The failures of markets

The early glimmers of economics were about the virtues of markets. The first economists understood that trade was good; that voluntary exchanges only made people better off; and that much of what seemed to call for dictators ordering people around could actually be done quite fine by just having people engage in trade with each other. It was a profound and compelling vision.

Financial markets are seen as the economist's ideal market. They have a large number of buyers and sellers; they have well defined and standardised commodities and come closest to the notion of a "frictionless market".

I was personally most disappointed when, somewhere in the early 1990s, I realised how difficult it is to actually obtain frictionless financial markets. It turns out that obtaining a sensible and liquid securities market is remarkably hard. We can divide the questions into two parts: the largest and most important products (e.g. the shares in Nifty or Nifty junior; or futures/options on Nifty; or the dollar-rupee; or a government bond) and the rest.

For the largest and most important products, the bulk of the difficulties are "man made"; they are caused by bad market design. There's a lot that we can do in making the Indian equity market more liquid, but it just isn't getting done. It isn't hard to obtain a liquid market for government bonds, but we persist with a bond market designed for illiquidity.

In these areas, in the late 1990s, I realised that even though getting to liquid and efficient markets isn't technically difficult, the political economy is often all wrong. There just isn't enough of a constituency that wants efficient markets, while there are major benefits to market intermediaries and other players from having inefficient markets.

For the remaining products, the best efforts that we can make still leaves securities in a state of illiquidity. I can imagine a world in which the top 1000 shares of India will trade fairly well and obtain decent market efficiency. The remaining 6000, in my judgement, are doomed. I don't know of a market design which will make these markets function sensibly.

This is a very troubling vision. Looking beyond finance, what about all the physical goods? If we can't trade many shares frictionlessly, we have little hope for the wide variety of ball bearings, tyres, organic chemicals or paints. For all physical goods, the frictions in trading can only exceed the worst securities. The labour market and the marriage market are beset with huge transactions costs, and seem to routinely malfunction. When we think of the usefulness and appeal of markets, we are faced with a tiny universe of products in the economy where "the market" actually achieves high levels of liquidity and market efficiency.

In this situation, there is a ray of hope in the form of new Internet-based initiatives to trade physical goods. They hope to bring the efficiencies of anonymous trading at an exchange, and thus eliminate the inefficiencies of people having to talk to each other on the telephone in doing trades. They are part of the "B2B" universe, or websites that work on business-to-business commerce.

These sites are innovating in all kinds of directions. The simplest idea is to use a variety of auction formats. There are those going towards true exchange-style order matching. There are murmurs of new "artificial intelligence" efforts in finding matches between trading intentions that are more fuzzily expressed than simple "limit orders" or "market orders".

It turns out that the core task in exchanges, of sewing up the clearing and settlement in order to make the trading truly anonymous and frictionless, is equally important in the B2B space. One of the most wonderful things about modern exchange trading is that it is anonymous. The identities, the politics, the cartel formation are all stripped out: all we face is prices and quantities. But why would you feel safe buying shares from a stranger on the NSE? The only reason to feel safe is that your counterparty is the National Securities Clearing Corporation (NSCC). NSCC does "novation", which means that everytime A buys from B, NSCC sets up the legal structure so that A buys from NSCC and NSCC buys from B. In this case, the default by either A or B does not affect the counterparty.

Novation at the clearing corporation is a prerequisite for anonymous trading. Novation also makes netting possible. Indeed, in general, it is dangerous to have netting without novation.

Most B2B efforts seem to be ignorant of the received wisdom in market microstructure about the design of markets. For example, B2B sites have often ignored the importance of anonymity, and B2B sites have done little about the problems of clearing and settlement. A market which relies on bilateral clearing and settlement will simply not become liquid (it will be like the OTC market for GOI paper).

In the case of physical goods in India, a proper streamlining of clearing and settlement is crippled owing to complex and confusing rules of sales tax, excise and octroi. The tax rates themselves, the types of forms, the form-dependent tax rates, tax incidence at central / state / city levels, multiple authorities, conflicting interpretations, all add up to an incomprehensibly complex system. There is not one sound software system today which encodes the knowledge of Indian sales tax, excise and octroi.

These issues are simply a death-knell for B2B trades in India. The whole point of B2B is to hook up buyers and sellers all over the country. But the tax rules that are prevalent today imply that it is actually foolish for a person in state X to trade with a person in state Y.

A lot has been written about the head start which B2B efforts in India have, by world standards. My sense is that unless we get radical surgery on the complexities of taxation, B2B in India is a non-starter.

Globally, the liquidity in B2B sites has been disappointing. I believe there is one electricity trading site which is obtaining fair volumes; apart from this, there is nothing happening on B2B sites which is like financial markets. I think part of the problem is that the transplantation of knowledge from securities to B2B has not yet taken place.

In summary, the promise of the Internet in improving the liquidity and market efficiency of physical goods is most alluring. As of yet, however, there is no sign that B2B trading offers anything like the liquidity that is known in financial markets. Part of this is because B2B sites have ignored the lessons of sound market design that are now well understood with financial markets.


Back up to Ajay Shah's media page