Smoke and numbers
Indian Express, 31st July 201
The Libor scandal is said to be a new high in criminality by bankers. Politicians are gearing up to imprison bankers for cheating. A more careful analysis of the story suggests that while the behaviour of banks was unsavoury, there were mitigating circumstances. Looking into the future, a fresh focus needs to be placed on the informational foundations of finance. Chartered accountants, credit-rating agencies and other data gatherers provide information that can play a vital role in the working of the financial system. Regulators need to be concerned about their conflict of interest and engage in a certain degree of oversight.
Libor is the interest rate at which banks borrow and lend from each other in London. Transactions take place through conversations on the phone. We can know the buying and selling price for shares of Infosys by glancing at the NSE screen. However, there is no comparable screen where we can learn the Libor. For the last 26 years, the British Bankers' Association (BBA) has computed Libor by asking dealers what they saw as prevailing market conditions, deleting the high and low values of the reports, and averaging the rest.
During the global crisis, the inter-bank money market broke down. Banks did not trust each other and stopped lending to each other. No trades were taking place. When bankers were asked to quote a rate, it is reported that they would often delay beyond the usual 11 am because they were in a bind. News reports also suggest that people often gave outlandish numbers and would be asked to reconsider. It was not just Barclays that reported unrealistically low rates, but all of the 160 or more people working in various banks. The US says it had pointed out that Libor rates were too low, and blames the UK regulator for not being alert enough and solving the problem. Four years later, the US is waking up to the fact that a wrong Libor means that many people received more or less money than they should have for contracts linked to the Libor.
From the viewpoint of the banks, there were two kinds of considerations. Some banks in distress were facing very high borrowing rates because others in the market were not sure they would survive. They were afraid of putting out signals of distress if those high rates were revealed to the world. Other banks were borrowing from the central bank at zero per cent interest, and had no idea what borrowing in the inter-bank market would actually cost. They liked to quote high rates because they earned more interest on Libor-linked loans. Some enforcement agencies, such as in the US, believe that giving out false quotations is a criminal act, and hope to send some dealers to prison.
With the benefit of hindsight, it appears that the right decision would have been to announce that the money market had broken down and that Libor could not be computed. But this would have exacerbated the perception that the global financial system was in collapse. Most contracts written in terms of the Libor rate did not have a clause covering a situation where Libor was not observed. To ignore the problems with setting the Libor was perhaps a practical decision taken by banks, the BBA, and the US and UK governments. After the issue came to light, the blame is being put solely on bankers, who no doubt were complicit, but maybe not much more than the regulators.
Producers of some other information products acted differently during the crisis. For example, a rate that ran into trouble in autumn 2008 was about inflationary expectations in the US market. Trading in inflation-linked bonds collapsed. The US Fed responded to the lack of liquidity in the US Treasury Inflation-Protected Securities (TIPS) market by halting the daily information release of the implied inflation rate. However, there was no large financial market predicated on this implied inflation rate, so few ripples were created by suspending the production of this information.
Looking forward, we should see that information production is a critical function underlying the financial system. Data production should be seen alongside credit rating, accounting, auditing etc as the vital infrastructure that shapes the markets. Governments and regulators should take interest in how data is generated in these fields.
Data is often collected by a market participant or a self-regulatory body. Such bodies may often serve the interests of their members; it may make sense to shift critical data functions away from them towards neutral bodies that have no stake in financial market outcomes. Exchanges are vital sources of data: this is one more reason why the ownership and governance of exchanges is a major issue that should concern policy-makers. When an exchange is conflicted by profit motive, there are incentives to undermine the integrity of information coming out of it.
Better methodologies need to be put in place to improve the statistical system underlying finance. Regulators need to engage in oversight of the methods and operational aspects of these information products. The mechanical transplantation of methods used in the West may often not be appropriate in India.
Even governments may misrepresent data on different macroeconomic variables to present a favourable picture of the economy. In India it is sometimes suspected that data such as the Index of Industrial Production and GDP are being wrongly reported. Revisions of past data downwards show higher growth in current periods, often strengthening this suspicion. The independence of the data-collection agency is crucial if the data is to be credible. In recent times there has been enormous unhappiness over the high volatility, frequent mistakes and repeated revisions in production data in India. If data is being collected and released by the government, which stands to benefit in some ways from it, and it keeps undergoing these frequent changes, it is difficult to have faith in the numbers being released. An important lesson from the Libor scandal is the importance of independent oversight of data collection, methods and release.