Wednesday, March 7, 2012

FAQ On High Frequency Algorithmic Trading

FAQ ON HIGH FREQUENCY ALGORITHMIC TRADING (HFAT) IN FINANCIAL MARKETS

Q1.Why must I know more about HFAT? Because if
you don't, you will be ignoring a major revolution that is going on in the financial markets and affects whole economies as well as individuals like you. HFAT is growing rapidly in market volume share all over the world. In America it now accounts for > 53% of all trades, in Europe 28%, Asia 18%. By asset class it accounts for 65 % of equity trades, 50 % of futures, 40 % of options, 25 % of forex and 20 % of fixed income instruments.

Q2.What does Algorithmic mean.?
Algorithmic simply means a set of instructions/rules. In our IT age it means having a software program that executes these instructions automatically. Algorithmc does not automatically equate with high frequency. Many fundamentals-based institutional traders also use algorithms merely to automate their trades. These could be stop-loss Algos, Algos that break up big blocks of buy/sell so that they won't affect the market, or arbitrage Algos that classify items with similarities and buy/sell when one of the items is 'out of step' with the rest in terms of valuation.

Q3.What does High Frequency mean?
'High frequency' trading means trading very often. Since the trading day is limited in hours, it also means high speed trading

Q4.How frequent is 'frequent'and how fast is 'fast'? Frequent and fast is about 1000 trades per second. To get an idea of the speed note that many HFAT traders site their servers as near as possible to the stock exchange so that they can shave a couple of micro-seconds of their transaction speed.

Q5.what strategies do HFAT pursue?
HFAT systems can pursue various strategies, including arbitrage, momentum, deception, predator-prey strategies

Q6.Can HFAT predict the markets?
In the sense of recognizing patterns and extrapolation into the future, the answer is yes. Nobody can predict tomorrow's price. But using complex non-linear mathematics it s possible to sometimes predict the price in the next hour. This assumes that nothing major, no external shocks happen in the next hour to affect the price. And this IS a reasonable assumption to make even in our fast-moving world. In any case, to a seasoned trader, such a prediction is very useful as a guide or confirmation of his view of the market.

Q6. Are HFAT systems profitable
? They can be or may not be depending on how they are used and who uses them.

Q7.What are the effects of HFAT on financial markets?
HFAT systems accentuate market volatility, and throws markets out of step with economic fundamentals for a longer and longer period. So stock market indices are not so much of an indicator of a country's economic health now. On the plus side, HFAT systems increase market liquidity and narrows the spread between buy and sell prices.

Q8.can people with no financial background or no trading experience use HFAT to make money?

NO. HFAT systems still need a human to manually override them sometimes. This is because a HFAT Is competing with many other HFATS in a war of the machines, and machines being machines, their interaction and response to each other may cause undesirable effect loops as they cannot interpret events in any context beyond the set of scenarios programmed into them. Like the mini earthquakes that happen every day, mini flash crashes caused by HFAT happen all the time. But they self-correct and go unnoticed until a really big one happens, to cause systemic disaster in the markets.

Q9.Can HFAT systems be built by anybody? if you have mathematical, statistical knowledge, and modeling experience with regard to nonlinear dynamic systems, there are off-the-shelf DIY kits that are available. Real time data is also available from vendors.

Q10. What does the future hold for HFAT? The rise of the machines is inevitable. That is why SGX is installing the world's fastest trading platform to attract Algo traders.

Q11. What must I do to adapt to the rise of the machines? Trade short term, emphasize less on fundamentals and learn technical analysis . TA cannot predict, but by filtering, compressing, de-noising, normalization etc technical indicators enable you to better visualize the data. The days of buy and hold based on fundamentals are over, unless you are Warren Buffet and plan to live for a very long time.

No comments:

Post a Comment