Skip to main content

High Frequency Trading...

HFT has been the way to trade for quite a while now in the developed markets. Stock markets have evolved from physically transferring papers between participants in closed rooms called 'Market Place' to now being capable of electronically trading shares in plain nano-seconds. While a few years ago, a sophisticated market participant was one who adopted disciplined investment strategies considering fundamentals and valuations of companies, investing with a time horizon, with patience being the rewarding factor, today its the quants that are considered sofisticated scraping alpha with the use of quantitative strategy models, identifying patterns of trades programmatically, interacting with the market through algorithmic trade execution systems, etc.

Stock exchanges in developed countries are being pounded with millions of trades in matter of seconds in the name of HFT. While the biggies in the trading community have systems of robotic compound that have the capacity to push ultra-low latency trade orders in tens of nano second intervals comparable to the speed of light, even the most developed stock exchanges in the world are only capable of executing trades in milli-seconds at most. And the asian markets are much slower. Although the technology stirs up a wow-effect, the crucial point that goes unnoticed here is that, the computer/exchange connectivity at my home is not one of those high-tech massive parallel processing systems. Neither is yours! So, it makes it so easy for these biggies to execute trades at that right moment they want to, taking the best advantage of the technology that is not available to the common investors who try to sneak in a trade here and there amongst this hostile gush of data.

An example of the advantage of HFT is, consider an LFT system placing an order for 10000 shares at whatever the current market price is. The HFT system can actually see this order before it becomes visible to the market and immediately grabs 10000 shares from the market at the current price and also places the sell order for the same 10000 shares at a price higher than its buy price. It may not do this for every market order it sees, but in the few milli seconds that it has, it finds out if it would be worthwhile taking a position. All of this happens before the LFT system's order hits the market, so ending up with a small premium per share from the LFT system without even its knowledge, a highly covert behaviour leading to its alias, 'Flash Trades'.

Another example could be an Algorithmic/Black Box system that places automatic low latency trade orders based on its program to identify buy/sell signals from the real-time prices before those signals are even visible to LFT systems.

When the exchanges grow up further to support the capabilities of HFT to their fullest, that might be an era which would simply outcast the LFT players from making profits out of day trading strategies or at least minimize to a great extent as they will mostly be paying a premium to the HFT systems. Brokerage houses may be forced to have HFT trading systems to become more competitive and probably will charge a premium from the customers availing the technology. And at the worst case, LF traders (You and I) would be forced to either speculate or remain conservative for medium to long term growth, which is why HFT is highly debatable, yet bound to conquer!

Comments

  1. Though speed matters, what is more important is the research done before algorithms are written. So, even if it's a HFT, people who can do brillaint research to support writing effective algorithms play a vital role in any changing mode of trading.

    ReplyDelete
  2. So true! That's the belief that keeps our jobs, probably! :)

    ReplyDelete

Post a Comment

Popular posts from this blog

Hedging Currency Risk

"How to have my foreign currency income converted at a fixed exchange rate vis-a-vis INR?" Exchange rates have been a worry for a lot of us for a variety of reasons. The most common cause of concern with foreign currency income is the fluctuating nature of the currency which at a lower exchange rate would translate into fewer Rupees. This is no brain-jammer, its relatively a simple problem to deal with and that's exactly what this post is all about. This is mostly useful when the Rupee is already quite depreciated, i.e. when you get more of INR for every USD as is the case now. The current Ex Rate is Rs. 46.81 / USD. Now, if you are happy with this exchange rate and are not willing to take a risk with the possible Rupee appreciation, you can even lock your future USD income at this rate. That is, even if the rupee appreciated to Rs. 45/USD when you get your next month salary, you can still have the USD converted at Rs. 46.81! Basically, when Rupee appreciates, it mean...

Nifty @ 4000

4000 levels on Nifty is a good support and it should be hovering around it for sometime, unless something drastic triggers in either direction. But I don't see any major activity or release of news/information in the near term calendar. In times like these, best strategy would be to sell an option so that its value would be decreasing due to the Nifty staying around the same levels without major movements and the decrease in the Days to Expiry. An option price depends on 5 factors out of which the underlying and the Day to Expiry are 2. So, now you may sell a well out of the money Call option, take a premium and square it off in a few days when the market is at the same levels, as its value would have decreased due to the lesser number of days to expiry. You shouldn't be waiting too much to book your profit, as the sentiment could turn other way over night. It's a very common mistake that people do, taking the first position at the right and miss to square off that poistion...