Equity research · India & US
Signals, not trade tips.
A Stock Score for every listed name, ranked against its own sector — read before you plan capital, not orders to place.
Research, not advice · Not a broker
| Ticker | P/E | ROE | Score | Grade | |
|---|---|---|---|---|---|
| DEMO102 | Copperleaf Software | 35.4 | 34.6% | 100 | A |
| DEMO211 | Northwind Electronic | 23.7 | 22.5% | 100 | A |
| DEMO313 | Larkspur General | 13.5 | 15.0% | 100 | A |
| DEMO404 | Halcyon Banks | 16.5 | 6.6% | 100 | A |
| DEMO113 | Redfern Software | 22.9 | 18.9% | 92 | A |
| DEMO205 | Obsidian Electronic | 30.9 | 31.0% | 92 | A |
| DEMO311 | Juniper General | 14.1 | 8.0% | 92 | A |
| DEMO400 | Tamarind Banks | 19.7 | 18.5% | 92 | A |
Stock Score
Judge quality before capital.
Thirty fundamentals in five categories — valuation, profitability, growth, balance sheet and cash returns — each ranked against same-industry peers in the same market. A 30x P/E means something different for software than for a utility, so nothing is judged against a fixed bar.
EXAMPLE · US · Software & business services
Example Software Co. (fictional)
Stock Score 85 · 3rd of 14 in Services · Top fifth of its sector
- Price
- $121.50
- Market cap
- $46.9B
- Coverage
- 28/33 metrics
- Index
- —
- As of
- 2026-10-15
Valuation
Profitability
Growth
Balance sheet & risk
Cash flow & shareholder returns
Each gauge is a percentile against same-industry peers in the same market, falling back to the sector, then the market, when fewer than 10 peers have data. The grade ranks the weighted composite within the sector.
A fictional company scored by the production engine against a synthetic peer group. Not a read of any listed company.
Options Signals
Built to hedge, not to chase.
A derived view — stance, ATM/IV context, ranked structures, why — built for hedging equity risk, not for chasing direction. Directional and volatility ideas stay quiet unless the setup grades A or B. US markets today; no option chain, no order routing.
How it works
A Signal has to earn its place.
- 1
Collect point-in-time history
Scheduled jobs write India and US fundamentals and prices, plus US option implied volatility, into one historical lake, dated as they were known, so nothing leaks from the future.
- 2
Lock a Study, then test forward
Every idea becomes a Study with a fixed universe, method and window, corrected for testing hundreds of names at once. Then it has to hold on data it has never seen.
- 3
Change a Signal only when it survives
Signals stay simple and explainable: a Stock Score, or an Options Signal for hedging. A new method only reaches them after its Study holds up. Ideas that fail stay published, not hidden.
Studies
We publish what fails, too.
886 India and US names studied so far. Each Study is locked — universe, method, window — and published with its result, whether the idea held up or not.
- S-001Indicator promotion gatePassed93 of 886 names passed the gate after multiple-testing correction.
- S-002Locked forward test of promoted namesDid not hold0 of 93 names held their edge. The model trailed a naive baseline.
- S-003Shallow neural models vs technical rules, after costsMixedNeural models beat indicator rules on 447 of 498 names, but the best learned model beat simply holding on only 50 after costs.
Start reading Signals.
Free to create an account. Stock Score and Options Signals for India and US markets.