symbols and cleaner structure had a better chance of aligning with what was actually working. For current opportunities, see live setups in the scanner.

SWEET SPOT REPORT

The usual timing window underperformed

The sweet spot cohort, defined by pressure below 60 and setups forming over 3 to 5 bars, materially underperformed its historical baseline. A 29.0% win rate versus a long-run 58.5% win rate is not noise. It is a signal that the usual balanced-launch condition did not translate into reliable expansion this week.

There are two likely interpretations. First, lower-pressure setups may have been too early in a tape that demanded stronger confirmation before resolution. Second, the market may have been producing many acceptable-looking consolidations without enough directional sponsorship to convert them. In both cases, the pattern says timing quality deteriorated even when structural definitions were met.

SYMBOL SPOTLIGHT

Where the edge concentrated

LLY stands out because it combined perfection with meaningful sample depth. Ten trades at a 100.0% win rate and roughly 1R average outcome suggests an unusually clean environment where breakouts were not just triggering, but resolving efficiently. This is the type of symbol behavior that often appears when institutional sponsorship stays consistent across multiple sessions.

GDX is notable for a different reason. Its 94.44% win rate across 18 trades is one of the larger high-quality samples in the report. That matters more than a smaller perfect sample because it indicates repeatability. When an ETF or thematic vehicle produces that kind of consistency, it often reflects broad participation rather than isolated single-name strength.

COST represents the opposite side of the ledger. An 18.18% win rate across 11 trades with negative average R is not just random underperformance. It suggests that the symbol repeatedly attracted setups that lacked expansion quality or failed quickly after trigger. That pattern is useful because persistent laggards often keep failing until a meaningful character change appears.

Other top names such as NOW, URI, CVX, ZM, V, ISRG, TGT, and ELF reinforce the same theme. The edge was strongest in symbols that kept proving themselves over repeated tests, not in symbols that merely printed frequent setups.

WHAT THE BOTS MISSED

Low count, high opportunity cost

The bots missed only five trades, but the cost of those misses was large at +12.1R total. That is an average of more than 2R per missed opportunity, which tells us the blocked trades were not marginal signals. They were among the week’s more productive outcomes. To watch the bots in the Edge Lab.

The key detail is that every miss was tied to the same filter family: rvol_threshold. That points to a possible calibration issue rather than random omission. In a week where broad breakout conversion was weak, the volume filter may have been too strict in the few places where quality setups were still able to run.

This does not automatically mean the filter should be loosened. More likely, it suggests the threshold needs regime sensitivity or symbol-specific tolerance. If opportunity is clustering in fewer names, a static relative-volume gate can reject exactly the setups traders most need to capture.

SECTOR HEAT MAP

Leadership leaned toward technology, but breadth was uneven

Sector breakout concentration was led by Technology with 18 breakouts, followed by Financials with 8, ETFs with 7, Energy with 5, Consumer with 4, and Healthcare with 2. The duplicated category labels in the raw feed, including separate entries for Technology and technology as well as ETF and etf, imply some classification fragmentation, but the directional message is still clear: leadership was concentrated in growth-oriented and liquid institutional groups.

Technology’s lead matters, but not because it was dominant in an absolute sense. It matters because the next tier fell off quickly. That shape usually reflects selective risk appetite rather than broad market conviction. Financials and ETFs contributed enough to confirm some cross-sector participation, while Energy added a smaller but still meaningful pocket of confirmation.

The larger pattern is that breakouts were present across multiple groups, but not in a way that supported indiscriminate exposure. Concentration remained high, and that reinforces the week’s central theme: edge was available, but only in pockets where participation and follow-through aligned.

RESEARCH NOTE

When headline win rates stay healthy but scanner conversion falls, focus on dispersion

The most important research takeaway from this week is the divergence between strong aggregate backtest performance and weak live scanner conversion. A 64.7% average win rate across the backtest universe can coexist with a 15.3% breakout rate in the scanner when the market is rewarding a narrow set of symbols while rejecting the majority of acceptable-looking setups.

That is a dispersion problem, not a simple market-strength problem. In dispersed conditions, the right question is not whether breakouts are working in general. It is whether edge is clustering in a small leadership basket while the rest of the tape produces noise. This week’s top performers and weak sweet spot results strongly support that interpretation.

For traders, the implication is straightforward. When dispersion rises, screening and ranking matter more than setup count. The best process is usually to prioritize proven leaders, de-emphasize mediocre structure, and treat broad scanner activity as a search field rather than a signal by itself.