SectorScope

1 · Quick start

This tool answers one question: which Indian sectors are attractively priced right now, relative to their own history? It doesn't pick stocks — it tells you whether an entire sector is trading cheap or expensive compared to what it has historically traded at.

Each sector gets a card. Green = cheap vs history. Red = expensive vs history. Cards are sorted cheapest-first.

Three-step workflow

1

Scan the Dashboard. Green cards at the top are candidates. Red cards at the bottom are sectors to be cautious on — or to consider reducing. Look for green cards where the EPS number is also healthy (not declining).

2

Tap any card to flip it. The back shows the full valuation breakdown: current P/E or P/B, the weighted historical reference, and how today compares to the 3yr / 5yr / 10yr medians side-by-side.

3

Verify on the Backtest page. For any sector that looks interesting, go to Backtest → pick the sector → Run. You'll see every past time the sector was at this valuation level and what happened to prices over the next 6 months and 1 year.

One thing to keep in mind before you start

Cheap can get cheaper. A "Very Cheap" signal tells you the sector is unusually depressed vs its history. It is not a prediction of exactly when it will recover. These signals work best for patient investors with a 6–12 month horizon, not for next-week timing.

2 · Reading a card

Each card has two sides. Here's an annotated example of Nifty Auto — every element explained.

Card front (numbers at a glance)

① sector name

Auto

② band · ③ signals

Cheap

P18 · 82% of ref

④ caution capsule (when shown)

⚠ often this cheap

⑤ metric tiles

P / E

24.5

P18 · 82% of ref

EPS

312

87% of peak
▲ +14.2% YoY

⑥ sparkline (3yr P/E + median)

Analysis ›

①

Sector nameWhich Nifty sectoral index this card represents. The coloured top border and card background tint both match the band verdict.

②

Band verdictThe headline signal — Very Cheap / Cheap / Neutral / Expensive / Very Exp. Driven by how today's P/E (or P/B) compares to its weighted historical median. See section 3.

③

Two signals: P18 · 82% of refP18 = 18th percentile of its own last 5 years (cheap end). 82% of ref = current P/E is 18% below its historical anchor. The second number drives the band. See section 3 for why.

④

Caution capsule (only appears when signals disagree)"often this cheap" means: yes, the sector is 18% below its anchor — but it has also spent a lot of the last 5 years at this level. Less contrarian than it looks. If there's no capsule, both signals agree.

⑤

Metric tilesLeft tile (tinted with band color): the current P/E or P/B. Right tile: current EPS, % of its all-time peak, and year-on-year growth. See section 4.

⑥

SparklineThree years of daily P/E (or P/B) as a line, with the 3yr median as a dashed horizontal. A quick visual to spot whether we're at the top or bottom of the recent range. See section 5.

The card back (tap to flip)

Flipping the card reveals the detailed breakdown: the exact weighted reference, and how today's value compares to the 3yr, 5yr, and 10yr medians individually — each shown as a multiple (e.g. 0.82× means current P/E is 82% of that window's median, or 18% below it).

Which metric does each sector use?

The tool picks the right metric automatically — you don't need to change anything.

P/E only: IT, Auto, FMCG, Pharma — stable, earnings-driven businesses where P/E is the primary lens.
P/B only: Bank, PSU Bank — for banks, earnings are distorted by provisioning cycles, so book value is the more reliable anchor.
Both P/E + P/B: Infra, Energy, Realty, Metals — asset-heavy cyclicals where you want both an earnings lens and an asset-value lens.

💡

Try tapping a green card on the Dashboard now. On the back, find the 3yr / 5yr / 10yr rows. If the 5yr median is significantly higher than the current value, that's the core signal.

The Detail overlay — deeper data for any sector

At the bottom-left of every card front there's a small button that looks like this:

Detail

Tapping it opens a full-screen overlay with more detail than the card can fit. There are three sections inside it. Here's what each one contains and how to read it.

① Current Snapshot

Four stat cards showing today's key numbers at a glance, each colored by its own band verdict:

P / E

24.5

Cheap

P / B

3.12

Neutral

EPS

312

+14.2% YoY

Max EPS (all-time)

358

87% of peak

P/E and P/B show their individual band labels — useful when a dual-metric sector has one expensive and one cheap, which the card-level verdict averages over.

② Historical Medians

One block per metric. Each row is a time window — the bar length shows its median relative to the others, and the right column shows current ÷ that median as a multiplier.

P/E Medians Weighted ref: 25.2

3-Year

26.2

0.93×

5-Year

25.1

0.97×

10-Year

22.8

1.07×

Current

24.5

1.00×

Read the multiplier column as: below 1.00× = today is cheaper than that window's median. Above 1.00× = today is more expensive. The 10yr showing 1.07× in the example means today is slightly above the 10yr average — but below the 3yr and 5yr, which is what matters most for the signal.

③ EPS Growth

Two bar charts showing how the sector's earnings have evolved — one at annual resolution, one at quarterly resolution. Green bars = positive growth, red bars = contraction. Use these to sanity-check the valuation signal: a cheap sector with accelerating EPS is far more interesting than a cheap sector with collapsing EPS.

EPS YoY Growth · annual · last 10 yrs

EPS QoQ Growth · last 4 quarters

Two chart types · two different questions

Annual YoY: Each bar = EPS today vs EPS 1 year earlier, stepping back 10 years. Answers "is this sector's earnings trajectory fundamentally up or down across cycles?"
Quarterly QoQ: Each bar = that quarter's average EPS vs the prior quarter. Answers "is earnings momentum right now accelerating, flat, or turning negative?" A run of green QoQ bars after a red YoY is a classic "earnings turning" setup.

Y-axis auto-scales to the biggest absolute value in the chart — so a quiet sector with ±8% moves shows the same visual weight as a volatile one swinging ±40%. Compare shape and sign, not bar heights across two sectors. Each bar also has its exact percentage printed above (or below, for negatives).

④ Quarterly Averages · Last 5 Years

A scrollable table showing quarterly average Price, P/E, P/B, and EPS going back 5 years. Most recent quarter is at the top. Two cells in each column are highlighted:

Quarter	Price	P/E	P/B	EPS
Q1 2025	22,840	24.5	3.12	312
Q4 2024	24,100	26.8	3.35	299
Q3 2024	25,600	28.1	3.58	287
Q2 2024	23,900	27.4	3.41	275
Q1 2024	21,200	27.0	3.22	265
…	…	…	…	…
Q1 2020	11,400	18.2	2.20	198

Green= best quarter in that column

Red= worst quarter in that column

"Best" and "worst" are inverted for valuation metrics: the lowest P/E and P/B quarter is green (cheapest = good), the highest is red. For Price and EPS it's the other way: the highest price and highest EPS quarter are green. Only the single best and worst quarter across the whole 5-year history gets highlighted — all others are neutral.

Best use of the Detail overlay

The overlay is most useful for dual sectors (Infra, Energy, Realty, Metals) where P/E and P/B can disagree. The card front shows a blended verdict — the overlay's Snapshot section shows the two metrics separately, so you can see if the cheapness is driven by one metric or both. The quarterly table is useful for spotting seasonal patterns — e.g. Auto tends to see PE compression every Q2 ahead of new model launches, which can look like a cheap signal but is cyclically normal.

Close the overlay with the ✕ button in the top-right corner, or press Escape on a keyboard.

💡

Try tapping a green card on the Dashboard now. On the back, find the 3yr / 5yr / 10yr rows. If the 5yr median is significantly higher than the current value, that's the core signal.

3 · Valuation bands & signals

The band (card color) is determined by one number: current P/E ÷ weighted historical reference. Call this the ratio. Here's how each ratio maps to a verdict:

Very Cheap≤ 80%

Cheap80–90%

Neutral90–105%

Expensive105–120%

Very Exp.> 120%

The Neutral band is slightly wider on the right side (up to 105% instead of a symmetric 100%) because sectors naturally drift slightly above their own median over long time periods. Being 5% above your median is not yet a warning.

Worked example — how the band is calculated

Nifty Auto's P/E today is 24.5. Its historical medians over different windows are:

Weighted Reference · Nifty Auto (example)

3-yr median

× 30%

26.2 → 7.86

5-yr median

× 60%

25.1 → 15.06

10-yr median

× 10%

22.8 → 2.28

Weighted reference = 7.86 + 15.06 + 2.28 = 25.2

Ratio = 24.5 ÷ 25.2 = 97% → Neutral

If the same sector had a P/E of 20.7 instead (say, after a sharp correction), the ratio would be 20.7 ÷ 25.2 = 82% → Cheap. That's the number you see on the card as 82% of ref.

Why 60% weight on the 5-year median?

The 5-year window covers roughly one full business cycle in India — close enough to reflect today's regime, long enough to include both peaks and troughs. The 10-year leg gets only 10% weight so that sectors that have permanently re-rated (like IT after 2020, or PSU Banks after 2022) aren't forever compared against a baseline that no longer applies.

The two signals on each card

Each card shows two valuation signals side-by-side, like P18 · 82% of ref. They measure different things and can sometimes disagree.

% of ref (primary)

What it asksHow far is today from the anchor?

Example82% of ref

Meaning18% below the blended median

Drives band?Yes — this sets the color

Measures magnitude — by how much is it cheap?

Percentile rank (secondary)

What it asksHow rare is this level, last 5 years?

ExampleP18

MeaningOnly 18% of last-5yr readings were lower

Drives band?No — info only

Measures frequency — how often does it get this cheap?

When the two signals disagree — the caution capsule

Sometimes a sector looks cheap vs its anchor (% of ref says Cheap) but is not unusual at all in its recent 5-year distribution (percentile says P45 — it's been here half the time). That's a much weaker signal than a sector that is both below its anchor and historically unusual.

Clean signal — no capsule

% of ref82% → Cheap

PercentileP14 → also Cheap

✓ Both signals agree. Sector is genuinely depressed — it's both below its anchor AND near a 5-year low.

Conflicted — capsule appears

% of ref83% → Cheap

PercentileP44 → Neutral

⚠ "often this cheap" — it IS below the anchor, but the sector has spent nearly half the last 5 years at this level. Less contrarian than the Cheap label suggests.

4 · EPS metrics

A sector can look cheap for two very different reasons: (A) price has corrected but earnings are healthy — a genuine opportunity. (B) earnings are collapsing and price is following — a value trap. The three EPS indicators help you tell them apart.

The three indicators

% of peak: Current EPS divided by the sector's all-time high EPS. 87% of peak means earnings are 13% below their historical best, with room to recover. Above 100% = earnings at a new high.
YoY growth: Current EPS vs EPS one year ago. Color-coded: red below 0%, amber 0–10%, green above 10%.
▲ TURNING: An amber pill that appears when EPS is rising over two consecutive ~90-day windows — i.e. today > 90 days ago > 180 days ago. This is an early-cycle inflection signal, often firing before YoY goes positive.

Three scenarios — what to look for

Best setup

Valuation bandVery Cheap

% of peak EPS68%

YoY growth−6%

Turning flag▲ TURNING

Early-cycle turnaroundEarnings are still negative year-over-year — but they've stopped falling and are inflecting. Sector is cheap. This is the highest-conviction setup. Classic Metals / PSU Bank early-cycle entry.

Watch out

Valuation bandCheap

% of peak EPS94%

YoY growth+12%

Turning flag—

Confirmed recovery, less upsideEarnings are healthy and growing. Sector is cheap — but since recovery is already confirmed, the market has partially priced it. Still a good setup, just less asymmetric than the one above.

Value trap

Valuation bandVery Cheap

% of peak EPS41%

YoY growth−28%

Turning flag—

Cheap for a reason — waitEarnings are collapsing. P/E looks low because E is falling fast. The sector will appear cheap and get cheaper as long as earnings deteriorate. Do not act until either TURNING fires or YoY stabilises.

The most powerful combination

Cheap/Very Cheap band + TURNING flag + YoY still negative. This fires at the exact inflection point, before the crowd notices. It works best on cyclicals: Metals, Auto, PSU Bank, Realty. For defensive sectors (FMCG, Pharma, IT), the turning signal matters less — earnings are steadier there and the valuation band alone is more reliable.

5 · Sparkline chart

The small chart on each card front shows three years of daily P/E (or P/B). Three things to read from it:

Annotated sparkline — Nifty Auto (example)

P/E line — colored to match the sector. Watch its trend direction over the 3 years.

EPS line (P/E cards only) — on its own scale. When P/E falls but EPS rises, it means the sector is getting cheaper for the right reason: price fell while earnings grew.

3yr median — the dashed line. The card's primary signal compares current P/E to the blended reference (not just the 3yr), but the dashed line gives you a fast visual anchor. Current below the dashes = cheap zone.

The shaded area under the main line is purely visual emphasis — it has no separate numeric meaning.

6 · Backtest page

The Backtest page answers: "Every time this sector was at this valuation level in the past — what happened to prices next?" It runs through history, finds every matching date, and shows you the forward returns.

The key connection to the dashboard

The threshold slider on the Backtest page maps directly to the % of ref badge on a dashboard card. If a card shows 82% of ref today, set the threshold to 82% on the Backtest page and ask: "the last time this sector was at 82% of its reference, what happened over the next year?" — that's your historical precedent for today's signal.

How signals fire — visual

Backtest timeline — when signals fire and what happens next (illustrative)

Each green circle is a signal date — a date when the ratio crossed your threshold. The numbers show what actually happened to the sector's price over the next 12 months from that date. The backtest table shows all of these with full detail.

Setting up a backtest — step by step

1

Pick your sector. Data loads automatically. The P/E and P/B checkboxes pre-tick based on what's appropriate for that sector — but you can change them freely.

2

Set the date range. Defaults to full history. Narrow it to test a specific period, e.g. 2018–today to exclude the pre-GST era for FMCG, or post-2022 for PSU Banks after their re-rating.

3

Set the threshold. This is the key input. A threshold of 80% means "fire a signal on every date when the ratio was at 80% of its reference." Start with whatever % of ref the dashboard shows today — that's the most relevant comparison.

4

Hit Run. The results show every past signal date, the valuation at that date, and the 6-month and 1-year annualised return from that point forward.

Reading the results table

Weighted ref banner: Grey strip above the stats. Confirms the formula used (30/60/10), the ±1% trigger tolerance, and shows the weighted reference as of the end of your selected window.
Signals count: How many matching dates the engine found. Very low counts (1–3) should be treated as anecdote, not statistics.
Avg 6m / 1yr CAGR: Annualised return averaged across all signal dates, at 6-month and 1-year horizons. Compare to Nifty 50's typical annual return (~12–14%) as a mental benchmark.
Hit rate: % of signals that were in profit at that horizon. Hit rate above 70% is strong. Below 50% means more losers than winners — inspect the EPS % column to see if value traps are responsible.
EPS % column: EPS as a % of the sector's max-ever EPS at that signal date. Low values (below 50%) flag potential value traps — the sector was cheap because earnings had collapsed. High values (above 85%) mean the cheapness was price-driven, not earnings-driven — usually higher quality signals.

Row shading — what the colors mean

🟢

Green rowEither the 6m or 1yr CAGR was above +15%. This was a strong outcome.

🟡

Yellow rowPositive returns but not exceptional — a win but modest. Still a hit.

🔴

Red rowEither the 6m or 1yr CAGR was negative. Check the EPS % column — low values here usually explain the loss.

Quick-reference: what to look for

Interpreting your backtest result

Strong setup: Hit rate above 70%, average 1yr CAGR clearly positive, most rows green. Proceed with conviction.
Noisy setup: High average return but low hit rate — a few huge green rows masking several red ones. A small number of outlier wins. Fragile; wait for an additional catalyst.
Broken setup: Mostly red rows, poor hit rate, low average return. Either the threshold is wrong for this sector, or the signals are clustering around earnings collapses. Check the EPS % column.
Too few signals: 1–3 hits is not a pattern — it's a handful of data points. Don't draw strong conclusions. Widen the threshold slightly or extend the date range.

How to use the EPS % column to filter

Uncheck the red rows where EPS % was very low (below 50%), then re-check the summary stats. If the hit rate and average return improve significantly when you exclude low-EPS signals, that tells you: this sector's cheap-valuation signals only work when earnings are healthy. That's important context for acting on today's signal — check the current EPS tile before committing.

7 · Limits & caveats

This tool is designed to sharpen decisions, not replace judgement. Here are the honest limits:

Valuation is not timing. A sector in the "Very Cheap" band can stay cheap — or get cheaper — for months before reversing. These signals are for patient capital with a 6–12 month horizon, not for predicting next week's move.
Index composition changes over time. Nifty sectoral indices reconstitute periodically. The 10-year P/E series for Nifty IT in 2025 includes different companies than it did in 2014. Very long histories blend old and new compositions — treat them as approximate, not precise.
No macro overlay. The model knows nothing about RBI rate cycles, the INR, global commodity prices, or credit cycles. A "Cheap" PSU Bank signal during a credit crisis is a different risk than a "Cheap" signal in a stable macro environment. Add your own macro read.
EPS can be distorted. The EPS series uses NSE's reported TTM earnings methodology. One-off asset sales, accounting changes, or a single large provisioning quarter can make EPS look unusually high or low for a quarter or two. Cross-check with reported quarterly results if EPS looks surprising.
Structural regime changes. The model puts only 10% weight on the 10-year median, so it adapts gradually to re-ratings. But if a sector undergoes a fundamental structural change — like PSU Banks after IBC/asset-quality clean-up — the new regime won't fully dominate the reference for another few years. During a transition, treat the signal as directional rather than precise.

Best practice

Use this as one structured input among several. The signals are most powerful when: (1) the valuation band is Very Cheap or Cheap, (2) the EPS tile shows a TURNING flag or healthy YoY growth, (3) the backtest shows a strong historical hit rate at this threshold, and (4) your own fundamental/macro view doesn't contradict the signal. All four together is a high-conviction setup.

India Equity · NSE sector indices · Valuations based on trailing TTM earnings & book value.