FRIDAY · Status

Return since cycle start

FRIDAY

—

Top 500

—

Against the index

—

Both measured from the same start date. Capital added mid-cycle is excluded from FRIDAY's return, so top-ups cannot flatter the number.

Model portfolio

Current value · FRIDAY

—

Top 500

—

Invested on day 0

₹100.00

A notional ₹100 compounded each session by the day's return — like a smallcase model portfolio. Real rupee balances are never published.

Portfolio stats

Risk measures computed from the same daily series, the way a smallcase model portfolio reports them.

Max drawdown—

Volatility · annualised—

CAGR · annualised—

Days in cycle—

Risk gauges

Read-only analysis run each session: the market's volatility regime, a one-day volatility forecast, concentration limits, and the position-sizing band. No names, no counts, no thresholds.

Volatility regime—

Vol forecast · 1 day—

Concentration—

Sizing band—

Return against the Top 500

Both lines start at zero on the first day of the cycle. The gap between them is the only number the covenant is judged on.

FRIDAY Top 500

Cumulative percentage return since the cycle began. Hover for any day.

Value of ₹100 invested

Both lines start at 100 on the first day of the cycle and move with each day's return, like a model portfolio.

FRIDAY Top 500

What ₹100 invested in each has become, computed from the cumulative returns above.

View the daily figures as a table

Date	FRIDAY	Top 500	FRIDAY ₹	Top 500 ₹

Day by day

Each session's return for FRIDAY and the Top 500, around the zero line — the swings behind the cumulative curves above.

FRIDAY Top 500

Daily percentage change, derived from the cumulative series. Hover a bar for the day.

What FRIDAY is

FRIDAY is an autopilot for a single equity account. It is not a fund, it takes no outside money, and it manages a single account and nothing else. It runs on a fixed daily schedule without a human in the loop: it forms tomorrow's plan the previous evening, and the next day it either executes that plan or does nothing at all.

Everything it does is delivery-based cash equity. No leverage, no derivatives, no intraday positions, no short selling. A position is bought outright and held until a rule sells it.

How the model portfolio is maintained

One score decides everything

Every candidate is reduced to a single number out of 100, built from trend quality, momentum relative to the index, participation from volume, relative strength and the reward available against the risk taken. Only high scores are bought, and the same score is re-applied to open positions every week. There are no favourites and no discretionary overrides.

Risk is fixed before entry

Position size is set from the stock's own volatility, so that every trade puts the same small fraction of capital at risk regardless of which stock it is. Size is recalculated from live account equity on every run, so exposure scales with the account rather than drifting.

Exits are mechanical

Losers are cut on a stop ladder rather than on judgment; winners are held while the rules that bought them still hold. Exits rest as exchange-held orders that a fifteen-minute sweep places or repairs, so a gap in protection never lasts out the hour. Falling equity tightens the system automatically, halving new position sizes and then pausing new entries entirely if the drawdown deepens.

It trades rarely

A broad market filter throttles or stops new buying when the market's internals are weak, whatever the individual scores say. The normal cadence is nought to two new positions a week, frequently none. Doing nothing is the default state, not a failure of the system.

The daily routine

The previous evening, FRIDAY forms the next day's plan from a broad nightly scan and a single composite score. In the morning a breadth filter checks the market's internals before any new entry is allowed, and buys happen only in two short windows. Every open position carries a resting exit order, and a sweep runs every fifteen minutes during the session to place or repair missing stop and target orders — protection is refreshed through the day, not set once and forgotten. A position review closes the loop each evening, and the covenant is marked against the Top 500 after the close.

Nightly scanbroad universe

One scoreout of 100

Next-day planformed the evening before

Breadth checkmorning filter

Buy windowstwice a day

GTT sweepevery 15 min

Daily markafter the close

The research layer

Alongside the mechanical rules, a small statistical layer runs read-only with every session — the closest thing to AI in the stack. A volatility model forecasts how much the market is likely to swing next, a correlation check looks for hidden overlap between positions, and a sizing band is learned from the trade journal and re-checked as that journal grows. Their outputs are the four risk gauges above, reduced to plain aggregates. The layer is advisory: it can tighten the system's behaviour but never loosen it, and it never places an order.

What this page will never show

No holdings. Not the names, not the number of positions, not entry or exit prices, not what was bought or sold on any given day. The published file contains aggregate figures only — dates, a day counter and two percentages — and there is no code path by which an instrument name could reach it. The account's absolute rupee balance is likewise never published: the page shows returns and a value-of-₹100 index, not an amount. The selection rules and the thresholds behind the score are also not published. What you get is the outcome, on the same terms the covenant is judged on.