Photo: Mint
Photo: Mint

The banking pickle

Contract duration importantly shapes the behaviour in office of a central bank governor

When Raghuram Rajan announced on 18 June that he did not intend to continue as Reserve Bank of India (RBI) governor beyond the end of his three-year term, he surprised a lot of people. His urbane response to a question at the monetary policy press conference on 7 June, that it would be cruel to deny the press the fun they had been having over the issue, suggested he had his extension in the bag.

He did not have a five-year contract, unlike the first three post-reform governors. It was his immediate predecessor, Duvvuri Subbarao, who was the first (post-reform) to be given a three-year contract, although in his case it was seamlessly extended by two years.

Manmohan Singh was the Prime Minister when the contract duration of the RBI governor got reduced in 2008 to three years, and again when Rajan was appointed with a three-year contract in 2013. In his own term as RBI governor starting September 1982, Manmohan Singh served for only two years and four months upon truncation of a longer contract, whose duration I do not know—probably three years. Perhaps his experience shaped his beliefs about contract duration. It is also entirely possible that the appointees in both cases wanted only three years.

Is there an optimal duration of contract for central bank governors? The economics literature does not address the issue. Janet Yellen in the US has a four-year contract, as did her predecessors—Ben Bernanke—who got two four-year terms, and Alan Greenspan, who got too many. Their four-year duration is aligned with that for President of the country. For governors of the Bank of England, the standard reported is eight years, renewable once. Most recent appointees have been in office for 10 years. The present governor, Mark Carney, is reported to have chosen a shorter contract of five years.

Stability in contract duration, as a signal of trust, matters even more than the duration itself. The wobble in India gives the government at this point a choice between staying with the recent practice of three years, or going back to the earlier practice of five. The US has had duration stability, but where they erred was in not having term limits. Greenspan craftily got four full terms by creating around himself a miasma of indispensability before a fifth and final half-term of two years. He had a full stretch of 18 years in which to ruin the financial sector in the US, and bring down the world with it.

Alignment with political terms of office points to five years for a central bank head in India. But quite independently of that, I believe five years is desirable. Contract duration importantly shapes the behaviour in office of a central bank governor. With five years rather than three, they are likely to adopt a more measured pace towards the changes they wish to bring about, and therefore to succeed more in that effort. The government too is likely to choose more carefully if the appointment is for five years.

Although the economics literature does not provide any answers to optimal duration, a famous set of papers by Alex Cukierman in the 1980s established the importance of allowing central bank governors to serve out their full term in office, whatever that length might be. The idea was that countries where governors lived under fear of recall would be less independent, and thereby less effective, and this was borne out by his econometric results. Following from Cukierman, the institutional strength of the central bank of any country continues to be measured by whether governors are allowed to serve out their full term in office, whatever its contracted duration might be.

By that measure India performs well. The last truncated term was that of Manmohan Singh in the early 1980s. But the frenetic press reporting on Rajan’s announcement made it look like a truncation rather than as the natural end of a term. We will have to wait for the memoirs of the principal actors to learn why it was not extended.

Press coverage did not help matters. For that matter, excessive press attention to the person holding office as governor has damaged the institution very severely. The incredible range of photographs of governors, from every angle and in every mood—pensive, anxious, scholarly, jocular, exercising, jogging—in large blow-ups on newspaper pages, is greater than that given to political leaders or movie stars in
other countries.

More seriously, the media have failed to convey to their reading or viewing public the decision-making process within RBI. The governor leads a team with four deputy governors who are his immediate advisers. The governor chairs a board with additional representation of executive directors from within and independent directors from outside. A subset of the full board constitutes the Board for Financial Supervision (BFS). In the interests of full disclosure, I should mention that I was an independent director for four years until September 2015 (but never a member of BFS). And there are the serving officers of RBI, excellent professionals in their own right. This collectivity had been working towards tightening banking regulation and had succeeded in getting changes in default recognition norms well before Rajan’s appointment in September 2013.

The notification ending regulatory forbearance, and requiring that all loans restructured even once be classified as non-performing assets (NPAs), was introduced on 30 May 2013, with a forward effective date of 1 April 2015. At the same time, it preserved and actually enhanced pre-existing forbearance for delays in the date of commencement of commercial operations (DCCO), on the grounds that government clearances remained uncertain and capricious. (All notifications are available on the RBI website.)

Thus, although the tightened NPA norms came in on Rajan’s watch, they were notified well before. Soon after taking over, governor Rajan introduced compulsory reporting by banks to a Central Repository of Information on Large Credits—a huge initiative. (Access to the database is, however, restricted to banks.) The guidelines accompanying this reporting requirement, issued on 30 January 2014, also introduced some new sub-categories in pre-NPA Special Mention Accounts, a pre-existing early warning system for loans in trouble. Subsequently, however, permissible DCCO delays were further extended quite substantially through a 26 June 2014 notification. There were also some relaxations of conditions for refinancing of project loans which had crossed DCCO, notified on 15 July and 7 August 2014.

Despite these pre- and post-DCCO relaxations, banks struggled to cope after the new NPA regime played in as scheduled on 1 April 2015. Avenues in place for settling bad debts, like Debt Recovery Tribunals and the Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interest Act got hopelessly mired in the legal system.

Asset reconstruction companies offered a seeming escape route dangerously bordering on a scam, which was finally recognized for the mirage it was and closed off. Conversion of debt into equity offered an escape route, subject to a cap of 10% of the restructured loan as part of the May 2013 tightening. Strategic debt restructuring brought in to enable banks to gain control of defaulting companies did not really help because it stretched banks into areas where they had little in-house competence.

Finally, the asset quality review initiated in December 2015 put pressure on large borrowers to sell off profitable assets to pay back banks. And that was also what forced passage through Parliament of the bankruptcy bill. Unfortunately, enough has not happened to change the upward trajectory of NPAs, as is clear from the latest financial stability report issued by RBI on 28 June.

The basic problem is that the Indian regulatory regime has historically been lenient in respect of single borrower limits. Internationally, an exposure limit for individual banks of 25% to a single borrower is the norm. In India, while the single borrower limit is the same, there is a group borrower limit of 55%. A single borrower just has to spin off another enterprise within his group to reach beyond his personal limit. Consortium lending by banks has meant that even while individual banks observed exposure limits, the banking system as a whole became heavily skewed towards industry groups. Reports by rating agencies and country reports of the International Monetary Fund under Article IV have repeatedly flagged
this issue.

Corporate India argues that the high group exposure limit is necessary in view of the poor state of development of the corporate bond market. Clearly, structural reform of the banking sector can happen only if it is embedded in reform of the larger financial sector.

Governor Rajan has led his regulatory team from the front. It would have been so much better for both the institution and the country if he had stayed on to follow through, so that bank funds are released for financially starved small enterprises. But RBI is a mature institution which has displayed policy continuity across leadership changes, so hopefully this effort will continue.

Indira Rajaraman is an economist.