The only screener that proves every number

Look up any stock. Every number verified.

Fundamentals, technicals, sentiment and macro for all 500 NIFTY-500 stocks. Every figure is computed from the actual filing and carries its proof. It explains stocks. It never tells you to buy them.

Free tier, full terminal. No card. · Not investment advice.

The whole product in 52 seconds. Sound on.

today · from the dataset

What the data shows right now.

500 stocks · … verified P/E · prices as of …

1
The filing
Every input is read from the actual BSE/NSE document, with its period and basis.
2
Deterministic compute
Code computes every figure. The AI only asks questions and explains answers.
3
Self-audit
Figures are cross-checked against the filing and price; failures are withheld, with the reason.
4
The receipt
Every answer ends with the exact query, parameters, source and timestamp that produced it.
01 · the problem

Don't trust numbers from a screener you didn't compute yourself.

Indian retail investors live between two bad options: tools that hand you a number with no proof, and chatbots that invent the proof. Both ask for the one thing you can't give them, blind faith.

opaque numbers

"Why are the numbers different?"

Standalone or consolidated? Trailing or FY? Static screeners give you a single figure and no way to reconcile it against the filing. You're left guessing whose math it is.

no AI you can trust

The chatbot just makes it up.

Ask a general-purpose AI for a P/E and it will confidently hallucinate one, wrong by a factor of ten, with no source and no way to check. Plausible is not the same as correct.

the paywall

₹4,999 to see one ratio.

The genuinely useful features (backtests, exports, full history) sit behind annual plans that price out exactly the people learning to invest. Research shouldn't be a luxury good.

02 · the wedge

Every number
opens up.

Click any figure and the terminal unfolds the arithmetic that produced it (the inputs, the operation, the answer), each input tagged with the exact filing it came from. Nothing is computed off-screen.

  • Deterministic, not generative. The same query gives the same answer, every time. Math, not vibes.
  • Sourced to the filing. Every input links to the BSE/NSE document and period it was read from.
  • You pick the basis. Standalone vs consolidated, TTM vs FY. Switch it and watch the math recompute live.
R Reliance Industries NSE: RELIANCE
P/E drill-down
Market cap ÷ PAT
P/E= Market cap ÷ PAT
= ₹16,83,304cr ÷ ₹95,754cr
= 17.6×
Market cap
₹16,83,304cr
NSE close × shares o/s
PAT
₹95,754cr
consolidated P&L · FY26
source: BSE/NSE filing · FY26 recomputed live
03 · the receipt

Every backtest comes with a receipt.

Every figure comes with an auditable trace: the exact query, the parameters, the data source, the result, and what it doesn't account for yet. If you can't reproduce it, it isn't research, it's a rumour.

Backtest · golden cross trace · bt_2204
backtest a golden cross on Infosys from 2022
Strategy return
−6.27%
Benchmark
−14.76%
signalgolden_cross
universeNIFTY-500
datacached EOD close · point-in-time
window2022 → 2026
result−6.27% vs −14.76% bench

Methodology honesty

The receipt also says what the number doesn't include. We tell you the strategy used point-in-time prices and modeled costs (and that survivorship isn't adjusted for yet) rather than print a glossy number you can't trust.

point-in-time prices · costs modeled · not survivorship-adjusted yet

printed on every backtest

Reproducible by design

Replay the same query tomorrow and get the identical result. The data is point-in-time, so nothing drifts under you. The receipt is the trust artifact: it's the difference between an answer and a claim.

Data integrity alert auto-flagged
Persistent Systems
NSE: PERSISTENT · FY26
flagged: verify
Reported revenue (filing parse)
₹1,47,484cr true ≈ ₹14,748cr
~10× too high, a units error in the source. Shown as verify, not fact.
checkvs sector + own history
actionheld back · not shown as fact
statusflagged verify · re-parse queued
04 · self-auditing

The only screener that verifies its own numbers.

Every figure is cross-checked against the filing and the traded price before you see it. That guard is how 101Wealth caught a feed showing PERSISTENT revenue of ₹1,47,484cr (a units slip, 10× the true figure near ₹14,748cr) and held it for verification instead of serving it as fact. Accuracy here is enforced, not assumed.

Most tools assume their data is clean and pass the error straight to you. We assume ours might be wrong, and prove that we checked, in front of you. No other screener audits its own numbers like this.

Research

Does privatisation work?

Haryana's power workers struck over a private licence in Gurugram. We went through every major Indian electricity privatisation and four strategic sales, before and after, with the consumer's bill in the middle of the table: Delhi, Odisha twice, Mumbai, Chandigarh, 28 franchisees, and the stakes the government kept in Hindustan Zinc, VSNL, Maruti and BALCO.

Seventeen exhibits, every source named, and a view: losses fell wherever the contract made them fall, consumers got supply and a deferred bill, and the stakes the government kept were the best thing it ever sold. Our previous reports, Do markets care about democracy? and The de-rating of Indian markets, are still free to read.

Read the research

Summary free · full report with a name and email · not investment advice

50.7% → 5.4%
Delhi's distribution losses, the regulator's figure in February 2002 against Tata Power-DDL in FY2025-26.
₹38,552 crore
The regulatory assets APTEL ordered Delhi's consumers to repay over seven years, in April 2026. About ₹55,000 a connection.
₹98,839 crore
What the 29.5% of Hindustan Zinc the government kept has produced, against ₹769 crore for the 44.9% it sold in 2002-03.
12 of 28
Distribution franchisees still operating, of all appointed since 2006. The rest were terminated or never started.

Data to 23 Oct 2026 · regulator orders, company filings, CSEP, Prayas and the Regulatory Assistance Project · prices and dividends from Yahoo Finance

the lab · research preview

We also let the AI
run a portfolio. In public.

Since June 2026 a team of Claude agents (three analysts, a portfolio manager and a macro strategist) has managed a ₹10 lakh paper portfolio of NIFTY-500 stocks, on our own capital. Every evening it reads the filings and the news, decides, and logs why.

  • Honest accounting. Time-weighted returns, costs and slippage on, deposits never counted as profit.
  • Published when it loses. The scoreboard is live and unedited. Right now it trails the index.
  • Not advice, not a signal service. It manages our money, not yours. The point is to learn what AI research is worth, in rupees.
AI-WM paper portfolio since 4 Jun 2026
live
Portfolio · TWR
-5.2%
NIFTY 500 · same window
-1.3%
Difference
-3.9 pp
Holdings
37 · 10 sectors
Cash
16.3%
Value
₹9,49,434
Last run
6 Oct 2026
time-weighted · costs on paper fills at next open ₹10,00,000 start

Research experiment on our own capital. Not investment advice; not open to outside money. Numbers refresh after each evening run.

Get the research when it publishes.

Leave your name and email and every 101Wealth report lands with you on the day it publishes, with the data-quality findings behind it. No spam, no card.

Free · not investment advice