Think you're right about a stock? Let's check.

Write your view on any Nifty 50 stock in three sentences. We test each one against real market data, tell you which parts hold up, and show exactly how often our own models are wrong.

Test your thesis See our track record

Free · No sign-up · Never tells you to buy or sell

HDFC Bank

Financials · close of 14 Oct 2026

₹711

Example thesis

HDFC Bank's margins recover as deposit costs ease after the merger.
It's a steady, defensive holding while the market wobbles.
The stock is in an uptrend and keeps grinding higher.

See why each one got its verdict

the full Nifty 50: 50 stocksthe full Nifty 50
of daily prices: 12+ yearsof daily prices
each tested against a simple baseline: 3 modelseach tested against a simple baseline
of out-of-sample testing: 8 yearsof out-of-sample testing

How it works

Three steps, about thirty seconds.

1. Write your view
Pick a Nifty 50 stock and give three short reasons. Plain English is fine.
The stock is in an uptrend and keeps grinding higher.
2. We test each sentence
Each one is checked against the data that can actually test it, and leans for or against.
Price trendUndermined
AgainstFor
3. Know what would prove you wrong
For every sentence you get the exact thing to watch, so you know when to rethink.
You're wrong if: A close below the 200-day average, ₹797.

Other apps show you a stock. We test your view of it.

Most stock apps tell you what a price did and leave the thinking to you. Prove Me Wrong starts with what you believe and checks it, sentence by sentence.

Feature	Typical stock app	Prove Me Wrong
Starts from	The stock's price	Your own view
Predictions	Targets and “buy” calls	Ranges, never tips
Model accuracy	Rarely shown	Published, losses included
When the data can't tell	Guesses anyway	Says “open” and why

We publish our own report card

Every model was tested on years it never saw, against a simple rule anyone could use. Here's how they did, including where they lost.

Full scoreboard

Lost
Calling up or down
Right 55% of the time, vs 57% for just assuming “up”. So we never make these calls.
Works
Forecasting how bumpy a stock will be
GARCH(1,1) cuts forecast error by 28% versus the simple rule.
Works
Reading the market mood
“Stressed” days were followed by 1.9× the swings of “Calm” days.
Marginal
Likely price range
Only 1% better than the simple rule, and its 80% range caught 75% of outcomes.

Biggest movers this month

Have a view on one of these? Test it.

Questions

Is this investment advice?

No. Prove Me Wrong analyses; it never tells you to buy, sell or hold, and gives no price targets. It is not registered with SEBI as a research analyst or adviser.

Where does the data come from?

Daily closing prices for the Nifty 50 and the index, updated end of day. Every model is re-tested on years it never saw before its numbers appear here.

Why do some sentences come back “open”?

Some claims, like “margins will recover”, need company filings and earnings calls that the price models can't read. Rather than guess, we say the question is open.

Is it free?

Yes. No sign-up, no account. It's a portfolio project built to show honest machine learning.

Got a view on a stock?

Find out which parts of it the data agrees with, in about thirty seconds.

Test your thesis