Our Methodology
Transparency is core to our values. Here's exactly how our AI predictions work and where our data comes from.
Data Sources
Our predictions are only as good as our data. We aggregate information from multiple verified sources to ensure accuracy and reliability.
BSE/NSE Official Data
Subscription status, allotment ratios, and listing prices directly from exchanges
Registrar Data
Real-time subscription updates from Link Intime, KFin Tech, and other registrars
Grey Market Sources
Verified GMP data aggregated from multiple dealer networks across India
News & Sentiment
AI-processed news articles, expert opinions, and social media sentiment
AI Prediction Model
Our machine learning model is trained on 5+ years of historical IPO data, analyzing patterns across 500+ IPOs to predict listing gains. The model uses an ensemble approach combining gradient boosting and neural networks.
Prediction Factors
Overall, QIB, NII, and Retail subscription levels
Grey market premium trajectory in the days before listing
Nifty/Sensex trend, sector performance, and volatility
Company financials, valuations, and peer comparison
Similar IPO performance in comparable market conditions
Expert reviews, news sentiment, and social media analysis
Market Sentiment Score
Our sentiment score ranges from -100 (extremely bearish) to +100 (extremely bullish). It's calculated by aggregating:
- Expert reviews from YouTubers, analysts, and brokerage firms (weighted by historical accuracy)
- GMP trend direction and momentum over the subscription period
- Subscription pattern analysis (early vs late subscription, category-wise trends)
- News sentiment analysis using NLP on IPO-related articles
Accuracy Tracking & Accountability
We believe in complete transparency. Unlike most prediction sites that hide behind vague claims, we track every prediction against actual listing-day gains and publish the full dataset on our Accuracy page. No cherry-picking, no survivor bias - just raw numbers.
Our AI is scored against the grey market premium quoted on the IPO close date - the number most retail investors actually see. This is the fairest comparison because it's what traders used to make their decisions before listing.
We track mainboard and SME listings across the dataset. Error is measured as absolute deviation from the actual listing-day close. Hit rate is ±5% from actual gain. Direction accuracy measures whether we correctly called gain vs loss.
Rather than claiming fixed accuracy percentages, we publish live dashboards showing year-over-year performance, recent head-to-head comparisons, and the full prediction log. Your investment decisions deserve better than marketing numbers.
Important Limitations
- Our predictions are probabilistic estimates, not guarantees. Past accuracy does not guarantee future results.
- Grey market premiums are unofficial and can be manipulated. We aggregate multiple sources to reduce noise.
- Extreme market events (crashes, circuit breakers) can invalidate predictions.
- SME IPOs have higher volatility and lower prediction accuracy compared to mainboard IPOs.
- This is NOT investment advice. Always do your own research and consult a SEBI-registered advisor.