What is it about high frequency trading that makes one rich overnight? Is it legal? If so, why is everyone not doing it?
I was always fascinated with Michael Lewis's writings... His Liar's Poker was truly a path breaker, giving fantastic insights into what happened in Wall Street in the 80s. On one of my TV channel surfing experiences, I stumbled on to an interview of Brad Katsuyama by Maria Bartiromo in Fox Business Channel on his new IEX exchange. In the course of this chat, I learnt that Michael Lewis had written about Brad and his company IEX in his new book titled ‘Flash Boys’ and that got me interested. If there is anyone who can explain a complex topic in simple, easy-to-understand terms, it is Michael Lewis.
I left for India shortly thereafter and while in Bangalore, picked up Flash Boys. I had intended to read it but got caught up with other stuff and could not get to it. On my flight from Bangalore to Mumbai, I came across an article of how 2 friends moved back from Wall Street to a tiny office in Bandra-Kurla complex and are generating business worth Rs4,000 crore a day using high frequency trading strategies. The numbers were truly mind boggling - even if they were generating 1% profit, that works out to Rs40 crores a day or assuming 200 trading days in a year, Rs8,000 crore (about $1.25 billion)! It was one of those moments, when you start asking yourself if you were in the right profession. I filed away this information for future reference as I flew back to the US.
I finally got time to pick up Flash Boys and got hooked right away. Despite being a work of non-fiction, I could not stop turning the pages. When I put the book down, I started thinking as to how many other countries are now having high frequency trading (HFT) and I suddenly remembered the article about RKSV and their amazing turnover as a discount broker in Mumbai. What is it about high frequency trading that makes one rich overnight? Is it legal? If so, why is not everyone doing it? But first, let us take a step back and understand the evolution of high frequency trading...
On 19 October 1987, the Dow Jones Industrial Average crashed by 22.6%,
the single largest one-day stock market decline in history. Many reasons were given for the loss, one of which was how globalization was affecting financial markets across the world. In those days, a stock was bought or sold using a broker whom you called up (or he called you!) and agreed on the stock, its price and how many. As the market crash was playing out, it was rumoured that many brokers stopped answering their phones! The US Securities Exchange Commission (SEC) addressed this and several other structural flaws and one of the remedies suggested was to have computerised trading. This set off a huge wave of modernisation and computerisation in Wall Street, which continues till date.
NASDAQ (and then NYSE) went public in 2000 and had their own profit targets and started finding ways to increase revenues. To increase competition, the US government also allowed setting up of more stock exchanges and by 2010, there were 13 stock exchanges. Some of these, were setup by brokers and high frequency trading companies, which should have raised red flags (after all these are intermediaries and ideally stock exchanges should be run by those for whom they exist, i. e. investors) as Michael Lewis observes in
this inteview, on how these 13 Public Stock Exchange have become an unfriendly place for the investor.
To be fair, there have been many benefits because of computerisation too. One of them is the cost of commission per trade. However, as technology advanced, faster computers and networks enabled some firms to start trading between computers based purely on algorithms (an algorithm can be thoughts of as a series of steps to solve a problem). At first, it seemed like all was well. In 2006, approximately 26% of all trades in the US were done by HFT firms and by 2010, this had reached 52%. What boggled the investors’ minds was how HFTs could have
4000 days of trading with NO losses!
By 2008, many institutional investors, such as Royal Bank of Canada (RBC) started observing that even simple trades were costing them more than before. They started digging into this and uncovered a remarkable truth. It was as if
HFT traders were lurking, waiting for them to put out an order and then act on it immediately, almost in a predatory fashion! What was even more frustrating was that they could do nothing about it. Brad Katsuyama, who headed the RBC trading desk at its New York office was determined to get to the bottom of this and with the help of other specialists, narrowed down the cause as being due to HFT companies. This led to his formulation of a process, which would make it fair for everyone and that is how IEX came about. However, more on HFT... I located this
excellent description of how High Frequency Trading works on YouTube. It is only 11 minutes long and I urge you to watch it because it will help you understand this article better.
For those of you who are interested in knowing the state of US Stock markets today, there is an ongoing investigation by New York State Attorney General's office and by the Commodities Futures Trading Commission in Washington DC on high frequency trading practices. Ben Bernanke, the former chairman of Federal Reserve, announced last week that he is joining Citadel Group, one of the biggest HFT companies, in an advisory capacity. Please note that
TD Ameritrade, one of the biggest online trading platform sells all its orders to Citadel Group! If an HFT is willing to buy all the orders from TD Ameritrade, there must be a very good reason. Today, in the US, HFT traders are essentially driving the market. For more, see the diagram below:
- HFT companies pay brokers to buy all their trades (e. g. Citadel Group buys all orders of TD Ameritrade)
- HFT companies pay stock exchanges huge sums of money to co-locate at the stock exchange in order to minimize the delay in monitoring orders
- Stock exchanges (since they need to generate profits) pay banks and institutional investors to direct orders their way. Sometimes, it happens the other way too!
Where is the investor in this picture? Was the stock exchange not established so capital could be directed to the company that needs it the most? So why are HFT companies paying brokers and stock exchanges? The answers to these questions can be found in Michael Lewis's book. The explanation is too long for this article. The number of Americans trading in stocks and derivatives has fallen to 52% from 63% about six years ago. The average investor is losing confidence in the market.
Now let us see how all this applies to Indian markets. HFTs make money in several ways and here are some of these:
1.
Front running - In a simplistic way, this is similar to scalping movie tickets. Let us take a real life example... Infosys trades on
NSE, BSE and several other exchanges in the country. An institutional investor (II) wants to buy for a mutual fund (MF) company about 10,000 shares of Infosys. The MF company, stipulates that the range of the purchase of the 10,000 shares be between Rs2,150 - Rs2,200. The institutional investor's trading desk runs an algorithm whereby it tries to buy a small lot (since it does not want to reveal its hand) of 100 shares at Rs2,150. Assume that the II is located close to NSE, and therefore the order appears first at NSE.
• An HFT, co-located at NSE, picks up this order at Rs2,150 and then tries to gauge the upper limit of the II. It may try selling to the II at Rs2,250 and the II computer will reject it (because it can go only as high as Rs2,200). The HFT computer lowers the price to Rs2,225 and gets rejected again. Then it tries Rs2,200 and II buys the order!
• Now the HFT knows the upper bound of the II and will out run the buy request of the II to all the other exchanges and buy up INFY stock and then sell it back to the II, all at Rs2,200, which the II accepts since it does not find any other lower priced shares. All this happens in a matter of milliseconds because the speed at which data runs between exchanges is slower than that of the HF traders.
• Who loses out? The investor, who has put money in the mutual fund hoping that they will do a good job of managing his investments.
• The profit margin may not be big but if you can do this on EACH and EVERY sale, the numbers do add up and that is what HFTs do.
2. Dark pools - Some of the HFTs get both sell and buy orders of stocks and instead of routing them to a stock exchange, do the trading themselves, thereby not exposing the real value of a share in the market. Regulatory bodies will not even be aware of a sale of this kind. There are 45 dark pools in the US, with companies such as Goldman Sachs and Morgan Stanley having at least one such pool.
This is not to say that all HFTs are doing this but it behoves the Securities and Exchange Board of India (SEBI) and Reserve Bank of India (RBI) to create a fair trading platform.
My suggestion is to create an exchange designed along the lines of IEX. IEX opened for business in 2013 and now does about 1% of all US stock transactions. They offer only three types of trade - Market, Limit and mid-point. IEX does not allow co-location nor does it offer any speed benefits for HFTs and therefore levels the playing field. Some of the investors in IEF are Bain Capital Ventures, Belfer Management, Brandes Investment Partners, Capital Group, Cleveland Capital Management, Franklin Resources Inc, Greenlight Capital, MassMutual Ventures, Maverick Capital, Pershing Square, Scoggin Capital Management, Senator Investment Group, Spark Capital, TDF Ventures, and Third Point Partners.
Goldman Sachs announced its support to IEX and even sent some business its way for a few days (Dec 18-19, 2013) before they changed their mind and went back to their (good or bad) ways. Moreover, one more thing - the new exchange (let us call it FAIR - Fair and Intelligent Retail platform) should be located geographically in the middle of the country so as to be equidistant (more or less) to all cities... Did I hear someone say Nagpur?
Final Note: If this subject seems complicated, I encourage readers to click on the URLs which have a lot of information. This video
stopped trading at US markets as it was playing out!
I had the privilege of being part of this team. This was my first hand exposure to both the workings of a stock market & algorithmic trading.
over-estimate. Profits in HFT are of the order of 0.001%. With high volumes that can add up to a respectable figure. But nowhere near his number.