It is trading by brokerage firms, hedge funds and to a lesser extent mutual funds using computer algorithms. They seek to optimize some function that is a proxy for profit, given whatever parameters they believe represent the real world. It happens so quickly that ordinary individuals should not be capable of outperforming the computer driven orders.
For example, if you were a firm with a goal of providing liquidity to the market for a profit and in your experience XYZ's trading range is lognormally distributed with an expectation of $10 per share. If the price exceeded some limit you would either step in and buy or alternatively sell or short the stock. While the goal is to be the counterparty to as many trades as possible, if the price deviates too much from the expectation the algorithm should execute trades to drive the price until it is close to the mathematical expectation eliminating any arbitrage opportunities that are created by unusual orders such as an order for 10,000 shares in a lightly traded stock. If you have priced the transactions well, you would trade the liquidity risk for the reward from buying those shares at a discount. The algorithm would guarantee that no one could hit the enter key faster than it could generate an order except another algorithm. The quality of the algorithm is strictly determined by the quality of the programmer and quality of the modelling done regarding the market's microstructure.
2007-04-09 09:15:31
·
answer #1
·
answered by OPM 7
·
0⤊
0⤋
I have a PhD in finance and learned Market Microstructure from A.S. "Pete" Kyle -- the guy who created the field -- and I have no clue what OPM just said in his response to you.
The theory behind algorithmic trading is simple. If you are buying or selling a big block of stock, you could move the market (alter the price) just by executing your trade.
Suppose that you are selling. If you are observed making a big sale, then the traders who are buying will think that either you know something bad about the stock that they don't know or that you have to dump the stock. Either way, they will low-ball you and try to get the stock at a cheap price.
However, there may be other people who want to buy or sell the stock, too. If there happens to be a lot of people buying that day, then their demand masks your need to sell. The fact that there are multiple sources of uncetainty in the market allows you to mask your need to buy or sell.
If volume is low, you may want to sell off at a slow pace in order to hide information about your trade. But if volume is high, you can trade more quickly.
Algorithmic trading allows you to enter a number of inputs about market conditions and then the algorithm tells you how quickly you can sell your position. If these conditions change during the course of trading, the algorithms adjust the speed of the trade.
There are all kinds of algorithms -- but this is the basic idea.
2007-04-09 09:34:10
·
answer #2
·
answered by Ranto 7
·
0⤊
0⤋
Hi,
Algorithmic Trading simply put is making a computer program of your (mechanical) trading system / strategy. Many time this term looks very big to people, however it is not. Algorithmic Trading is also known as Automated Trading / Computerised Trading.
For more information on this, please visit
http://quantmaster.in/QuantitativeFinance.aspx
Regards,
Gaurav
2013-09-30 20:44:41
·
answer #3
·
answered by gaurav b 1
·
0⤊
0⤋