Reading the results responsibly

Algo Py is an educational and screening tool. Nothing on it is a recommendation to buy or sell any security. We are not a SEBI-registered investment adviser or research analyst. Any decision you make is your own.

When a strategy flags a stock, it means the stock matched a rule — a breakout, an uptrend template, a fall from its high. That is a reason to research further, not a reason to buy. Markets are driven by news, results, liquidity and sentiment that a price rule cannot see.

Before acting on anything here, check the broader market and sector, the company's fundamentals and shareholding, recent news and events, and the valuation. The Trade Plan page lists exactly what to verify. Cross-check on screener.in and the company's own filings.

• Never risk money you cannot afford to lose.
• Decide your stop-loss before you enter, and respect it.
• Size positions so a single wrong trade cannot hurt you badly.
• Past performance and back-tested rules do not guarantee future results.

Prices and fundamentals come from third parties and can be delayed, adjusted differently, or occasionally incorrect. Treat every number as “approximately right, verify before you act”. See Data & third-party APIs.

The Astro Almanac exists to show you a mistake, not to help you trade. A 2013 book titled "Predicting the Nifty & Sensex" states about 1,700 rules of the form “when the Sun is in Gemini and the Moon is in Ardra, the market rises”, and claims 70% accuracy. Every input is astronomy, so every rule is checkable — so we checked all of them against 25 years of real data: the Nifty for the NSE page, the Sensex for the BSE page. (The book's title promises both, but every table it printed was the Nifty.)

They do not work — on either index. The market rises on about 53.8% of all days regardless of the sky. 65 of the book's rules look “statistically significant” — and about 80 is exactly what pure chance produces when you test that many. Not one survives a correction for the number of tests run, on the Nifty or the Sensex.

The book was published in 2013. On the years its author had never seen, its calls win 48.3% of the time — below both a coin flip and the market's own 53.8% up-day rate — and lose money. Trading it in the years he did study looks magnificent, which is the entire lesson: anything can be made to fit the past.

The page can never emit a buy or sell call, by design. It is the clearest example we have of the trap this whole site tries to avoid: run a thousand ideas, keep the winners, never mention how many you tried.

For advice tailored to your situation, consult a SEBI-registered investment adviser. When something sounds too good to be true — a guaranteed tip, a “sure” multibagger — it usually is.