Methodology

1. Data Sources and Update Cadence

U.S. listings are collected from the Nasdaq Trader directory, U.S. daily prices and classifications from Yahoo Finance, and Korean listings, prices and market caps from Naver Finance. Calculations use daily bars after each market closes. The displayed as-of date, rather than the current calendar date, identifies the latest observation. Holidays and early closes follow each market calendar.

• Split/dividend back-adjustment: when a split or dividend retroactively adjusts past prices, the stock's entire price history is re-downloaded and unified to the same adjustment basis. This keeps moving averages and RS from being distorted by an adjustment mismatch.
• Timing: it runs automatically right after the close, and on market holidays the previous trading day's data is kept.

2. Stock Universe and Exclusion Criteria

For the metrics to be statistically meaningful, the population must be stable. Trend Screener limits the universe with the following criteria.

• U.S.: listings from NYSE, NASDAQ and AMEX are filtered using ETF flags and security-name rules. These metadata rules do not guarantee removal of every SPAC, ADR, unit or other unusual security. Verify the instrument type before using a result.
• Korea: KOSPI and KOSDAQ stocks. So that barely-traded micro-caps don't distort the ranking, market cap is not used to cut the population — instead, when an industry or sector return is cap-weighted, names under roughly ₩10 billion carry no weight, so barely-traded micro-caps cannot swing a group's return. The RS population additionally requires valid price history and the data checks below.
• Under 250 trading days (about a year) excluded: without a year of tape there is no comparable base, so the name is left out of the RS population. With no RS rating it cannot meet the 'RS 70 or above' condition, and so never reaches the 8/8 list.
• Ultra-low-priced stocks excluded (U.S.): stocks closing under $1 are dropped from the U.S. universe — their volatility and noise are too high. Korea has no price floor.

3. Computing the Weighted Relative Strength (RS)

Relative Strength (RS) is a momentum measure that converts how strongly a stock moves within the whole market into a percentile rank. Instead of a simple one-year return, it uses a weighted average that puts more weight on recent periods.

The center of gravity is the 3-month window (40%) — leadership usually establishes itself over a quarter. On top of that, 1 month (30%) and 1 week (10%) confirm a recent turn, while 6 months (20%) checks that the advance has roots. The point of this split is to stop crediting a stock for an advance it made a year ago and has since given back. Because RS is a relative rank rather than an absolute return, it can surface strong candidates that "fall less" even in a down market. An RS of 99 means the top 1% by momentum.

4. Trend Template — the 8 Trend Conditions

An 8/8 in the table means a stock passed all 8 conditions of Mark Minervini's Stage 2 uptrend, applied identically to U.S. and Korean stocks.

1. Price > 150-day & 200-day MAPrice is above the mid- and long-term moving averages.
2. 150-day > 200-day MAThe mid-term average is above the long-term one (proper alignment).
3. 200-day MA risingThe current SMA200 is greater than its value 21 trading sessions earlier. This does not require a continuous rise between those dates.
4. 50-day > 150-day & 200-day MAThe short-term average is above the mid- and long-term ones.
5. Price > 50-day MAPrice is above the short-term average.
6. Price ≥ 52-week low × 1.30At least 30% above the floor.
7. Price ≥ 52-week high × 0.75Within 25% of the new high.
8. RS Rating ≥ 70Relatively strong versus the market.

Taken together, the eight conditions quantify a single state: the short-, mid-, and long-term moving averages are aligned and rising, price sits near a new high, and the stock is strong relative to the market.

5. Optional Filters — Volume and ADR%

Volume is an optional filter separate from the 8 conditions. Pick a floor for average volume over the last 50 trading days to see only stocks above it — 100K / 500K / 1M / 3M shares on the U.S. screener, 50K / 100K / 300K / 1M on the KR screener (tuned to each market’s volume distribution). It does not affect the pass (8/8) determination itself, so thinly traded but strong-trending names still appear on the default screen.

ADR% (Average Daily Range) is an optional filter of the same kind. It measures how much higher the high was than the low on an average day over the last 20 trading sessions, and is shown per stock in the ADR column.

ADR measures historical intraday range, regardless of direction. It does not set a maximum future return or a guaranteed stop distance. A stock can gap beyond its historical range, and a stop can execute at a worse price. The available thresholds, 3% to 5% in 0.5% steps, are interface options rather than validated optimal settings. ADR does not change the 8-condition pass count. See the risk-management examples.

6. Industry Classification and Industry Rankings

The Industry Rankings page groups stocks into a two-level classification (sector · industry) and ranks each group on one number — the cap-weighted composite return of its members, 0.10×1W + 0.30×1M + 0.40×3M + 0.20×6M, the same windows and weights as the RS above. That return is converted to a min-percentile within the day's cross-industry distribution to give the 0–100 Industry Score, and the rank is taken on the raw return rather than the rounded score. Leadership density and breadth are still computed and shown on the detail panel as evidence, but they are no longer pillars of the score. Sector ranking (added 2026-08-02) measures the level above with the same formula: the four cards atop the industry rankings page are sector ranks #1–#4. The U.S. ranks the 11 SPDR Select Sector ETFs (XLK, XLF, XLV, XLY, XLP, XLE, XLI, XLB, XLRE, XLU, XLC) on their own closes — Yahoo's U.S. sector taxonomy maps onto them 1:1, and the return of a tradable instrument is the truer sector performance. Korea's sectors are this site's own themes (secondary batteries, K-culture, defence, holding companies) rather than an index provider's, half of them have no usable ETF, and a mixed basis would make sector-versus-sector rank meaningless — so all 15 are measured as cap-weighted member baskets. Sector RS (1–99) is the same yardstick as a stock's RS: it is where the sector's composite falls in that day's all-stock RS distribution, so RS 70 means the sector beat 70% of stocks — and in a broad decline a sector can carry a high RS on a negative return, because falling less is still relative strength. The table lines up this week's, 3-week-old and 6-week-old ranks to catch leadership rotation (the 5-session change Δ5D lives on the sector cards and the rank-trend charts) — because a trend is most reliable when a strong stock sits inside a strong industry.

This Industry Score ranking is also what the screener uses. The ordering of the screener's industry filter list (the rank number before each item) and the leader basket on the Market Environment page (§7) both rank industries by this 0–100 Industry Score. Industry relative strength (industry RS, 1–99), which ranks groups on relative strength alone, remains a separate auxiliary metric used only as a fallback when ranking data is unavailable.

7. Defining a Leader

"Leader" quantifies William O'Neil's "Leader or Laggard" principle. It is the definition shared by the leader basket and per-session leader lists on the Market Environment page and by the Leadership pillar of the Industry Rankings, covering stocks that meet both of the following at once.

• The stock's RS Rating is 90 or higher.
• Its industry ranks within the top 10 of the Industry Rankings.

Here "top 10" follows the same Industry Score (0–100) ranking as the Industry Rankings page (§6). Past sessions are judged by that day's industry ranking — the industries ranked 1–10 on the industry rankings page are exactly the "strong industries." To narrow the screener the same way, set the RS filter to 90+ and check the top-ranked industries in the industry filter (the list is already sorted by this ranking).

8. Market Environment Score (TSS)

The Market Environment page condenses whether the current tape favors trend-following and breakout trading into a single 0–100 TSS (Trend Strength Score). It sums three pillars — Breadth (55) · Institutional Flow (30) · Index Trend (15) — and reads 77+ as a strong trend, 59–77 healthy, 41–59 neutral, 23–41 weak, and below 23 as risk-off. Because the total is 100× the weighted average of the three axes, all-neutral inputs score exactly 50, and the bands are placed at even 18-point steps around that midpoint (recalibrated 2026-08-01 — the old 85/70/55/40 was tuned around the downward bias of the axes that have since been removed). Through 2026-08-01 it also carried Breakout Health (35), Industry Leadership (20) and Opportunity Set (10); all three scored derived judgments (did a breakout follow through, are leaders clustered, how many setups qualify) that daily bars can only approximate, so they were dropped. What remains are direct counts of the tape. It measures breadth and index trend with an in-house market-cap-weighted composite rather than an external index, so the U.S. and Korea stay on the same yardstick.

9. Limitations and Cautions

Worked calculation and data limitations

For illustrative returns of 2%, 5%, 10% and 20% over 1W/1M/3M/6M, the raw score is 0.10×2 + 0.30×5 + 0.40×10 + 0.20×20 = 9.7%. This is a weighted input, not an investment return or RS 97. The engine ranks eligible raw scores using average ranks for ties, multiplies the percentile by 99, rounds and clips to 1–99. Without the other stocks, a standalone RS rating cannot be reproduced.

A screener row needs at least 200 observed closes, while the RS population needs at least 250. Missing RS inputs or an extreme one-day move in the six-month window can remove a stock from RS ranking; its displayed fallback RS is 1. Missing closes may be forward-filled; early gaps in the moving-average baseline may be filled with the first available close. Price corrections and changes in the universe can therefore alter results. The weights are design choices, not evidence of superior performance.

Revised September 7, 2026. Report a discrepancy.