Search interesting materials

Sunday, January 19, 2014

The biggest exchanges in the world

There are two ways to measure turnover: dollar value and number of trades. By dollar value, advanced economies dominate the rankings. But the number of trades is more important than meets the eye. The number of trades is a measure of what's going on : 1000 trades/s is a lot more than 100 trades/s. And, as far as the IT complexity of an exchange is concerned, number of trades is all that matters. For the folks building and running exchanges, or consuming a feed from an exchange, the IT complexity is determined only by the number of trades.

Here's data from the World Federation of Exchanges for the top 10 exchanges by the number of transactions (measured in millions):

Exchange2012 transactions2012 rank2013 transactions2013 rankChange (%)
NSE India 14071144913.04
NYSE Euronext 1375211883-13.59
NASDAQ OMX 1268311515-9.17
Korea Exchange 1219410326-15.38
Shenzen SE 93651289237.82
Shanghai SE 92661153424.61
BSE India 35673458-3.07
Tokyo 3508599771.39
London 222921110-4.87
TMX Group 2161023599.15

This shows that NSE was #1 in the world in both 2012 and 2013. With 1449 million transactions spread over roughly 250 days of roughly 20,000 seconds each, this is an average intensity of 290 trades per second. There is no other part of Indian finance where a win of this scale has come about. I used to be nervous about the vulnerability of these achievements, but now the danger has subsided. We have a good equity market and it's now unlikely to get messed up. As the Indian Financial Code gradually falls into place, the equity market will become much better.

The listing underlines the well known fact that the electronic limit order book won. There are no market maker exchanges of note any more. This is a serious problem for academic finance as a lot of our intuition is still rooted in the old world of market makers. I see a striking contrast between the importance of market makers in the finance literature and their irrelevance in the real world. The only game in town is the anonymous limit order book market, and academic finance is weak on it.

If you add up the two Chinese exchanges, there is more activity than the two Indian exchanges. In 2012, the two Chinese exchanges added up to 1861 million transactions, which went up sharply (by 31.25%) to 2442 million transactions in 2013. The two Indian exchanges, in contrast, added up to 1762 million transactions in 2012 but grew by only 1.81% to 1793 million transactions in 2013. In 2013, the two Chinese exchanges added up to a trading intensity that was 36% greater than India. If Shenzen and Shanghai keep up their blistering growth, and NSE stays at low growth, then by 2015 or so, NSE will be displaced from the #1 slot. The question mark for China lies in establishing something like the Indian Financial Code and the rule of law.

This ranking implies that NSE is a great lab for doing research. It is now a bigger exchange than the NYSE and NASDAQ by number of transactions. In addition, liquidity in the US is fragmented across numerous trading venues, which makes is much harder to understand what is going on. In contrast, the order flow for spot and derivatives is largely consolidated. If one can marshal data for NSE and BSE there is 100% coverage. There is no OTC trading and no dark pools. In fact, if a stock is not on the stock derivatives list, it is essentially impossible to take a leveraged position on it. This makes India a clean laboratory where the working of an equity market can be understood.

The US is the ideal lab for understanding what happens when liquidity is fragmented; research projects focusing on fragmentation of liquidity should be done using US data.

The heart of the action in modern exchanges is algorithmic trading. I see one world for the really big exchanges where on average there are over 200 trades/s with peaks of over 10,000 trades/s, and another world for all other exchanges. There is one cadre of finance and IT folks, spread all over the world, who are building these hairy systems around the top exchanges. If you setup a conversation on algorithmic trading between the folks in Bombay, New York, Korea and China, they'd have a lot to say to each other. A new world of financial firms is going to emerge, where a common set of finance & IT skills are deployed across the four locations.

A very large number of transactions yields economies of scale. It is not surprising that the charges in India are low by world standards. This is a competitive advantage in the production of transaction services. If India makes the right moves on the policy bottlenecks, a greater fraction of India-related activity will come to India, and India can be a platform for trading global products. The competitive advantage of NSE and the algorithmic firms surrounding NSE comes from a combination of world class transaction intensity but low revenues/trade. This forces exchanges and algorithmic firms in India to be more intelligent.

No comments:

Post a Comment

Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.