Analytical Methodology
How Rovana AI analyses every CSE stock — the three pillars that power the platform. All outputs are informational; nothing here constitutes investment advice.
Every Saturday, all 296 CSE stocks are analysed across nine quantitative dimensions. Each dimension captures a distinct aspect of a company's quality — from balance sheet strength to statistical price behaviour. The dimensions are combined into a single composite score, which maps to a five-tier output. No single dimension drives the result; the model is deliberately multi-factor to avoid single-metric distortions.
How efficiently the company generates returns from its assets and equity. A strong profitability dimension signals a business with durable competitive advantages.
Balance sheet resilience — debt levels, liquidity, and the ability to service obligations through economic cycles.
Whether the stock is attractively priced relative to its earnings, book value, and sector peers — identifying value before the market does.
Revenue, earnings, and free cash flow growth across multiple horizons — identifying compounding businesses with expanding economic moats.
How effectively leadership deploys capital. Great businesses earn superior returns on every rupee reinvested.
Price momentum and trend strength — not used as a standalone signal, but as confirmation that the fundamental picture is playing out in price action.
The relationship between volume and price movement. Institutional accumulation leaves footprints in volume data before it appears in price.
Exposure to the current CBSL macro environment. Rate-sensitive sectors behave very differently depending on the monetary policy cycle.
Incorporates the Granger causality verdict — whether a statistically significant volume→price lead relationship exists for this specific stock.
Composite Tier Output
Tiers are relative — APEX and CAUTION represent the top and bottom of the current universe, not absolute thresholds. Scores are recalculated every Saturday and reflect the most recent publicly available data.