🪐 MARS V41.8.5 · Manish's Analytics & Research System

Four institutional-grade quant modules. All compute-only — zero new external scrapers, zero new dependencies. Each module is independently cached (5-30 min TTL) and gracefully degrades on partial upstream data.

Select a window and tap "Load from NSE Official" to fetch upcoming corporate actions for that period directly from NSE.

V23 Hardened cascade · cookie warm-up · UA rotation · NSE→BSE→Screener fallback

Plan a trade before you place it. The journal locks your thesis, stop, target, holding period, and invalidation event. It then alerts you when your stop or targets hit and tracks whether you actually honored your own plan.

Once locked, the plan is immutable in spirit. You can edit notes/targets, but the journal records every change so you can audit your own discipline.

Column guide: P&L = profit/loss in ₹ · R = risk-multiple (P&L ÷ planned-risk; e.g. +2R means won 2× what you risked, +0.00R = no stop was set so R can't be computed) · Days = holding duration · Exit: manual_target = hit your profit target, manual_stop = hit your stop-loss, imported = from broker, exit reason unknown

Most retail apps show your trades. Pro tools show your stats. This shows your psychology vs outcomes — were you actually right when you said "80% confident"? Are you adding to losers? Holding past targets? It learns from your closed trades only.

A perfectly calibrated trader has stated≈actual on each row. Most traders are systematically over-confident in the 70–90% bin. The Brier score is the average squared error — lower is better, <0.20 is good calibration.

Discovers stocks being accumulated by smart money. Composite score 0-100 blends 5 signals: famous-investor matches, MF deal proxy, FII deal proxy, residual institutional flow, and (per-stock only) promoter-insider net buying. Source: NSE bulk + block deals classified by counterparty against a curated 40-investor list.

Timeframe:

Every bulk/block deal from the past ~60 days, sortable & filterable. This is the raw data the Discovery scores above are computed from.

Shows ALL stocks where the selected marquee investor appears on the disclosed shareholder list. Limitation: only positions >1% of company are disclosed by SEBI; smaller positions invisible. Click any investor card to see their holdings.

WHO actually owns this stock? Per-stock view shows promoter%, FII%, DII%, MF%, public%, plus promoter-insider activity timeline. Group rollup sums your portfolio exposure across the major Indian business houses (Tata/Adani/Reliance/etc) so you see hidden concentration.

Fetches live data for every stock in the selected group at once. Large groups (e.g. Birla = 39) can be slow on mobile. Cells that can't resolve live fall back to the last CSV snapshot, marked stale.

Decomposes a daily move into sector pull (sector index × beta), macro pull (USD/INR, oil, gold, US 10Y × betas), and residual (stock-specific). Overlays headlines from the same window. Multi-factor attribution like a pro desk would run, computed in seconds locally.

Pre-trade slippage estimate using a square-root market-impact model: cost % = κ × σ × √(participation). Tells you what your trade actually costs vs the LTP you see on screen — especially important for mid-caps where the displayed price is often fiction for any meaningful size.

Quantifies whether a stock has historical edges around Indian-market-specific calendar windows. For each event window: average return, win rate, sample size, and a t-statistic for significance (|t|>2 ≈ 95% confidence). Significant calendar edges are real and exploitable — most retail traders never measure them.

Hard pass/fail audit of any stock against 10 codified rules from legendary investors. Different from a screener — this is deterministic vetting: each rule shows its actual value, threshold, and reasoning. Composite Score is 0–10 (count of rules passed, weighted). Investor-Fit detector identifies whose framework the stock best matches.

Defensive scanner — surfaces governance, accounting, and solvency red flags BEFORE they implode. Runs 7 academic-grade checks. Catches Yes-Bank / DHFL / Vakrangee-style disasters early. Pair this with Rulebook: Rulebook says go, Forensic says no-go.

A compact decision layer that combines macro flow, volatility, news and risk into one tradeable view.