Colocation in Trading: Do Retail Traders Need It?
Blog by Aliceblue News Desk
20 November 2026
9 minutes

Colocation means a broker or trading firm keeps its server inside the stock exchange’s data centre, so its orders reach the exchange a fraction of a second sooner. Only exchange members can buy it, and high-frequency firms are its main users. If you trade through a mobile app or web platform, you don’t need it.
You place a buy order at ₹250, and it fills at ₹250.50. Then you read that big firms get faster access to the exchange, and you wonder if speed is costing you money. For most retail traders in India, it isn’t. This guide explains colocation in trading, how it works, what the NSE case was about, and what to fix in your own setup instead.
What Is Colocation in Trading?
Colocation is a paid service where a stock exchange lets trading members place their servers close to its own servers. ET describes it as traders being able to place servers in proximity to those of an exchange, which gives them a time advantage.
It exists because of distance. An order travels from a trader’s computer to the exchange’s matching engine, the software that pairs buyers with sellers. A longer path means a longer trip. The delay between sending an order and the exchange receiving it is called latency. Colocation cuts that delay to a very small number.
How Does Colocation in Trading Work?
- A member applies to the exchange. Brokers and approved trading firms can apply. Individual traders can’t.
- The firm places its servers in the exchange facility. The exchange rents out the space and the connection.
- The servers receive prices almost at once. The data barely has to travel.
- Software places the orders. A program called an algorithm reads prices and sends orders by fixed rules. Nobody clicks.
- The exchange matches orders by price and time. At the same price, the order that arrived first gets filled first.
Step 5 is the reason firms pay for speed. A swing trader who holds a stock for days won’t notice a gap of a few millionths of a second. A firm that earns a few paise on thousands of trades a day will.
What Was the NSE Colocation Case?
In 2015, a whistleblower wrote to SEBI alleging that NSE gave preferential access to a few brokers and high-frequency traders. SEBI’s main complaint was about the exchange’s tick-by-tick data system, which sends order information as it happens. SEBI said the system was open to manipulation, and that members on crowded connections were at a disadvantage against those on less crowded ones.
In April 2019, SEBI directed NSE to pay ₹687 crore and barred it from accessing the securities market for six months. It also banned two former NSE chief executives from associating with any listed company or exchange for five years. SEBI’s order set out the disgorgement as ₹625 crore plus 12 percent annual interest.
The case then moved through appeals:
- In January 2023, the Securities Appellate Tribunal set aside the disgorgement. It still asked NSE to deposit ₹100 crore for lack of due diligence.
- SEBI appealed to the Supreme Court.
- NSE settled the colocation and a related dark fibre case (unauthorised fibre lines laid for brokers) for ₹1,491.21 crore. The Supreme Court disposed of the colocation appeals on September 18, 2026.
Settlements in India are usually reached without admitting or denying guilt. The exchange’s facility stayed in use. The case was about who got access first.
Colocation vs Limit Orders: A Simple Example
These numbers are made up to show the idea. They are not from a real trade.
Part 1: Why speed matters to firms. Only 200 shares are on offer at ₹250.00. Three buyers each bid ₹250.00 for 200 shares. Buyer A’s order reaches the exchange first and takes all 200. Buyers B and C get nothing at that price. A co-located firm is trying to be Buyer A, thousands of times a day.
Part 2: Why speed isn’t your problem. A thinly traded stock shows these sell offers:
- 200 shares at ₹250.00
- 500 shares at ₹250.50
You want 500 shares.
Order type | What happens | Cost |
Market order for 500 shares | 200 fill at ₹250.00 and 300 fill at ₹250.50 | ₹50,000 + ₹75,150 = ₹1,25,150 |
Limit order at ₹250.00 for 500 shares | 200 fill now, 300 wait for a seller at ₹250.00 | ₹50,000 for the 200 filled |
With the market order, your average price is ₹250.30. You paid ₹150 more than a ₹250.00 fill (300 shares × ₹0.50). A faster server wouldn’t have fixed that. A limit order would have.
Compare that with a wrong view on direction. You buy 20 shares at ₹250, and the stock falls ₹15. You lose ₹300, with no speed gap involved. Before any trade, check the charges in the brokerage calculator. It shows brokerage, STT (securities transaction tax), GST, and your breakeven price.
Key Features of Colocation
- Inside the exchange premises. The servers sit next to the exchange’s own systems.
- Paid and member-only. Brokers and firms pay the exchange. Clients can’t buy it.
- Used with algorithms. No person can react in millionths of a second, so software does it.
- Run by exchange rules. Access is meant to be equal for every member who pays for it.
- Lower delay, same risk. A fast order that’s wrong still loses money.
Advantages and Risks
Why firms pay for it
- Orders and cancellations reach the exchange sooner.
- Market makers, who keep buy and sell quotes in the market, can update prices quickly.
- Arbitrage traders, who profit from small price gaps between two places, rely on speed.
What can go wrong
- The cost only makes sense for firms with very large volumes.
- A software bug can send thousands of wrong orders in seconds.
- Unequal access harms other participants, as the NSE case showed.
For a beginner, the bigger risk is behavioural. Chasing faster tools pushes you to overtrade, and each extra trade adds brokerage and taxes.
Do Retail Traders Need Colocation?
No. Your order goes from your phone or computer to your broker’s server and then to the exchange. Colocation sits on the broker’s side of that path, and you can’t buy it.
If you run algos, the delay that matters is the one between your software and your broker’s API, the link that lets software send orders to the broker. One hosting guide quotes under 30 ms for that link, but it sells hosting, so treat the number as a guide only.
SEBI’s retail algo rules point the same way. Brokers must allow API access only through a unique API key and a static IP address (a fixed internet address) that the broker has whitelisted. NSE set the threshold for a self-built algo at 10 orders per second per exchange. Below it, a retail algo doesn’t need separate exchange registration. The rules apply to all brokers from April 1, 2026.
Colocation vs Retail Setup
Parameter | Colocation (firms) | Retail setup |
Who uses it | Brokers, market makers, high-frequency trading (HFT) firms | Individual traders and investors |
Holding time | Fractions of a second to minutes | Minutes to years |
Order path | Server inside the exchange | App, then broker, then exchange |
Cost | Paid to the exchange | Brokerage and platform charges |
Main edge | Speed | Stock choice, patience, risk control |
Can you buy it? | Members only | No |
Tips for Beginners
- Use limit orders in thin or fast stocks. You set the price, so the ₹150 surprise in the example can’t happen.
- Stay away from the first few minutes after the market opens, when prices jump and the gap between buy and sell prices is wide.
- Pick liquid stocks, where many people buy and sell, so you can get in and out easily.
- Keep mobile data as a backup, and update your app before a busy day.
- Put no more than 1 to 2 percent of your capital at risk on one trade, and set a stop loss (a price where you exit to cap the loss) before you enter.
- Test any algo on paper first. Backtesting means running your rules on past prices to see how they would have done.
- Add up the charges before you trade often. The calculators hub helps.
Common Mistakes to Avoid
- Blaming speed for losses. Late entries, no stop loss, and oversized positions cost far more. Review those first.
- Trying to beat HFT firms on speed. You’ll lose that race. Hold for longer, where seconds don’t matter.
- Using market orders in thin stocks. You can pay several price levels up. Use a limit order.
- Paying for “ultra fast” signals. Ask whether the provider is registered and how its tool connects to your broker.
- Ignoring charges. Small costs repeat on every trade.
Who Should Use Colocation?
Brokers, market makers, and HFT firms use it. They trade in huge volumes, and the speed gain covers the cost.
It doesn’t suit beginners, long-term investors, swing traders, or most intraday traders. If you’re starting, open a demat account, place small orders, and watch how fills work. If you’d like to automate later, Alice Blue’s tie-up with Tradetron and Quantman is one route. Third-party platforms carry their own risks, so read their terms.
Conclusion
Colocation puts a firm’s server next to the exchange’s systems to cut order delay. High-frequency firms and market makers use it, and the NSE case showed why equal access matters. Retail traders can’t buy it and don’t need it. Your results depend more on how you place orders and which stocks you pick. Begin with small trades and learn how fills work. Look at algo tools only when the basics feel easy.
Frequently Asked Questions
It’s a paid service where exchange members place their servers close to the exchange’s own. This shortens the time it takes for their orders to reach the matching engine.
No. Exchanges offer it to members such as brokers and trading firms. Retail traders send orders through a broker’s app or API.
Yes, exchanges offer it to members under their rules. The NSE case was about alleged unequal access to the facility, not about colocation itself.
A whistleblower alleged in 2015 that some brokers got faster access to NSE’s data than others. SEBI passed orders in 2019, the tribunal set aside the disgorgement in 2023, and NSE settled in 2026.
It’s the delay between sending an order and the exchange receiving it. It hardly matters for a long-term investor. For a firm making thousands of trades a minute, it does.

Blog by Aliceblue News Desk



