Manipulation: what can we prove?
One powerful theme of the recent stock market crisis has been an attempt at blaming a small group of evil players as being primarily responsible for the events on the market. To what extent might such a characterisation be justified?
The easiest case is a stock which is illiquid. A manipulative cartel develops which rigs the liquidity and price. They actively trade it amongst themselves, to a point where "innocent bystanders" feel the stock is liquid and valuable. This tempts innocent bystanders to step in and buy shares. At this point, the manipulative cartel has made profits because the innocent bystander has been persuaded to part with his money at a falsely elevated price.
There are many variations on this theme. Such efforts often involve collusion with journalists and the senior management of the company, so that glowing stories about the company appear on the front page of the pink papers. Such efforts often involve a short squeeze: where brave speculators, who see the stock is overpriced and sell it, are trapped with short positions they cannot profitably get out of.
In this case, it may be possible for an enforcement agency to pinpoint the culprits and enforce against them. The evidence that can be amassed is: (a) illiquidity of the stock, (b) orders and trades which are highly concentrated within a cartel (c) coordination of trades between cartel members, (d) collusion with journalists and senior management of the company. The ideal enforcement agency would be able to put together evidence about these four issues, and it could then have a strong case.
It is extremely hard to find evidence of coordination of trades between cartel members, except where cartel members are foolish enough to speak into voice recording systems. It is hard to prove links between the cartel and journalists, and the senior management of the company. This leaves us with orders and trades, which are completely observable on the electronic exchange. In the case of highly illiquid stocks, it is sometimes possible to pinpoint the domination of a small cartel in the orders and trades of the stock.
Now, in the recent stock market crisis, the stocks under question have not been obscure illiquid stocks. They have been the most highly traded stocks of the country. This puts a completely different complexion upon the problem. For month after month, we had daily trading in excess of Rs.5,000 crore on the top stocks which are said to have been manipulated. Every single NSE terminal in the country produced orders for these stocks, and records at NSDL will show hundreds of thousands of investors trading in them.
These top stocks were extremely liquid. Liquidity impedes market manipulation: if a manipulator tossed Rs.10 crore of buy orders into Global Telesystems in January 2001, he would move the price by just Rs.2. A rational profit-maximising manipulator would be tempted by the fact that the profit rate in attacking less liquid stocks is higher.
There were hundreds of thousands of economic agents in India trading these stocks; the price formation of these stocks swirled in an ocean of liquidity. I find it extremely implausible when a small manipulative cartel is accused of distorting these prices. The most lurid depictions of the resources controlled by Ketan Parekh and his associates appear to be implausibly small when compared with the depth of the market in that period.
If it can be proved that Ketan Parekh and his associates obtained fake pay orders from banks, or if it can be proved that they engaged in illegal badla transactions, then these are unarguably criminal activities, and they should face the full penalty of the law for this. However, to make a case against them on the grounds of market manipulation is extremely hard.
In judging when a manipulative cartel dominated price formation of a stock, SEBI has tried to use the net outstanding position at the end of the day as the benchmark against which positions of accused cartel members are compared. This is incorrect. Intra-day, the impact of an order upon the price should be judged in terms of market impact cost. When a market order for 10,000 shares is thrown at Global Telesystems, it is possible to use NSE's monthly CDs to compute the magnitude of the price impact, and it is possible to make a case that the market order for 10,000 shares caused a price movement of (say) 0.5%. The end-of-day outstanding position is an incorrect foundation for making a case about market manipulation. This measure ignores the enormous trading activity which takes place intra-day, and it ignores market liquidity. If this measure were consistently applied, most mutual funds would be found guilty of manipulation of most stocks in the country.
Prevention versus cure. Enforcement against manipulation is extremely difficult, and it requires honest, competent enforcement staff. In India, our capacity to produce honest, competent investigations is extremely limited. Hence, it makes more sense for us to push on the frontiers of prevention rather than cure.
- The old BSE floor was a hot-bed of corrupt practices. We could have, in principle, stationed a hundred policemen on the floor to make things better. Instead, it made more sense to switch over to an anonymous, electronic exchange, where such policing is not required.
- RBI's philosophy of favouring non-transparent markets is a recipe for enforcement difficulties. RBI's "negotiated dealing system" (NDS) is a mistake in market design, insofar as it lacks anonymity, and makes it possible for cartel members to coordinate trading strategies. Non-anonymous markets makes it possible for cartel members to observe each other and ensure that cartel members behave in line with pre-arranged conspiracies. Markets should always be anonymous, so that cartel members do not know who they are dealing with, and constantly worry about being stabbed in the back by their fellow cartel members. When X and Y agree to artificially buy on the market tomorrow, X should worry (thanks to anonymous trading) that it is actually Y who is selling to him.
- Rolling settlement makes many classic manipulative episodes infeasible. All trades of Monday are finally and irrevocably settled on the coming Monday, so the build-up of leveraged long and short positions jostling against each other in a strategic game simply does not take place. It is not an accident that our brief experience with the stock market without badla (from December 1993 until February 1995) was a calm period without market disruptions. The elimination of badla, and the move to rolling settlement, are important preventive measures.
- All efforts on improving disclosure impede market manipulation.
In summary, it is perhaps innate in human nature to try to personalise events. However, the equity market is a great achievement of institution building in India. It reflects the views of lakhs of economics agents spread across the country. Each participant only looks at prices and quantities, and does not know who he is dealing with. A lot of what happens on the equity market goes beyond what any small group can engineer. We should take "innocent until proven guilty" seriously, and only accept logically sound explanations.
Back up to Ajay Shah's media page