Methodology & tax rules — how the return-potential composite + MTF screener works

The composite — fundamentals ARE the score

Screener.in shows metrics one at a time; it can’t rank the whole market by a single blended score. That blended score is the point of this site — and it is computed directly from the raw fundamentals, with no abstract sub-score layer in between.

1 · Percentile each metric, in its GOOD direction. Every metric is percentile-ranked 0–100 across the loaded universe. For “lower is better” metrics (PE, PB, PS, EV/EBITDA, P/FCF, PEG, D/E, pledge, beta, distance-below-52W-high) the percentile is flipped so cheap/safe = high. For “higher is better” metrics (ROE, ROCE, margins, growth, momentum, FCF-yield, upside, Piotroski, institutional stake) the percentile is used as-is.

2 · Weighted percentile average. The composite is the weighted mean of those percentiles, renormalised over the metrics the stock actually has:

pct[m]     = percentile of the metric across the universe (0–100, in its good direction)

composite  = Σ( wₘ · pct[m] )  /  Σ( wₘ )     over the metrics present (weight > 0)
higher composite = higher return potential (0–100)

A metric with weight 0 is excluded. A metric missing for a stock is dropped and the weights renormalise over the rest — never penalised with a zero. A stock with fewer than 4 present weighted metrics ranks last (too little to score fairly).

Default weights — value-heavy

Out of the box the weights tilt to value (cheapness), then quality, then growth / momentum / analyst light:

Group	Metrics (weight)
Valuation	PE 20 · PB 16 · PS 8 · EV/EBITDA 12 · P/FCF 10 · PEG 6 · Div % 2 · Mcap ₹cr 2
Quality	ROE % 10 · ROCE % 8 · ROA % 3 · Net mgn % 4 · Piotroski 4
Growth	EPS g % 4 · Rev g % 3
Health	D/E 8 · Int cover 3 · Pledge % 3
Momentum	1M % 1 · 6M % 3 · 1Y % 3 · off 52wH % 2 · off 52wL % 1 · Beta 4
Ownership	Promoter % 2 · FII % 2 · DII % 2 · FII+DII % 2
Analyst	Upside % 4

Percentile ranges (slider bands)

Each metric’s slider spans the 1st–99th data-percentile of real values, so one outlier can’t stretch the scale. A band left at its full min–max applies no filter (it’s ignored) — only a moved band screens.

Live re-weighting

The whole universe ships as one JSON; the percentiles and the composite are computed in your browser, so dragging any weight re-ranks instantly with no refetch. The heat gauge in the rank column shows the 0–100 composite and the thin strip under it shows which metrics drove the score.

Presets, bands & weights

Three controls drive the screen, two-way synced:

Metric weights = scoring. Each metric’s weight sets how much it drives the composite. Weight 0 excludes it. Not a filter — nothing is dropped, the ranking just re-tilts.
Metric filter bands = screening. Each metric’s dual-thumb band sets a numeric min/max filter by dragging. Left at full range = off. Reuses the same filter state the advanced builder writes.
Presets = scans. Each preset sets the filter bands and a weight profile and a sort, and surfaces the top ~100 so you narrow to your final 5. Picking one moves the bands + weights; dragging any band or weight switches to custom.

Presets are grouped Featured / Core / Named / Strategy. Best Return-Potential loads by default. Exact rule for each:

Featured

Preset	What it screens	Exact rule
Deep Value ★	PE < 6 and PB < 1.5 with a quality guard (ROE > 12 OR Piotroski ≥ 6 OR D/E < 1) to weed value-traps out of the cheap tail. Value-heavy weights, ranked by composite.	pe > 0 · pe < 6 · pb > 0 · pb < 1.5 — sort composite ↓
Best Return-Potential ★ (default)	Default. Ranks by after-tax, after-MTF-interest 1Y return (1Y − 12% interest − 12.5% LTCG on the gain), leaning MTF-eligible + beta < 1.2. The leveraged buy-and-hold shortlist.	MTF/F&O-eligible · beta < 1.2 — sort aftertax_1y ↓
Quality-Value + turnaround ★	Quality (ROE > 15 & low D/E) that is cheap and turning. Turnaround condition is selectable — default: 6M momentum up (r_6m > 0 and below 52W high).	roe > 15 · de < 1 — sort composite ↓

Core

Preset	What it screens	Exact rule
Value	Pure cheapness — value-heavy weights, no filter, ranked by composite.	no filter — whole universe — sort composite ↓
Quality-Value	Cheap AND strong: ROE > 15%, D/E < 1. Cheap without the value-trap risk.	roe > 15 · de < 1 — sort composite ↓
MTF Buy-Hold 1yr	MTF/F&O-eligible, beta < 1.2, ranked by after-tax after-interest 1Y return.	MTF/F&O-eligible · beta < 1.2 — sort aftertax_1y ↓
Deep-Value net-net	Asset-cheap: PB < 1, D/E < 0.5 — near/under book with a clean balance sheet.	pb > 0 · pb < 1 · de < 0.5 — sort composite ↓

Named

Preset	What it screens	Exact rule
Graham	Graham defensive: PE < 15, PB < 1.5, D/E < 1.	pe > 0 · pe < 15 · pb > 0 · pb < 1.5 · de < 1 — sort composite ↓
Magic Formula	Greenblatt: cheap (EV/EBITDA) + high return on capital (ROCE).	no filter — whole universe — sort composite ↓
Piotroski (F≥6)	Fundamentally strong: Piotroski F-score ≥ 6. Enriched shortlist.	f_score ≥ 6 — sort composite ↓

Strategy

Preset	What it screens	Exact rule
GARP	Growth at a reasonable price: PEG < 1, ROE > 15%, positive EPS growth.	peg > 0 · peg < 1 · roe > 15 · eps_growth > 0 — sort composite ↓
Turnaround	6M momentum turning up (r_6m > 0) with positive revenue growth, still cheap.	r_6m > 0 · rev_growth > 0 — sort r_6m ↓
Contrarian	Beaten-down but solid: 1Y return < 0, D/E < 1, cheap.	r_1y < 0 · de < 1 — sort composite ↓
FII/DII Accumulation	Smart-money-held + cheap: FII+DII ≥ 25%.	fii_dii ≥ 25 — sort composite ↓
High FCF-Yield	Top free-cash-flow yield: lowest P/FCF (most cash per rupee of price).	p_fcf > 0 — sort composite ↓

The three featured scans

Deep Value — PE < 6 AND PB < 1.5 with a documented quality guard (ROE > 12 OR Piotroski ≥ 6 OR D/E < 1) to drop value-traps; value-heavy weights, sorted by composite.
Best Return-Potential (default) — MTF/F&O-eligible, beta < 1.2, ranked by after-tax after-interest 1Y return (1Y − 12% MTF interest − 12.5% LTCG on the gain).
Quality-Value + turnaround — quality (ROE > 15 & D/E < 1) that is cheap and turning. The turnaround condition is selectable: default 6M momentum up (r_6m > 0 and below the 52W high), or earnings turn (EPS & revenue growth positive), or both.

Every metric — plain language

Metric	What it is	Why it matters	What cheap / good looks like
PE (P/E)	Price ÷ trailing earnings per share.	How many rupees you pay per rupee of annual profit.	Lower = cheaper. <15 cheap, >30 rich — but banks/utilities run low, growth runs high. Loss-makers have no meaningful PE.
PB (P/B)	Price ÷ book value (net assets) per share.	What you pay over the accounting net worth. Best for asset-heavy names (banks, PSUs).	<1 = below book. High-ROE businesses justify higher PB.
PS (P/S)	Market cap ÷ annual sales.	Valuation independent of margin/accounting — useful when earnings are noisy or negative.	Lower cheaper; compare within a sector only.
EV/EBIT	Enterprise value ÷ operating profit.	Capital-structure-neutral — counts debt, so a leveraged and an unleveraged firm compare fairly.	Lower cheaper. Our EV cheapness factor (keyless EV/EBITDA proper is not exposed).
P/FCF	Market cap ÷ free cash flow.	Free cash flow is harder to fake than earnings — the cash actually left after capex.	Lower cheaper; negative FCF = excluded from the FCF factor.
PEG	PE ÷ earnings-growth %.	Cheapness adjusted for growth — a high PE can be fair if profits are compounding.	<1 = cheap for the growth. Needs positive growth + positive PE to be valid.
FCF yield	Free cash flow ÷ market cap.	The inverse of P/FCF — the cash return on today’s price.	Higher = more cash per rupee invested. >5% is strong.
Div yield	Dividend ÷ price.	Cash paid out. DISPLAY-ONLY here — dividends are slab-taxed, so high yield is tax-inefficient for this strategy.	Higher pays more, but drags after-tax return; kept out of the core rank.
ROE	Net profit ÷ shareholder equity.	How well the firm turns retained capital into profit — the compounding engine.	>15% good, >20% excellent. A quality gate, not a value factor.
ROCE	Operating profit ÷ capital employed.	Return on ALL capital (debt + equity) — harder to flatter with leverage than ROE.	>15% good. Should exceed cost of capital.
D/E	Total debt ÷ equity.	Leverage. High debt magnifies both returns and the risk of a margin-call spiral under MTF.	<1 comfortable; >2 risky, especially leveraged buy-and-hold.
Piotroski F	9-point fundamental-health score (see below).	One number for improving profitability, leverage and efficiency — screens value-traps.	≥6 healthy, ≤3 weak.
Beta	Sensitivity to the index.	High beta = larger drawdowns = higher margin-call risk when you’re leveraged.	<1 defensive, >1.2 we exclude from the MTF preset.
1Y return	Price change over the last year.	The gross number the after-tax helper starts from.	Context only — past return is not future return.

Piotroski F-score — 9 points

One point per test that passes, comparing this financial year to the last. It screens value-traps: a cheap stock with a rising F-score is improving; a cheap stock with a falling one is often cheap for a reason. A test whose data is missing is skipped, never faked — the row also reports how many of the 9 were computable.

Group	Test	Passes when
Profitability	ROA positive	Net income > 0
Profitability	CFO positive	Operating cash flow > 0
Profitability	ROA rising	ROA this year > last year
Profitability	Accruals	CFO > net income (earnings backed by cash)
Leverage / liquidity	Lower leverage	Long-term-debt ratio down YoY
Leverage / liquidity	Current ratio up	Current assets / liabilities improved
Leverage / liquidity	No dilution	Shares outstanding did not rise
Efficiency	Margin up	Gross (or operating) margin rose YoY
Efficiency	Turnover up	Asset turnover (sales / assets) rose YoY

Value vs quality

Value is cheapness (the composite). Quality is business strength — ROE, ROCE, Piotroski, low debt. They’re kept separate on purpose:

The primary rank is pure value, so genuinely cheap names surface even if unglamorous.
A Quality flag (ROE>15% & ROCE>15% & Piotroski≥6 & D/E<1.5) marks the strong businesses.
Cheap and quality = the sweet spot; cheap without quality = possible value-trap. The Quality-Value preset intersects both (value ≥70th pct, ROE>15%, D/E<1).

The MTF 1-year buy-and-hold strategy

MTF (Margin Trading Facility) lets you hold a leveraged long position and pay interest (~12%/yr assumed here) rather than square off intraday. This screener is tuned for buying quality-value on margin, holding just over 12 months, then rebalancing annually. Why the 12-month hold and annual churn:

Tax — a hold past 12 months converts the gain from STCG (20%) to LTCG (12.5%). The annual churn is deliberately LTCG-optimal.
Interest drag — you must clear ~12%/yr just to break even on the borrowed portion. The after-tax 1Y column subtracts it: 1Y return − MTF interest − LTCG on the gain. A name that doesn’t clear interest is a losing leveraged trade even if the stock rose.
Margin-call risk — leverage magnifies drawdowns; a sharp fall can force liquidation at the worst time. Hence the preset caps beta ≤ 1.2 and tilts to quality.
Value-trap risk — cheap + weak fundamentals can stay cheap for years while interest accrues. The quality tilt + Piotroski guard against it.

The MTF Buy-Hold 1yr preset (flagship) restricts to MTF-eligible, β ≤ 1.2 names and ranks by after-tax, after-interest 1Y return — not gross return.

Tax rules (India equity) — lowest-tax by design

This strategy targets the lowest tax drag. The rules baked into the after-tax column and the presets:

LTCG vs STCG — hold >12 months → long-term capital gains at 12.5%. Sell inside 12 months → short-term at 20%. Holding past a year nearly halves the tax; the annual churn is timed for it.
₹1.25L LTCG exemption/yr — the first ₹1.25 lakh of long-term gains each year is tax-free. Harvesting gains up to the threshold realises them at 0% tax.
Dividend drag — dividends are taxed at your slab rate (up to ~30%+). For a high earner, a high-dividend stock is tax-inefficient, so div-yield is display-only, kept out of the core rank; high-div names are flagged as tax-inefficient for this strategy.
MTF interest deductibility — interest paid on the margin facility may offset gains under a business-income treatment. Informational only, depends on how you file — not tax advice.
Churn after 12 months — rebalance just past the one-year mark so every realised gain is LTCG, not STCG.
Tax-loss harvesting — before the annual churn, book losers to offset the winners’ gains and cut the net taxable amount.

Rates reflect the current India LTCG/STCG regime for listed equity. Verify against the latest Finance Act and your own situation before acting.

Disclaimer

Not investment or tax advice. This is an educational screen, not a recommendation. Low multiples don’t make a stock cheap in reality; a high composite is a starting point for research, not a buy signal. PSUs are cyclical and policy-exposed — they dominate low-PE screens for reasons that don’t always reverse. Leverage (MTF) can lose more than your capital. Data is best-effort from keyless public sources and can be stale, partial, or wrong. Do your own research; consult a registered adviser and a tax professional.