AI-WM Lab · research preview · live since 4 June 2026

We let the AI run a portfolio. In public.

A team of Claude agents, three analysts, a portfolio manager and a macro strategist, manages a ₹10 lakh paper portfolio of NIFTY-500 stocks on our own capital. Every evening it reads the filings and the news, decides, and writes down why. Every return below is time-weighted with costs on. We publish it on the days it loses too.

Paper-traded, no live orders yet. Not investment advice. Not open to outside money.

Scoreboard since 4 Jun 2026 · last run 6 Oct 2026
updates after every evening run
Portfolio · time-weighted
-5.2%
₹10,00,000 start · costs and slippage on
NIFTY 500 · same window
-1.3%
^CRSLDX total return · the benchmark
NIFTY 50 · same window
-2.7%
^NSEI · for context
Portfolio minus NIFTY 500
-3.9 pp
the only number that matters here
Holdings
37 stocks · 10 sectors
Cash
16.3%
Portfolio value
₹9,49,434
Scorecard rows matured
47,475

The portfolio is behind the index so far. We show that because a scoreboard you can edit is not a scoreboard. The bar we set, three to five points a year over the NIFTY 500 across two market regimes, is not yet met, and four months is far too short to judge either way. Full scorecard on the dashboard.

why run it at all

A research tool should be able to
put its money where its numbers are.

The question

Can a disciplined team of AI analysts, reading the same filings and news our terminal serves, beat the index net of realistic costs? Not in a backtest. Day by day, forward, with the rules fixed in advance.

The honesty test

The terminal promises every number is proven. The lab is where that promise gets expensive. If the research is wrong, the portfolio says so in rupees, in public.

What we learn

Which signals carry information and which are noise. The scorecard tracks each analyst's hit rate at 5 and 10 days, so we promote or drop inputs on evidence, never on a hunch.

how one evening works

Seven steps, every
trading day, after the close.

The models do the reading and the judgement. Code does the gating, the risk limits, the tax maths and the fills. That boundary is deliberate: an AI may decide for our own book, but it never gets to bend a rule.

Also running, report-only
  • Macro strategist · one brief per session, prepended to every prompt
  • Devil's advocate (Claude Haiku) · argues against each decision, in shadow mode
  • Holdings correlation and F&O-ban tags · flagged, not yet enforced

Anything that could change a decision ships in shadow first and is promoted only after at least 30 trading days of scorecard evidence.

  1. 1
    Gatedeterministic

    Of ~150 tracked stocks, only those with fresh filings, news or a price event (typically 11 to 66) go to the analysts. The rest hold. Attention is rationed, like a real desk.

  2. 2
    Fundamental analystClaude Sonnet

    Reads the sector card, the filings pack, valuation versus peers and analyst consensus. Returns a thesis with growth and risk grades and a conviction score.

  3. 3
    Sentiment analystClaude Sonnet

    Reads event-classified headlines, management tone and the earnings surprise. Returns tone, drivers and conviction.

  4. 4
    Technical analystClaude Sonnet

    Reads indicators and derived pattern labels. Returns signal, setup and conviction. Price context only, never the thesis.

  5. 5
    Portfolio managerClaude Opus

    Synthesises the three views against current positions and cash, inside a macro brief the strategist writes once per session. Proposes buy, trim, sell or hold with size.

  6. 6
    Risk and tax overlaydeterministic

    Code, not a model, applies the caps: 5% per stock, 25% per sector, the cash rule, three new positions a day. Then estimates the tax cost of any sale. The model cannot override it.

  7. 7
    Fill and publishdeterministic

    Approved orders fill at the next open with transaction costs and slippage. The decision, the reasoning and the trace are pushed to the public log and the dashboard.

the rules it cannot break

Fixed in advance. Enforced in code.

A track record is only as honest as its accounting. These are the constraints the portfolio has lived under since day one, and the bar it has to clear before we call it anything.

5%
max in one stock
Position cap at cost. No hero trades.
25%
max in one sector
Diversification is enforced, not hoped for.
3
new positions a day
Slows the book down. Forces the manager to choose.
T+1 open
fill price
Decisions after the close fill at the next open. No look-ahead.
On
costs and slippage
Brokerage, STT, exchange fees and a slippage model on every fill.
TWR
the only return we quote
Time-weighted. A deposit is never a gain.
NIFTY 500
the benchmark
The benchmark is the opportunity set. No cherry-picked index.
+3 to 5 pp
the bar, per year
Net of costs, across at least two market regimes, before we call it anything.
the reading so far · cumulative

More filings than any analyst
reads in a career.

Agent analyses
6,149
fundamental · sentiment · technical · PM
Filings indexed
29,355
BSE / NSE, point-in-time
Documents deep-read
7,229
annual reports, results, concalls
Headlines analysed
71,988
event-classified, not keyword-scored
Events classified
54,269
order wins, regulatory, capex, guidance…
Price observations
2,03,600
daily closes across the universe
Latest run 6 Oct 2026
Decisions logged
509
0
buys
4
trims
4
sells
157
holds

Every decision carries the analyst views, the manager's reasoning, the risk verdict and the tax note. Read today's log.

What it is
  • A research experiment on the founder's own capital.
  • A forward, out-of-sample test of the same data the terminal serves.
  • Fully logged: every prompt, view, decision and fill is versioned.
What it is not
  • Investment advice, a tip service or a model portfolio to copy.
  • Open to anyone else's money. We are not SEBI-registered and do not manage client funds.
  • Live yet. Fills are simulated at the next open with real costs; no broker is connected.
Where it goes next
  1. 1Now: paper trading, scorecard accumulating, shadow agents under evaluation.
  2. 2Next: live execution of the same decisions, on our own capital, through an Indian broker.
  3. 3Then: the full decision log and scorecard published here, so the terminal's research can be judged by what it did, not what it said.
Built on Claude.

The analysts run on Claude Sonnet, the portfolio manager on Claude Opus, the devil's advocate and news classifier on Claude Haiku. The gate, the risk limits, the tax maths and the fills are plain code.