Raghu Kumar Raghu has been featured on Forbes’ annual “30 under 30″ list. He handles RKSV’s (www.rksv.in) PR and media communication, manages RKSV’s algorithms at its proprietary trading desk, and works closely with the business team. A numbers freak, he’s always looking to find ways to apply statistical analysis into his work. He’s an avid fan of classical Carnatic Indian music, teaching his two dogs new tricks, and working out at the gym. His goal is to one day venture into sabermetrics and to also actively get involved with promoting sports in India.

# Understanding High Frequency Trading (HFT)

Many have read the book “Flash Boys” by Michael Lewis, which gives the reader an insider’s view into the murky world of High Frequency Trading, more commonly known as HFT (and if you haven’t, I highly suggest that you do!). Here, I’ll try to break down what HFT is, how it works, and its prevalence on the India’s capital markets today.

## It’s a Speed Game

HFT, on its own, is exactly what the name suggests- extremely fast, high frequency trades. Instead of placing a trade per second, High Frequency Trading firms have reached a point where they can place multiple trades within nanoseconds. Not milliseconds (1/1000th of a second) or microseconds (1 millionth of a second), but nanoseconds- that is, 1 billionth of a second.

So how does one do High Frequency Trading? Well, as it stands now, it’s not quite that easy to do high frequency trading. Typically, it involves the following.

1. Co-location: HFT firms try to deploy their servers (basically super fast computers) as close to an exchange as possible. The method they do this is by having their servers co-located at the exchange.Each exchange typically has a collocation center, including the NSE and BSE. Firms rent out space for their servers in order to get their servers into the exchange’s collocation “rack”. This gives them an advantage over regular traders who are not as close to the exchange.
2. Algorithms: HFT firms deploy algorithms that are pre-programmed by extremely smart (and well paid!) programmers who understand the markets and look for strategies to deploy in a high frequency fashion.There are all sorts of algorithms that HFT uses, including market making, statistical arbitrage, news trading, or to put it even more simple- any strategy that the programmer finds profitable. Due to the fact that the firm is getting market data faster than non HFT players and can send orders faster than non HFT players, HFT algorithms will be able to beat a non HFT algorithm any day of the week. In a nutshell, if the strategy can be coded, it can be programmed arithmetically and the HFT firm has an automatic edge.

Let’s take an example. Suppose the Reserve Bank of India announces a GDP figure that beats analyst expectations by a huge amount. All traders are expecting the markets to go up. HFT firms will be able to process that information and fire off trades within nanoseconds, grabbing the lowest market price before others get to it. By the time the human, sitting in front of his terminal, is even able to process the information- the market has already shot up and the HFT firm is earning a profit.

3. Automated trades: This is where HFT really has a huge edge over retail investors and traders. By constantly scanning the markets for discrepancies in pricing, HFT can deploy arbitrage, market making, and sorts completely automated strategies. It might sound like this is unfair, and one can make a strong argument for that case; however, from a price discovery point of view, nothing beats HFT in ensuring that the price of a stock, future, option, or any financial product is priced efficiently and as quickly as possible. Think about it this way: would you rather buy shares of Reliance shares at Rs. 1001 when it should be priced at Rs. 1000? HFT firms would go in and immediately sell Reliance at 1001 and buy it back at 1000; this, in turn, helps investors who are looking to genuinely purchase Reliance shares as a buy and hold strategy for the long run.

## HFT in India: Understanding the Differences

Many try to group HFT with algorithmic trading; others try to group it with Flash Trading. Neither is completely accurate.

HFT falls within algorithmic trading, and Flash Trading falls within HFT. There are all sorts of algorithmic trading strategies being employed in India, and HFT certainly does not comprise a lot of it.

1. Algorthmic trading is actually done mostly by FII’s or fund managers looking to purchase a large quantity of a specific asset. They will inform their broker that they are looking to execute X number of shares at a price of Y or below; in return, the trader working at the broker’s desk will use DMA (Direct Market Access) to setup an algorithm using the fund manager’s preset rules (for example, to only purchase trades in slices of 1000 shares at a time). In other words, speed is usually not the requirement. The fund manager wants to ensure that his trades get placed within a specified period of time for his investors. Many times, he does not care whether the firm employs automated or HFT technology.
2. Automated trading, which falls under algorithmic trading, is typically used by hedge funds of proprietary trading funds. They look for patterns in price, quantity, and other market information to generate automated, algorithmic strategies. even for these guys, ultra fast speed is not necessarily a requirement.
3. HFT falls under automated trading, and as written earlier, is dependent on a fast price feed. Speed is the name of the game for HFT. Not only are the strategies almost always automated, but every second, millisecond, miscosecond and sometmes, nanosecond matters. HFT firms often compete with each other to see who can get their order to the exchange the quickest.
4. Flash Trading: Finally, we get to the big elephant in the room. Flash trading was prevalent in the West among HFT firms. To get a good understand of how it works, reading “Flash Boys” gives a good glimpse on how it worked. In a nutshell, Flash Trading allowed some firms to get an unfair advantage over other firms by being able to get market information fractions of a microsecond/nanosecond faster than other and reacting to the information. The result was that the US markets were not an open playing field. Fortunately for us, Flash Trading has never been an issue in India because SEBI never allowed such types of unfair practices to be granted permission.

So, in conclusion. If you ever hear of people complaining about HFT and grouping it with “Flash Orders”, you can immediately inform them that while HFT is prevalent, flash trading is not allowed in India. And seeing how the US and European regulatory bodies are clamping down on Flash Trading in a massive way, it is safe to assume that Flash Trading won’t be a problem in India anytime in the near future.

And if he’s still not convinced, ask him to read “Flash Boys”. That should do the trick.

About the Author: Raghu Kumar is a Promoter of RKSV, the fastest growing low-cost brokerage firm in India. RKSV’s clients process close to 4500 Cr. of turnover daily Pan-India. RKSV offers free trades and also allows investors and traders to do trades for Rs. 20 per order executed, also also offers unlimited trading prcing plans for active traders. Learn more by visiting www.rksv.in

Raghu Kumar Raghu has been featured on Forbes’ annual “30 under 30″ list. He handles RKSV’s (www.rksv.in) PR and media communication, manages RKSV’s algorithms at its proprietary trading desk, and works closely with the business team. A numbers freak, he’s always looking to find ways to apply statistical analysis into his work. He’s an avid fan of classical Carnatic Indian music, teaching his two dogs new tricks, and working out at the gym. His goal is to one day venture into sabermetrics and to also actively get involved with promoting sports in India.

## Finfolab Technologies developed API Solutions for Goodwill

Finfolab Technologies developed Arrow API solutions for Goodwill. Our API solutions enable retail traders and Goodwill commodities partners to design and deploy advanced algorithmic...

## Angel Broking Multi-Client Amibroker Trading Module

Got frequent requirements from Angel Broking Customers to automate Amibroker based trading system where one single generation from charts could punch orders in multiple...

## 12 Replies to “Understanding High Frequency Trading (HFT)”

1. Raghunath says:

Hi ,
Can retail investors do automated trading like statistitical arbitrage, or basket trading like option strategies. does NSE or BSE prevent it. If yes Who offers it, is it linked to platforms like amibroker or other trading platforms.

If no are brokerages planning to bring automated trading to the retail traders

1. Hi Raghunath,

It’s a very grey area at the moment. In a nutshell, unfortunately, regulators are not allowing fully automated trading for retail clients. Once they do allow it, brokers will begin linking the functionality to their platform with the help of APIs.

And yes, RKSV (I’m a promoter of it) is working diligently on it to get the necessary approvals from the exchange. Right now, we are trying to get the full approval to do fully algorithmic trading through charts (this won’t take more than a couple weeks). As with everything else when it comes to regulations, it’s a step by step process 🙂

Best Regards,
-Raghu

2. Shailesh says:

If retail traders will be allowed, how much initial investments would be required to start HFT?

1. Hi Shailesh,

It depends on the type of strategy you run. You can probably do algorithmic trading (non HFT) with an amount of around Rs. 20,000 to get things started. For HFT, the amount would probably have to be in the range of lakhs since you’ll be looking to place a lot of orders within a small amount of time.

Best Regards,
-Raghu

1. Shailesh says:

Thanks a lot sir for response

One more thing I want to know is that, I am living in Ahmedabad & I want to start HFT online?

Can you please send me any contact number?

1. HI shailesh,

Just like that one cannot start doing HFT. It requires a co-location server near to exchange server. Understanding about micro level markets, market participants. Deep quant thoughts , mathematical and statistical skills, proper risk management, understanding about your competitors and tons of Exchange and SEBI regulations.

Only Institutional traders qualify to do HFT trading. For an ordinary man he can play on with non HFT based autotrading strategies.
If you are very new to non HFT based autotrading start here Understanding Autotrading in NSE Future markets

1. Shailesh says:

Thanks a lot sir,

Any kind of contact number to reach you as I want some more details

3. Raghunath says:

Hi Raghu,
This question deviates from the HFT topic. FII are currently 22.7% invested in Indian market. with regulation and FII limits of trading in many counters they are trading in derivates. How much much more cash market position They are allowed to take. Are FII limits reaching 90% of the allowed limits for FII.

Can FII further take 10% in Indian equities for listed companies. Does this give room for another 50 billion of positive inflows

4. Hi Raghunath,

The RBI closely monitors FII investments on a daily basis. There are some key takeaways from RBi’s policies.

1. Although the limit by default is that FII’s can invest up to 24% in Indian firms, the actual cap is 2% points lower, meaning 22%. Once 22% is reached, banks across India are cautioned to not purchase more shares of that company unless the RBI gives the approval.

2. 89 companies fall under the 24% ceiling. Out of the 89 companies, none have reached the 22% limit.

3. The RBI has increased the ceiling (usually based on the sector) for many other firms. For example, 19 firms have the ceiling at 30% (including Asian Paints, for example); 14 firms have the ceiling at 40% (such as Hero Honda), and another 16 have it at 49%, such as Zee Entertainment)

4. Out of all these companies, a small handful have hit their upper ceilings; the large majority have not.

So, in a nutshell, I don’t think FII limit breaching is of a major concern at the moment. These last two months of the fiscal year should see strong FII inflows.

Best Regards,
-Raghu

5. rajeev says:

hello sir
is there any software by which we can see hft firms buy and sell signals.

1. HFT Firms Buy and Sell Signals? First try to get the concept of HFT they are different game players(Market Makers , Arbitrage).

There is no point of analysing what they are doing in the market. They are just liquidity providers.

6. Barun Kumar De says:

Recently FPGA based platform is getting traction. The biggest advantage of FPGA based HFT trading platform is one can create unlimited parallelism. In any processor based trading platform maximum number of operation can run is limited by the number of cores the CPU has. In case of FPGA, it can go to infinity. Hence even when there is a burst of market data FPGA based platform will not miss any data

Regards,
Barun