Step 10. Clearing Up the Trade
Step 10. Clearing Up the Trade
To settle the trade, the parties involved use a clearing firm.
Funds are moved from the buyers account to the sellers account.
The buy-side firm uses a custodian as an intermediary for the clearing of the trade.
SOURCE: Cisco Systems
Step 9. Reporting Back
Step 9. Reporting Back
After a trade is executed, the matching engine sends a notice of execution report back to the sell-side firm.
Step 8. Going Direct
Step 8. Going Direct
In some cases, a sell-side broker gives its customers (the buy-side trading firm) direct access to the market.
In this case, the buy-side firm triggers the trade.
The buy-side firm uses the sell-side firms infrastructure to perform the pre-trade risk check and then execute the order.
Step 7. Getting Closer
Step 7. Getting Closer
To speed up the matching process, a sell-side firm places its strategy engines and pre-trade risk systems at the execution venue.
The trade in that case goes to the matching engine via an exchange cross-connect.
This is a high-speed connection between co-located market participants and the exchange systems.
Step 6. Matching the Order
Step 6. Matching the Order
The order is sent to the matching engine of the execution venue.
The engine pairs sell orders with buy orders of the same quantity; and, vice versa.
Step 5. Placing the Order
Step 5. Placing the Order
The order is sent in FIX format to a Financial Information Exchange (FIX) gateway at the exchange or other execution venue.
The venue sends back an acknowledgement message.
The round-trip time for the acknowledgment is measured by monitoring tools to determine if the execution venue is fast enough.
The order management system might cancel the trade, if the round-trip time is too high.
Step 4. Processing a Trade
Step 4. Processing a Trade
If it passes the risk check, the trade is sent to an Order Management System.
The system logs the order and passes it to a computer known as a smart order router. This system chooses which exchange or venue in which to execute the order.
The decision is based on price, available liquidity, speed, cost and other criteria.
Records of the criteria used and result achieved are kept for compliance reasons. The idea? Prove that the best (possible) execution of the order was achieved.
Step 3. Triggering a Trade
Step 3. Triggering a Trade
A trading engine processes the data and triggers a trade if pre-set conditions are met. The trade goes to a risk management application, as a checkpoint.
Step 2. Sucking In Data
Step 2. Sucking In Data
The buy-side and other users of market data pull in the data from their own market data messaging buses.
The consumers can be strategy engines, algorithmic trading applications, pricing engines and pretrade risk management applications.
Buy-side firms take in the data to manage risk and model portfolios, generally.
Step 1. Pushing Out Data
Step 1. Pushing Out Data
Exchanges, alternate trading venues and market data providers send out pricing information to market participants.
NYSE Euronext, Nasdaq OMX Group, Thomson Reuters and Bloomberg are all major suppliers of high-speed data on trades.
Market data is taken in by feed handlers at a sell-side firm.
The data gets cleaned up and sent back out to the buy-side and other consumers, using a market data messaging bus on a server.
Execution Time: The Ten Steps in a Trade
Execution Time: The Ten Steps in a Trade
BATS Global Markets now acknowledges orders in 200 millionths of a second or less. Nasdaq OMX Group says it has deployed technology in Scandinavia that takes that down to 100 microseconds and 1 million messages a second.
All told, sell-side firms will spend $14 billion this year to speed up how fast they can execute orders, according to Tabb Group. At stake: Millions of dollars in trading profits, on an annual basis. Per millisecond.
Heres a step by step look at what goes into high-performance execution of a trade, in every fraction of a second.
SOURCE: Cisco Systems, High Performance Trading Fabric