The work a portfolio manager does. On your own book.

State the mandate. Every strategy proves itself against the benchmark before anything funds it, and the winner runs forward on live data, with a desk of analysts on it, a bull and a bear arguing each name, and a written reason behind every proposed trade.

Flagship momentum, against the S&P 500
54 months, $100,000 start. Stored backtest, not a simulation run for this page.
  • Flagship Momentum
  • S&P 500
  1. 01

    Define

    A mandate is a statement of what a portfolio should be: universe, strategy, how many names, cash reserve, stop-loss.

  2. 02

    Prove

    Every strategy is backtested against the benchmark and ranked on risk-adjusted return. Nothing funds a book on a hunch.

  3. 03

    Run

    The winner runs forward on live market data, and each proposed rebalance arrives with the reasoning attached.

The job

What a portfolio manager actually does.

Six of them, and what does each one here. Nothing on this list is a plan.

Sets the mandate and holds to it
You state the universe, sectors, risk profile and how many names. Every proposal is checked against it, and a drift is called out before it is accepted.
Forms a view on each name
A committee scores the security, then a bull and a bear argue it and a judge rules, so the weak side of a thesis is visible rather than buried.
Watches risk, regime and macro
A desk of agents covers market regime, macro, concentration and news, each writing from the same fact pack rather than its own impression.
Rebalances, and says why
Proposed trades arrive with turnover, sector drift and a written rationale per name. You accept or reject; nothing moves on its own.
Answers questions about your own book
Ask about exposure, risk or a holding and the answer is computed from your positions, not generalised from the internet.
Reports, benchmark-relative
The book marked daily in its own market and currency, against the index it claims to beat, with realised and unrealised P&L kept separate.

It does the work; you make the decisions. Every trade is proposed and waits for you, nothing is executed anywhere, and this is research tooling rather than an advisory service or a brokerage.

What it holds

Five surfaces, one terminal.

Portfolio

From what you want, to a book that runs.

Mandates

Sector allocations, risk profile, universe and position count, the way an allocator actually thinks.

Strategy Lab

Backtest one mandate under every engine at once, ranked by risk-adjusted return.

Paper trading

Every portfolio runs forward on live market data before a dollar moves.

Rebalancing

Proposed trades with turnover, sector drift and a written rationale for each name.

Research

The evidence under every decision.

Security analysis

A 24-factor read per security, each weight explained in plain English.

Scenario analysis

What-if from revenue through margin to EPS and multiple.

Diversification study

Monte Carlo across size and sector: how many names actually diversify.

Strategy backtests

Stored runs, benchmark-relative, equal and cap weighted.

Earnings intelligence

Not just the number. What was said, and what changed.

Earnings calendar

Who reports next, with our forward-EPS forecast beside the street's consensus.

Call and release sentiment

Every line of the earnings text scored and weighted by materiality: the arc of the call, and the guidance that moved it.

AI read of the 8-K

Why results changed, as structured drivers pulled from the filing itself.

Post-earnings drift

The one earnings signal that survived walk-forward testing, used as such.

Ownership and linkage

Who owns it, and what it is connected to.

Institutional holdings

Form 13F positions per security, quarter by quarter, by manager.

Beneficial owners

Schedule 13D/13G in the US and SAST in India: who crossed 5%.

India shareholding

Promoter, FII, DII, mutual-fund and retail splits, plus promoter pledging.

Ripple

An event in one company, and the correlated names it reaches, with the business linkage and not just the correlation.

AI layer

Grounded in what the platform computed.

Portfolio co-pilot

Ask questions about your own book and get answers grounded in computed facts, never freehand.

Analyst desk

Several agents, covering regime, macro, risk and opportunities, each writing from the same fact pack.

Bull vs bear debate

Opposing cases argued per security, then judged, so the weak side of a thesis is visible.

Every answer grounded

The model narrates numbers the platform computed. It is not asked to know anything on its own.

Underneath

The sources, loaded and reconciled, not linked out to.

SEC EDGAR

10-K, 10-Q and 8-K filings, XBRL facts, earnings press releases, Forms 3/4/5, 13F and 13D/G

NSE and BSE

Bhavcopy prices, corporate filings, Reg-31 shareholding XBRL, pledge and SAST disclosures

Market data

Daily bars for both markets, corporate-action adjusted, with a data-quality verdict per security

Macro and news

FRED macro panel, market news mapped to tickers, and sentiment scored against it

Start with a mandate.

Signing in with Google opens a ten day trial of the full terminal.

Open the terminal