Register
Product

Manage Orders from the QuantPower Platform

Order management is everything that happens to an order after you place it: whether it was accepted, filled, partly filled, rejected or cancelled, at what price and when, and what you do next. QuantPower's order book shows every order, manual or automated, in one list with its status, fill price and time, and lets you change or cancel orders that are still open.

QuantPower Research · Updated 12 Nov 2026 · 18 min read

An order is not a position. Check the order book after every entry and exit: an order that is partly filled, rejected or still open means your position differs from your plan, and you need to act before the market moves further. In QuantPower every order, from the ticket, the option chain, the chart, a basket, the Scalper, a GTT, an algo, a signal or a prepared order from your AI app, goes through one order path with the same checks, and lands in one order book. This page explains the life of an order, each status and what you can do with it, how partial fills and rejections work, and the daily habits that keep your orders and your positions in step.

What order management in QuantPower does

PartWhat it does
One order pathEvery order, manual or automated, goes through the same checks and the same broker connection
Order bookOpen, executed, GTT, SIP and basket orders in one list, with filter chips and a Strategy column showing which algo, signal or SIP placed each order
DetailsThe status trail, the fills and the reason for a rejection in plain words
Modify and cancelFor any order still open or waiting for its trigger
Duplicate guardEvery order carries a unique key, so a retried request never places twice
TradesToday's fills from each broker's trade book and the Forward-test book
Several brokersA broker picker in the order sheet; each order and exit goes to the right broker
Live updatesOrder events reach the app over a live connection as the broker reports them

A short history of order management

For most of market history an order was a slip of paper. A client called a broker, the broker wrote a ticket, and a clerk carried it to the floor, where a dealer shouted it into the crowd. The ticket's journey was the order's status: written, sent, executed, confirmed by a note the next day.

Computers changed the journey before they changed the market. In the late 1980s and early 1990s, large brokers and fund managers built order management systems to track orders from the client to the exchange. In 1992 Fidelity Investments and Salomon Brothers began work on the Financial Information eXchange protocol, FIX, a common language for sending orders and reporting their status electronically.1 Its states, new, partly filled, filled, cancelled and rejected, are still the ones every trading app shows.

In India, the exchange itself moved to screens. The NSE started screen-based trading on its NEAT system in November 1994, matching orders by price and time priority on a computer.2 Orders could now be accepted, partly filled and cancelled in seconds, and an order book became something a broker's dealer watched on a terminal. Online trading brought that book to clients from 2000.

Automation added new problems: orders sent too fast, orders duplicated after a timeout, and orders that changed too often. SEBI's 2012 guidelines on algo trading asked exchanges to watch order-to-trade ratios and to put limits on order rates.3 Under SEBI's framework for retail algo trading, in force from 1 April 2026, a client's own algo must stay below 10 orders a second unless it is registered.4 Good order management today means the same as it did with paper tickets: know where every order is, and never send the same order twice. The tools are faster; the discipline is the same, and an order book you read every day is still the cheapest risk control there is.

The life of an order

  1. Built. A screen, an algo or a signal builds the order: the instrument, the side, the quantity, the price and the product.
  2. Your tap, for a manual order. The tap is tied to your login and to the exact order, works once and lasts 60 seconds. Algos and armed signals do not need a tap, because you switched them on.
  3. Prepared. The instrument, the product, the lot size and the tick size are checked.
  4. Risk checks. The segment is on; the kill switch, the daily loss limit, the maximum orders a day and the F&O ban list allow it. A square-off is never blocked.
  5. Sent to your broker through its official API, from your registered static IPs. Your broker's risk system checks margin, price, quantity and order rate.
  6. At the exchange. The order waits in the exchange's book until it fills, is changed or is cancelled.
  7. Recorded. Each change of status reaches the app over the live connection and appears in the order book.

A failure at any step ends the order there, and the reason shows where it stopped.

Order status and what you can do

Today's orders in the Orders tab: time, instrument, side, quantity, price, status and the reason for a rejectionToday's orders in the Orders tab: time, instrument, side, quantity, price, status and the reason for a rejection
Today's orders in the Orders tab: time, instrument, side, quantity, price, status and the reason for a rejection
StatusWhat it meansWhat you can do
PendingSent, and waiting for the broker or exchange to confirmWait a moment; do not send it again
OpenWaiting in the exchange's book at your priceWait, change the price or quantity, or cancel
Trigger pendingA stop-loss order waiting for its trigger priceMove the trigger, or cancel it if you exit another way
Partly filledPart of the quantity has filled; the rest waitsWait, change the price for the rest, or cancel the rest
AMOAn after-market order waiting for the next sessionChange or cancel it before the session opens
FilledThe whole quantity has filledNothing to change; manage the position
CancelledCancelled by you, by the system, or at the session closePlace a new order if you still want the trade
RejectedRefused by QuantPower's checks, your broker or the exchangeRead the reason, fix the cause, and place it again

Open the order's details for the status trail, each fill with its price and time, and, for a rejection, the reason in plain words. The details sheet is the first place to look when an order did not do what you expected, before you check your broker's app or ask support.

Order types at a glance

Order typeWhat it doesUse it for
MarketFills at once at the best price availableExits that must happen now, in liquid contracts
LimitFills at your price or better, or waitsMost entries, and all orders in thin strikes
SL (stop-limit)Sends a limit order when the trigger is reachedStops on options, where NSE does not accept SL-M
SL-M (stop-market)Sends a market order when the trigger is reachedStops on futures and shares

Validity can be Day, IOC (fill what you can now and cancel the rest) or GTD. Products are Intraday, Delivery for shares, and Overnight for F&O, commodity and currency. The order types guide explains each, with examples of when it fills and when it does not.

Changing and cancelling orders

An open order can be modified from the order book, or by dragging its line on the chart. The order sheet opens with the change for your confirmation; nothing changes until you confirm. Cancel asks first.

  • A price change costs queue position. Orders at the same price fill in time order, so a changed order joins the back of the queue at its new price.
  • Changes are limited. Each change is a separate request to your broker, and brokers can cap how many times one order is modified.
  • Cancelling stops only the unfilled part. Anything already filled stays as a position.
  • Cancel orphaned stops. If you exit a position by hand, cancel its pending stop-loss order. Otherwise the stop can trigger later and open a new position the other way.

Never twice: the duplicate guard

The most expensive order mistake is placing the same order twice. It usually happens after a slow response: the app seems to hang, the trader taps again, and both orders reach the exchange.

QuantPower guards against this at two levels. Every order carries a unique key, so if the same request reaches the server twice, only one order is placed. And if the answer from the broker is lost, the app asks the server by that key before it says anything. It never tells you "nothing was sent" when it does not know, and it never retries an order on its own. If the result is still unknown, it tells you to check Orders.

A timeout is not a rejection: the order may already be at the exchange. Check the order book, and your broker's app if needed, before you send it again.

Partial fills

A limit order fills only as far as there are sellers or buyers at your price or better. The rest waits.

Take a limit order to buy 200 shares at ₹1,416.55. Eighty fill at once, then the price moves up and the other 120 stay open. Your position is now 80 shares, not 200, so any stop or exit must be for 80 until more fill. You can wait, raise the limit for the remaining 120, or cancel them. The figures are a sample.

An algo makes this choice by its rules. Its entry can be set to "modify until filled": if the order is still open after a set number of seconds, its price moves to the last price, up to 10 times, and after that it is cancelled or sent at market, as you chose.

Iceberg orders

An iceberg order in the order window: the total quantity, the number of legs and the size of each legAn iceberg order in the order window: the total quantity, the number of legs and the size of each leg
An iceberg order in the order window: the total quantity, the number of legs and the size of each leg

An iceberg splits one order into 2 to 50 legs and sends them one at a time. Each leg is a normal order with every check, and the next goes only after the one before has fully filled. There are no retries: a leg that is rejected, cancelled or partly filled, the kill switch or a closed market stops the rest, and the order book shows why. Cancelling an iceberg stops the later legs; the leg already at the exchange is cancelled only if you tick that. An iceberg lasts for the day. The iceberg orders page explains when to use one.

Rejected orders

A rejected order never reaches the market and opens no position. The reason comes from QuantPower's own checks, your broker or the exchange, and usually falls in one of a few groups:

GroupExampleFix
Account controlsThe kill switch is on, or the daily loss limit or order cap is reachedWait for the next session, or review your limits
MarginNot enough margin or premiumAdd funds, reduce the size, or place the hedge leg first
PriceA limit outside the exchange's price bandSet a price closer to the market
QuantityNot a multiple of the lot size, or above the freeze quantityFix the lots, or use an iceberg
ContractExpired, or a stock in the F&O banChoose another contract, or only reduce positions
SessionThe market is closed, or the broker sign-in has lapsedPlace it in the session; sign in again
ConnectionThe static IP route or the broker is not answeringWait and check; orders are held, never sent from another address

When one leg of a multi-leg order is rejected and another fills, your position is not what you planned. A hedged basket avoids the worst case by sending bought legs first and sold legs only once the hedges fill.

GTT orders, baskets and SIPs in the order book

The order book has filter chips for the orders that wait or repeat:

  • GTT. A GTT waits for its trigger for up to 365 days, then sends its order through the same path as any other. An OCO GTT holds two legs, such as a stop and a target, and the first to fire cancels the other. You can have up to 100 active. Creating or changing one needs your tap, and a change whose trigger the price has already reached is refused, because it would fire at once. On the chart, a GTT is a line you can drag to a new trigger.
  • Baskets. A saved list of up to 20 legs, sent together with one tap. The tap is tied to the exact version you reviewed; if the basket was edited since, it is refused. In a hedged basket, bought legs fill before sold legs go.
  • SIP. A stock SIP's instalments appear as orders on their due dates. In reminder mode you are told on the due day and tap to place it; in auto mode the instalment is placed for you, with a record of when and how you agreed to it.

Orders on the Forward-test account

Your Forward-test account has its own order book, with the same statuses and the same order types, kept apart from your real money. Market orders fill at once at the price you see; a resting limit fills at its limit when the price crosses it; SL-M fills at the price once triggered; SL becomes a resting limit once triggered and fills only within its limit. Day orders left open are cancelled at the session close, IOC orders never rest, and an AMO goes live at the next session open. It is the place to practise modifying, cancelling and reading a partial fill before real money depends on it.

Orders prepared by your AI app

If you link your own AI app, such as Claude or ChatGPT, through Connect AI and give it the permission to prepare orders, it can write an order card into your app. A card waits up to 15 minutes, and at most 5 wait at once. Nothing reaches your broker until you confirm the card with your tap and your second factor. A confirmed card goes through the same order path and the same checks as a manual order, and it is decided once: a card being placed is never retried, so a crash cannot cause a second order. The AI app itself cannot place, change or cancel any order.

Manual and automated orders together

Orders from your algos and signals go through the same order path as your manual orders, so they appear in the same order book. The Strategy column names the source of every order: the algo or signal algo by its name, a SIP, a basket, Smart Exit, or Manual for orders you placed yourself. Filter by it to see one algo's orders on their own, and check them against its rules: the entry time, the strikes, the quantity and the exits.

  • Exits are reduce-only. An algo's exit order can only reduce a position. It can never open a new one.
  • Refused exits are retried. If an algo's exit is refused, it tries again up to 8 times with a growing gap, and the algo's card shows what is still owed.
  • Refused entries are unwound. If an entry leg is refused, the algo closes the legs already in, stops, and shows the reason.
  • Stop or pause before you intervene. If you change or cancel an algo's order by hand, the algo may still act on its rules. Pause it or stop it first.
  • Keep algos off contracts you trade by hand, or a manual exit may leave an algo managing a position that no longer exists.
  • Watch the request count. Each new order, change and cancellation is a request. Your own algos must stay below 10 orders a second, and brokers can set lower limits.

Orders the system places for you

Some orders in your order book are placed by QuantPower's account controls, not by you or an algo. Each says why.

  • Smart Exit. When your account's target, stop, locked profit or trail is reached, Smart Exit cancels the open orders and exit GTTs in its segments, stops the algos there and squares off every position, each at the broker that holds it.
  • Exit all. When you press it, running algos stop first, then every position, or every position in one segment, is squared off.
  • The kill switch cancels the open orders in the segments you stop, at every connected broker. It places no new order and leaves positions as they are.
  • Algo exits. At an algo's exit time, or when its stop or target is hit, its exit orders appear with the algo as the source.

A square-off is never blocked by the account's own limits. The daily loss limit refuses new orders, but an order that reduces or closes a position always goes, so a limit can never trap you in a position.

Orders across several brokers

With two or more brokers connected, the order sheet shows a broker picker, and nothing is ever sent to a guessed broker. An algo sends every leg to the broker you picked when you switched it on. Each exit goes to the broker that holds the position. The order book shows each order with its broker, and the risk checks count your positions at every broker together. If a broker does not answer, its orders are refused rather than assumed, and "no answer" is never read as "no positions".

A worked example: an entry that half-fills

At 9:20 am an algo sells one lot each of the Nifty at-the-money call and put, with limit entries and "modify until filled" after 5 seconds, up to 10 times, then market.

  1. The call fills at once. The put's limit is not reached.
  2. After 5 seconds, the put's price moves to the last price. It still does not fill.
  3. On the third modification, it fills.
  4. The order book shows the put's trail: open, modified, modified, filled, with each price and time.
  5. Both legs are now positions, each with its own stop.

Had the put been refused instead, for margin, the algo would have bought back the call, stopped, and shown the reason, so you would not be left with one sold option.

A worked example: an exit after a timeout

At 2:55 pm you tap Exit on a position. The connection drops before the broker answers.

  1. The app does not say "failed" or "nothing sent". It asks the server about the order by its key.
  2. The server has the answer: the order reached the broker and filled.
  3. The order book shows the exit as filled, and the position as closed.

If you had tapped Exit again before the answer came, the duplicate guard would have stopped a second order. That second order, on a closed position, would have opened a new position the other way.

Checking fills against your plan

Compare each fill with the price your strategy expected. The difference, multiplied by the quantity, is your slippage. A few rupees a trade looks small, but over hundreds of trades it decides whether a strategy that worked in a backtest still works live. For example, slippage of ₹0.50 on a Nifty option lot of 65 units is ₹32.50 a trade. Over 400 trades a year, in and out, that is ₹13,000, before brokerage or tax. Add a realistic slippage to your backtests, and compare it with what your fills actually show. If live slippage is much larger than the backtest assumed, find out why: thin strikes, market orders at busy moments, or entries at the same minute as many other traders.

An end-of-day order check

  1. No order is left open that you meant to fill or cancel.
  2. No stop-loss order is pending for a position you have already closed.
  3. Every rejection is understood, and its cause is fixed.
  4. Algo orders match their rules: entry time, strikes and quantity.
  5. Fills are close to plan, with any large slippage noted.

Day orders that have not filled are cancelled at the session close, but a filled part stays as a position until you close it. Read the orders and the positions together: position monitoring shows every open position with its live P&L, so an order that filled when you thought it had not is easy to spot. GTTs stay active until they fire, you delete them, or they reach 365 days.

Common mistakes

  • Tapping again after a slow response. Check the order book first.
  • Reading a trade at your price as a fill. A trade at your price means someone traded there, not that your order reached the front of the queue.
  • Exiting by hand and leaving the stop. The pending stop can open a new position later.
  • Changing an algo's order while it runs. Pause or stop the algo first.

The guide to algo trading risks covers these and other operational risks, from broker outages to runaway re-entries.

  • Ignoring partial fills. Size the stop and the exit to what filled, not to what you ordered.

Where order management is going

Order management is moving from the order to the account. Brokers and exchanges now check order rates, position limits and margins during the day, and traders need the same view: every order from every source, at every broker, under one set of limits. The other change is that more orders are prepared by software, including AI apps. In QuantPower, an AI app can only prepare an order card; your tap and your second factor turn it into an order, through the same path and the same checks as any other. Whatever prepares the order, the questions that matter stay the same: where is it, did it fill, at what price, and is my position what I planned?

Pricing

The order book, order entry, modify, cancel, iceberg, GTT and baskets cost no tokens. Tokens (1 token = ₹1, GST extra) pay for automation and AI: a live algo costs 200 tokens per 28 days with orders unlimited inside it, and a backtest 2 tokens with the first 10 each week free. See pricing.

Start with 500 free tokens

Connect your broker, build a strategy and test it on paper before a single live order goes out.

Frequently asked questions

What does "Pending" mean in the order book?

The order has been sent and QuantPower is waiting for your broker or the exchange to confirm it. It usually lasts a moment. Do not send the order again while it shows Pending; the duplicate guard would stop a repeat, but a different order would not be stopped.

Can I change an order after placing it?

Yes, while it is open, partly filled or waiting for its trigger. You can change its price, trigger or quantity, or cancel the unfilled part. Filled quantity cannot be undone; you close it with a new order.

Why is my order still open when the price has traded?

A limit order fills at your price only after the orders ahead of it at that price have filled. A trade at your price on the screen means someone traded there, not that your order reached the front of the queue.

What happens to unfilled orders at the close?

Day orders that have not filled are cancelled when the session ends. Any quantity that did fill stays as a position, so check positions as well as orders before the close.

Why did my stop-loss order trigger but not fill?

An SL order becomes a limit order once its trigger is hit. If the price jumps past the limit, the order waits unfilled. Leave a buffer between the trigger and the limit, and act at once if a triggered stop shows as open.

Are orders placed in QuantPower also in my broker's order book?

Yes. Every order is placed in your broker account, so it also appears in your broker's app, where you can match it by order ID.

References

  1. FIX Trading Community: history of the FIX protocol, begun in 1992 by Fidelity Investments and Salomon Brothers (fixtrading.org)
  2. NSE: history, screen-based trading on NEAT from November 1994 (nseindia.com)
  3. SEBI, circular "Broad Guidelines on Algorithmic Trading", 30 March 2012 (sebi.gov.in)
  4. SEBI, circular on safer participation of retail investors in algorithmic trading, 4 February 2025, in force from 1 April 2026 (sebi.gov.in)
  5. QuantPower: order path, statuses, iceberg, duplicate guard and algo order rules, from the app's code, checked 12 November 2026
Order statuses come from your broker and the exchange and can lag in fast markets, so check your broker's order book when in doubt. This page is for information only and is not investment advice. Trading involves substantial risk.