Search interesting materials

Sunday, September 18, 2011

Paying for liquidity provision on exchanges

Market making versus the electronic limit order book

Exchanges in India all operate as electronic limit order book markets. There are no `market makers'; there is just a publicly visible limit order book. Anyone is free to supply liquidity, by placing limit orders. The person who places market orders is the consumer of liquidity: he pays market impact cost. [A guide to the jargon].

Prior to the rise of the anonymous limit order book, there used to be a great deal of effort on thinking about the market maker. Market makers played a big role in many old markets. E.g. at the NYSE, the `specialist' was obliged to provide liquidity. RBI established `primary dealers' thinking that they would provide liquidity.

These market structures involved complicated problems of measuring the liquidity provision by market makers, correctly compensating them, avoiding monopoly power in the hands of the market market, and enforcing against market manipulation by the market maker. The rise of the open electronic order book cut through this Gordian knot.

For many years, there used to be a debate about whether the anonymous open limit order book market (where anyone can provide liquidity) is better or worse than a market maker market (where limit orders can only be placed by one or more market makers). That debate died down in the 1990s with the success of the electronic limit order book.

Market making on the electronic limit order book

But even on a limit order book, does it make sense to pay one or more market makers to provide liquidity? The public would be free to place limit orders, but one or more market makers would be paid to place limit orders.

The positive argument runs like this. In the life of every contract, at first there is a lack of liquidity as various market participants are reluctant to take the plunge and trade on an illiquid contract. This leads to a chicken and egg problem. Illiquidity inhibits participation, and the lack of participation is illiquidity.

From a regulatory perspectives, exchanges might try to make payments for liquidity provision (or outright turnover) by various underhand means. If that is going to happen, then it is better to have this come out into the open.

But there are also important problems that can come out by going down this route. The resources that an exchange puts into portraying tight spreads or high turnover could potentially be used to improve services for customers. Market participants would make wrong decisions about an investment decision when they see a product as looking liquid on screen, whereas this liquidity is actually artificial: the screen would be falsely portraying liquidity. When exchanges compete on payments to market makers, this can degenerate into a slugfest where the deepest pockets win.

The artificial liquidity pushed by mercenary market makers would tend to lull the exchange into complacence. In the absence of market making, the exchange would run harder to solve problems of market mechanisms and contract design, and to get the word out about the contract.

Recent developments in India

On 2 June 2011, SEBI chose to move ahead with the specification of a `Liquidity Enhancement Scheme' (LES).

By these rules, LES is applicable for individual stocks where the trading volume on the last 60 days is below 0.1 per cent of the market capitalisation. (How would this be scaled to derivatives such as currency futures, where market capitalisation cannot be defined?) I think this makes sense. The LES would be used to kickstart liquidity when it is abysmal. The moment a small amount of liquidity comes about, the LES would step aside.

Based on these rules, NSE announced a program for market making on the derivatives products recently launched at the exchange: on the S&P 500 and the Down Jones Industrial Average (launched in partnership with the Chicago Mercantile Exchange). These incentives are over and above the absence of charges by the exchange. I was disappointed to see a payment based on mere turnover. This would give the market maker an incentive to do circular trading and thus show a lot of trades. But turnover is not liquidity.

This program came into effect on 15 September. It may matter more in the coming week, given that new contract series start trading from tomorrow.

Will it matter? How will we know that it mattered?

Derivatives on the S&P 500 and the Dow Jones indexes have gotten off to a surprisingly good start, even though there was no such program. This has perhaps been helped by unusual levels of volatility in the US after the launch of these contracts.

The early days of a contract can be a rollicking ride and even after these time-series fall into place, it will not be easy to tell whether LES was useful in the history of these contracts or not.

Similar thinking is taking place at BSE also: See Will BSE's biggest initiative work? by Mobis Philipose in Mint. The text there -- obligations such as providing two-way continuous quotes within specified parameters for quote size and spread -- sounds good, but here also there are payments per crore of turnover. By and large, the payments being made at BSE look much bigger than those at NSE.

In the case of BSE, if LES is able to lift BSE out of zero market share in derivatives trading, even after the six month period has expired, then it would be a clear proof that the LES helped. So this experiment is unlike that of NSE where it will be hard to evaluate whether or not the LES mattered.


  1. BSE has provided for payments based on active and passive orders even in the Market Making framework. It will be useful if you could read the BSE circular in detail and comment. Current blog seems to be based on super-ficial newspaper readings of the subject as quoted in the blog.

  2. NSE started the program suggests that their trading on Dow and S&P indices did not remain liquid. Neither, to start with there was much liquidity.

    BSE's incentive scheme provides for active and passive order based payments and not simplistic payment on turnover with fears you have espoused. In essence, you seem to have a-priori position which is being justified with proofs contrary to the position using obfuscation.


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.