Prayog functional overview

What a person can do on Prayog, what the exchange does on its own, and when each thing happens. The model is an NSE-style cash equity market, based on public information only. Prayog is a simulation and is not affiliated with any exchange.

How to read this

This page describes the target product: what a production exchange in this style offers. Each feature is tagged with when Prayog builds it:

MVP in the current build plan (sessions S1–S22).
PHASE 2 next after the MVP. Needs your approval before it enters the plan.
LATER production features we describe so the design leaves room for them, but do not build yet.

Some things people think of as exchange features are really broker features: the broker's system holds the order and sends it to the exchange later (GTT, after-market orders, bracket orders). Those are marked as such. Prayog plays both roles, but the line matters for the design, because broker features live outside the matching engine.

The companion page, technical.html, explains how each feature is built and how data flows through the system.

1. The big picture: exchanges and trading firms

Videos about HFT describe the whole industry at once: exchanges, trading firms, feeds, FPGAs, monitoring, reporting. These belong to two different organisations on either side of a network cable:

The exchange (NSE, BSE, Nasdaq) runs the market. It keeps the one official order book per stock, matches orders fairly, publishes prices to everyone, and makes sure trades settle. It does not care whether you win or lose.
A trading firm (an HFT firm, a hedge fund, your paper-trading algo) is a customer of the exchange. It reads the exchange's price feed, decides what to trade, and sends orders. Its whole job is to make money, and in HFT, to do it faster than other firms.

Both sides have an "order book", which is a common source of confusion. The exchange's book is the real one. A trading firm rebuilds its own copy from the exchange's feed (a "book builder") so its strategy can read it without asking the exchange.

What each component does, and where Prayog stands

Component	What it does	Prayog
Trading firm side
Colocation and network	Servers rented inside the exchange's data centre, cables cut to equal length, microwave links between cities. Every metre is about 5 ns	Not built. Everything runs on one machine
Time sync	GPS and PTP clocks so every server agrees on time to within nanoseconds. Needed to measure latency and for regulators' timestamps	Not built. Prayog uses a simulated clock
FPGAs and fast network cards	An FPGA is a chip you program as a circuit rather than as software. HFT firms use them to decode the price feed and fire an order in well under a microsecond, faster than any CPU program. Fast network cards skip the operating system ("kernel bypass") to save microseconds	Not built. It's hardware. The software ideas behind it (no allocation, busy-spinning, one thread per core) show up in our engine
Feed handler (market data ingestion)	Reads the exchange's raw feed, decodes it, spots missing messages by sequence number, and recovers them	Simplified: the SDK's WebSocket client with gap detection (S12)
Book builder	Rebuilds each stock's order book from the feed so the strategy can read it locally in nanoseconds	Simplified: the SDK keeps a local book from snapshot + updates (S12)
Strategy / signal engine	Decides what to buy and sell: market making, arbitrage, momentum. This is the firm's secret ("alpha")	Yours: your bot. Our agents (S13) are simple strategies that create the market
Pre-trade risk checks	Blocks orders that break limits (position, order size, price far off the market, loss limit) before they leave the firm. Required by regulators after real disasters (Knight Capital lost $440m in 45 minutes in 2012)	Exchange-side checks are built (max quantity, bands, rate limits, kill switch). Firm-side checks in the SDK could be a Phase 2 item
Order management (OMS / EMS)	Tracks every order's state, handles partial fills and rejects, and decides which exchange to send to ("smart order routing")	Simplified: the SDK tracks your orders. No routing, because there is only one exchange
Order gateway	Speaks the exchange's order protocol (FIX or binary) and manages the session	Simplified: REST and WebSocket JSON via the SDK
Tick data store	Records every price update for research. Often terabytes a day, in specialised databases	Not built. Prayog's journal holds every event, so it could be exported later
Research and backtesting	Replays history through a strategy to estimate how it would have performed, before risking money	Not built. A natural Later item: replay a recorded Prayog day into your bot
Monitoring and alerting	Dashboards for latency, fills, errors and P&L; alerts when anything drifts	Built for the exchange (S21). Your bot's own monitoring is up to you
P&L, risk and reporting	Real-time P&L and exposure, end-of-day reconciliation with the broker, regulatory reports	Built on the broker side: positions, P&L, charges, blotter, leaderboard (S16–S19)
Exchange side
Order gateway	Accepts connections, authenticates, rate-limits, checks message shape	Built (S10)
Sequencer	Puts all orders from all members into one official arrival order, which is what makes "first come, first served" fair	Built (S7)
Matching engine	Keeps the official order books and applies the rules: priority, bands, sessions, self-trade prevention	Built (S4–S6). S4 is done
Market data publisher	Sends book changes and trades to everyone at the same moment	Built (S11)
Journal and replay	Records every input so the market can be rebuilt exactly after a crash	Built (S8)
Sessions and ops control	Opens, halts and closes the market; kill switches	Built (S6, S19)
Post-trade, monitoring, reports	For an exchange: trade reports and statistics. For a broker: positions, P&L, charges, contract notes	Built (S15–S21), simplified
Market surveillance	Looks for manipulation: spoofing, wash trades, pump-and-dump	Not built (Later)
Clearing and settlement	A clearing house guarantees every trade and moves cash and shares (T+1 in India)	Not built (Later)

Prayog's focus in one line: build the exchange properly (the same architecture production exchanges use), add a thin broker layer for positions and P&L, and give trading firms (you, your bots, our agents) a starter kit. That kit is the SDK, which covers the trading-firm pipeline in simplified software form. The HFT-specific hardware (FPGAs, colocation, nanosecond clocks) is out of scope: it is a different, very expensive engineering field, and it doesn't change how the market itself works.

A possible later track: if you want to go deeper on the trading-firm side, the natural next steps are a backtester that replays a recorded Prayog day into a bot, firm-side risk checks in the SDK, and latency measurement from your bot to the exchange and back. It's listed in open decisions.

2. Prayog at a glance

The same picture at the level of what people do. Everything inside the exchange box happens in the moment an order arrives. The "after the trade" boxes update a moment later and can never slow down trading.

The results of "after the trade" (P&L, blotter, leaderboard) show up in the trader's terminal and through the API, read from those services rather than from the exchange.

3. People and roles

Role	Who	How they connect	What they can do
Trader MVP	A person trading by hand	Web terminal at `app.prayog.localhost`, signed in through Keycloak	Place, modify and cancel orders; watch the book and charts; see positions, P&L and the leaderboard
Bot developer MVP	A person running their own algorithm	Python SDK or plain REST + WebSocket, with a client ID and secret as the API key	Everything a trader can do, but from code. Higher message rates, so rate limits matter
Simulated trader MVP	Prayog's own agents that create the market	The same public API as bots: no back door	Quote prices, trade on news-like shocks, add noise. They make the market feel alive
Ops MVP	The person running the exchange	Ops page in the web app (role `ops`)	Open, halt and close the market; kill switch per account; change the clock speed; watch health
Viewer PHASE 2	Anyone not signed in	Public market data and leaderboard	Watch only
Risk / compliance LATER	Surveillance staff	Reports and alerts	Spot wash trades, spoofing and unusual activity; review audit trails

4. Instruments

An instrument is something you can trade (here, a share of a made-up company). Every instrument has static settings that the exchange checks on every order.

Setting	Meaning	Example	Status
Symbol	Short name, upper case	`ABC`	MVP
Tick size	Smallest price step. Every price must be a multiple	5 paise: ₹100.05 is valid, ₹100.03 is rejected	MVP
Lot size	Smallest quantity step. 1 for Indian cash equities	1 share	MVP (fixed at 1)
Reference price	The price bands are measured from (normally yesterday's close)	₹100.00	MVP (static)
Price band	How far from the reference price an order may be priced today. NSE uses 2%, 5%, 10% or 20% bands depending on the stock	10% band on ₹100 → valid ₹90.00 to ₹110.00	MVP
Max order quantity	Largest single order. Protects against "fat finger" mistakes	100,000 shares	MVP
Status	Active, suspended, or delisted	A suspended stock accepts no orders	PHASE 2
Daily reference rollover	Next day's reference = today's closing price	Close ₹104.20 → tomorrow's band ₹93.80–₹114.60	LATER

5. The trading day

An exchange is not always open for continuous trading. The day has sessions, and each session allows different actions. NSE's cash market day, from public information (check current NSE circulars before relying on exact minutes):

Session	What happens	Orders allowed	Status
Closed	Nothing trades. Open orders from the day are gone (DAY validity)	None. Broker-side "after-market orders" queue outside the exchange	MVP
Pre-open: order entry (NSE 09:00–09:08, ends at a random moment in the last minute)	Orders are collected but nothing matches. The ending is random so nobody can game the last second	Place, modify, cancel limit and market orders	PHASE 2
Pre-open: matching (≈09:08–09:12)	A call auction picks one opening price, the one where the most shares can trade, and all crossing orders trade at it	None	PHASE 2
Buffer (≈09:12–09:15)	A quiet gap before continuous trading starts	None	PHASE 2
Open / continuous (09:15–15:30)	Every incoming order is matched immediately against the book by price-time priority	All order types and actions	MVP
Halted	Trading paused by ops or by a circuit breaker. The book is kept	Cancels only (common policy). New orders and modifies are rejected	MVP (ops halt)
Closing price	NSE sets the official close from trades near the end of the day; in 2026 it began moving some stocks to a closing auction	Depends on method	LATER
Post-close (≈15:50–16:00)	Trading only at the closing price	Limit orders at the close price	LATER

Simulated time. Prayog's clock is simulated. At 1× it follows real time; ops can speed it up (for example 10×, so a full day takes 37.5 minutes). Every session change is driven by that clock, so a replay of the day sees exactly the same session changes.

6. Order types and validity

An order has a type (how its price works) and a validity (how long it lives). Prices are always whole paise.

Type	How it works	Example	Where it lives	Status
Limit	Buy at this price or lower / sell at this price or higher. Whatever does not fill immediately rests on the book	BUY 10 ABC @ ₹100.00	Exchange	MVP
Market	Take whatever the book offers, best price first, until filled. Never rests. Prayog stops at the price band; any remainder is cancelled	SELL 10 ABC at market	Exchange	MVP
Stop-loss limit (SL)	Sleeps until the last traded price touches the trigger, then becomes a limit order	SELL 10, trigger ₹95.00, limit ₹94.50	Exchange (NSE holds stop orders)	PHASE 2
Stop-loss market (SL-M)	Like SL, but becomes a market order when triggered	SELL 10, trigger ₹95.00	Exchange or broker	LATER
Disclosed quantity (iceberg)	Only part of the order shows on the book; the rest refills as it trades. Lets large orders hide their size	BUY 10,000, show 1,000	Exchange	LATER
GTT / OCO, bracket and cover orders	Conditional orders held by the broker, sent to the exchange when the condition is met	"If ABC hits ₹120, sell 10"	Broker	LATER

Validity	Meaning	Status
DAY	Lives until the market closes, then expires. The default on Indian exchanges	MVP
IOC (immediate or cancel)	Fill what you can right now; cancel the rest at once. Never rests. Bots use this a lot	PHASE 2
FOK (fill or kill)	All or nothing, right now. Not offered on NSE cash; common elsewhere	LATER
GTC (good till cancelled)	Survives across days. On Indian markets this is a broker feature (GTT), not an exchange one	LATER (broker)

7. Order lifecycle

Every order moves through a small set of states. Each arrow is an event the exchange publishes, so the order's state is always rebuildable from its events.

Open and Partially filled are the live states: the order can still trade, be modified or be cancelled. The other four are final. "Expired" is published as OrderCancelled with a reason; we will add an EXPIRED reason when sessions arrive in S6.

The quantities always add up: ordered = filled + open + cancelled. Prayog's property tests check this for thousands of random order flows.

8. Actions

Every action is described the same way: who does it, when it is allowed, what happens step by step, and what comes out.

A. Access

Sign in MVP

Who

Trader, ops

When

Any time

What happens

The web app sends you to Keycloak's login page.
Keycloak checks your password and returns a signed access token (a JWT) naming your account and roles.
The web app sends that token with every request. The gateway checks the signature; it never sees your password.

Edge cases

Tokens expire after minutes and are refreshed silently. A disabled account cannot get a new token.

Create bot credentials (API key) MVP

Who: Bot developer (ops issues them in the MVP)
What happens: A Keycloak client with a client ID and secret is created. The bot trades those for a token ("client credentials" flow) and then calls the same API as the web app.
Why it matters: Bots and humans use one API, so a bot can never do something a human cannot.

Rate limiting MVP

Who: Applied by the gateway to every account
What happens: Each account has a bucket of tokens that refills at a fixed rate (for example 50 per second). Each order, modify or cancel takes a token. An empty bucket means RATE_LIMITED, and the order never reaches the engine.
Why: One runaway bot must not slow the market for everyone. Real exchanges cap message rates and charge for excessive order-to-trade ratios.

B. Orders

Place a limit order MVP

Who

Trader, bot, simulated trader

When

Session OPEN (and pre-open in Phase 2)

What happens

The gateway checks the token, the rate limit and the message shape.
The engine validates it: known symbol, quantity 1 to max, price on tick, price inside the band, session open, account enabled. The first failure decides the reject reason.
If valid it gets an order ID and an OrderAccepted event.
It matches against the opposite side while prices cross: best price first, oldest first at a price. Each match is a Trade at the resting order's price.
Whatever is left rests on the book at the back of the queue for its price.

Outcomes

OrderRejected, or OrderAccepted followed by zero or more Trades

Example

Asks: 5 @ ₹100.00, 5 @ ₹100.10. You BUY 8 @ ₹100.05 → trade 5 @ ₹100.00; 3 rest as the new best bid at ₹100.05.

Place a market order MVP (S5)

When: Session OPEN
What happens: Same as a limit order, but with no price of your own: it sweeps the opposite side best price first. It stops at the price band (this protects you from buying at absurd prices in a thin book), and any remainder is cancelled with reason NO_LIQUIDITY. It never rests.
Edge cases: Empty opposite side → accepted, then fully cancelled with NO_LIQUIDITY. Many brokers turn market orders into limit orders a few percent away ("market protection"); Prayog's band limit plays that role.

Place an IOC order PHASE 2

What happens: A limit order that never rests: it fills what it can at its limit or better, and the rest is cancelled at once.
Why bots like it: No forgotten orders left on the book if the bot crashes.

Place a stop-loss order PHASE 2

What happens

Accepted but kept off the visible book, waiting.
When a trade prints at or through the trigger price (≤ trigger for a sell, ≥ trigger for a buy), it becomes a normal limit order and joins the back of the queue.
From there it behaves as a limit order.

Edge cases

One trade can trigger many stops, and their fills can trigger more ("stop cascade"). The engine must process triggered stops in a fixed, deterministic order (trigger price, then arrival).

Modify an order MVP (S5)

Who

The order's owner

When

Session OPEN; the order is still live

What happens

Reduce quantity: keeps its place in the queue. You are giving up size, which harms no one queued behind you.
Increase quantity or change price: goes to the back of the queue at the (new) price, as if new. Otherwise you could join a queue early with 1 share and later grow to 10,000, jumping everyone.
New price crosses the book: it trades immediately, like a new order.
Reduce to the filled amount or below: treated as a cancel of the open part.

Outcomes

OrderModified (with any Trades), OrderCancelled, or OrderRejected (UNKNOWN_ORDER if already filled)

Cancel an order MVP (S5)

When: OPEN or HALTED; the order is still live
What happens: The open part is removed from the book. Shares already filled stay filled.
Race: A cancel can arrive just after the order fully filled. That is normal and gets OrderRejected(UNKNOWN_ORDER). The client must handle "too late to cancel".

Cancel all (mass cancel) PHASE 2

What happens: Cancels all of an account's live orders, optionally for one symbol. Bots use it on startup and before shutting down.

C. Exchange rules (things the exchange does by itself)

Price-time priority matching MVP

Rule: Best price first; at the same price, the earliest order first. Trades happen at the resting order's price, so an aggressive order can get a better price than its limit ("price improvement").
Why: It rewards whoever offers the best price, and then whoever offered it first. That encourages people to show liquidity early.

Price band rejection MVP (S6)

Rule: A limit price outside today's band is rejected with PRICE_OUTSIDE_BAND. Market orders stop at the band edge.
Why: It limits damage from mistakes and panic: no single day can move a price more than the band.

Self-trade prevention MVP (S6)

Rule: If an incoming order would trade with a resting order from the same account, the incoming order's remainder is cancelled with SELF_TRADE_PREVENTION. Trades already made with other accounts stand.
Why: Trading with yourself creates fake volume (a "wash trade"), which regulators forbid. Other modes exist (cancel the resting order, or both); cancel incoming is simplest and the most common default.

Expiry at end of day MVP (S6)

When: When the session moves to CLOSED
What happens: Every live DAY order is cancelled, in a fixed order (by symbol, then order ID), so a replay produces the same events.

Opening call auction PHASE 2

What happens

During order entry, orders collect without matching.
At the end, the exchange tries every candidate price and picks the one where the most shares can trade. Ties go to the smallest leftover imbalance, then the price closest to the previous close.
All crossing orders trade at that one price. The leftovers move into continuous trading.

Why

Overnight news piles up. One fair opening price is better than a chaotic first second of trading.

Circuit breakers LATER

What happens: Stock level: the price band already acts as a circuit filter. Market level: when the index falls 10%, 15% or 20% the whole market halts for a set time (SEBI's index circuit breakers). In Prayog this would be an automatic HALTED state.

D. Market data

Order book (depth) MVP (S11)

What you get: On subscribe, a snapshot of every price level (price, total shares, number of orders). After that, an update for each level that changes. Each message has a sequence number, so a gap tells you to ask for a fresh snapshot.
Levels: Aggregated by price ("level 2"). Individual orders are not shown, and neither are account IDs.

Trade prints and ticker MVP (S11)

What you get: Every trade, anonymous: price, quantity, time, and which side was the aggressor. The terminal builds the last price, day high/low, volume and candlestick charts from these.

Your own order updates MVP (S11)

What you get: A private stream of your own orders' events (accepted, filled, cancelled, rejected). Only you see it.

E. After the trade

Positions MVP (S16)

What: Shares held per account and symbol: buys add, sells subtract. A negative position is a short sale.
When: Updated from each trade event a moment after the trade (asynchronously, off the order path).

P&L MVP (S16)

Realised: Profit locked in by closing a position. Computed against your average buy price (or FIFO lots; see decisions).
Unrealised: Profit on paper for what you still hold, valued at the last traded price. It moves with every trade print.
Net: Realised + unrealised − charges

Charges MVP (simple) · LATER (Indian detail)

MVP: A flat brokerage per trade plus a percentage fee, set in config.
Later: The full Indian stack: brokerage, STT, exchange transaction charges, SEBI turnover fee, stamp duty (buy side) and GST on brokerage and fees. Each has its own base and rate, and rounding rules matter.

Trade blotter and reports MVP (S19)

What: A list of your orders and trades with times, prices and charges, plus a P&L summary. The production version adds a daily contract note (the legal trade confirmation) and CSV export.

Leaderboard MVP (S17)

What: Accounts ranked by net P&L, updated live. Shows totals only, never individual trades.

Settlement LATER

What: In India, shares and cash actually change hands one business day after the trade (T+1). Until then the shares are "unsettled". Prayog would simulate cash and holdings balances and a daily settlement run.

Buying power and margin checks PHASE 2

What: Before accepting an order, check that the account can pay for it (cash minus what open orders have already reserved) or has enough shares to sell. In the MVP, accounts have unlimited money: fine for a game, wrong for a realistic broker.
Design impact: This check must happen on the order path and be deterministic, so balances live inside the engine (or a risk stage before it), not in the database.

F. Operations

Open, halt, resume, close the market MVP

Who: Ops, or the sim clock at the configured times
What happens: A SessionStateChanged event. Halting keeps the book but stops matching; closing expires DAY orders.

Kill switch MVP

Global: Ops sets the market to HALTED. Everything stops at once.
Per account: Ops disables one account: all its live orders are cancelled (reason KILL_SWITCH) and new ones are rejected (ACCOUNT_DISABLED). This is the tool for a misbehaving bot.

Change clock speed MVP

What happens: Ops changes the multiplier (1×, 10×, 60×). Sim time runs faster; nothing else changes. Because ticks are recorded, a replay is still exact.

Replay a session MVP (S8)

What happens: Feed the recorded input journal back into a fresh engine. The output must be byte-for-byte identical (checked with a SHA-256 checksum). This is how bugs are reproduced and how the exchange recovers after a crash.

Add or suspend an instrument PHASE 2

What happens: In the MVP, instruments come from config at startup. Later they become commands that go through the journal, so they replay too.

G. Simulated market

Simulated traders MVP (S13)

Kinds

Market makers keep a buy and a sell order around a fair price and earn the spread. They are why the book is never empty.
Momentum traders buy after rises and sell after falls.
Mean-reversion traders do the opposite.
Noise traders trade randomly in small size and keep volume flowing.
News shocks move each stock's hidden "fair value" now and then, so prices trend.

Rule

They use the public API like any bot: no special access. Their random choices are seeded, but because orders travel over the network their arrival order varies, so a run is not exactly repeatable. The exchange's replay is.

9. Worked example: one trader's morning

Instrument ABC: tick 5 paise, reference ₹100.00, band 10% (₹90.00–₹110.00). Charges for the example: ₹10 per trade.

Sim time	Action	Book before (best)	Result	Position · avg price
09:15:00	Ops opens the market	empty	`SessionStateChanged(OPEN)`	—
09:15:02	Market maker quotes 100 @ ₹99.95 / 100 @ ₹100.05	—	Two orders rest	—
09:20:10	You BUY 10 @ ₹100.10 (limit)	ask ₹100.05	Trade 10 @ ₹100.05 (resting price: 5 paise saved)	+10 · ₹100.05
09:31:00	You BUY 10 @ ₹101.95	ask ₹102.05	No cross; your bid rests and becomes the best bid	+10
09:33:40	Momentum bot SELLS 10 at market	bid ₹101.95 (yours)	Trade 10 @ ₹101.95; you are filled	+20 · ₹101.00
10:05:00	You SELL 15 @ ₹103.00	bid ₹103.10	Trade 15 @ ₹103.10	+5 · ₹101.00
10:30:00	Last trade on ABC prints at ₹102.50	—	Unrealised P&L is revalued	+5

Average price after both buys: (10 × 100.05 + 10 × 101.95) ÷ 20 = ₹101.00.
Realised on the sale: (103.10 − 101.00) × 15 = ₹31.50.
Unrealised on the 5 left: (102.50 − 101.00) × 5 = ₹7.50.
Charges: 3 trades × ₹10 = ₹30.00. Net P&L: 31.50 + 7.50 − 30.00 = ₹9.00.

Internally every one of those numbers is whole paise (₹101.00 = 10,100 paise). An average price that isn't a whole number of paise is kept as a total cost (paise × shares) instead, so no rounding happens until display.

10. Functional decisions still open

These change behaviour users can see. Each has a recommended default; we decide when the session that needs it starts.

Decision	Options	Recommended	Needed by
What can you do while HALTED?	Nothing / cancels only / cancels and modifies	Cancels only: people can reduce risk, but nothing new trades	S6
Realised P&L method	Average cost / FIFO lots	Average cost: simpler, and what most Indian broker P&L screens show for intraday	S16
Short selling	Allowed freely / allowed intraday only (square off by close) / not allowed	Allowed freely in the MVP; intraday-only with auto square-off in Phase 2 (that's how Indian cash shorts work)	S16
Starting cash	Unlimited / fixed (e.g. ₹10 lakh) with buying-power checks	Unlimited for MVP; fixed in Phase 2. It makes the leaderboard fairer	Phase 2
Leaderboard metric	Net P&L / return % / risk-adjusted (Sharpe-like)	Net P&L first; return % once cash is fixed	S17
Trading-firm track	None / backtester replaying a Prayog day / SDK-side risk checks / bot latency measurement	Decide after the MVP; the backtester gives the most learning for algo work	After MVP
Which Phase 2 features first	IOC, stop-loss, pre-open auction, buying power, mass cancel	IOC and mass cancel (small, and bots need them), then buying power, then the auction	After MVP

11. Glossary

Term	Meaning
Bid / ask	A buy order's price / a sell order's price. Best bid = highest bid; best ask = lowest ask
Spread	Best ask − best bid. Narrow spread = cheap to trade
Resting / passive / maker	An order waiting on the book; it "made" liquidity
Aggressive / incoming / taker	An order that trades on arrival; it "took" liquidity
Crossed book	Best bid ≥ best ask. Must never last in continuous trading: those orders should have traded
Depth / level 2	Total quantity at each price, per side
Tick size	Smallest allowed price step
Price band / circuit filter	Today's allowed price range for a stock
Leaves quantity	The part of an order still open
Call auction	Collect orders, then trade everything at one price that maximises volume
Wash trade	Trading with yourself; blocked by self-trade prevention
Paise	1/100 of a rupee. All Prayog prices are whole paise
T+1	Settlement one business day after the trade