An SMS joke that did the rounds recently: “What happened when HDFC Bank declared a 25% earnings growth as against the ~30% that it has been delivering for the past 40 quarters? Steve Ballmer lost his job as this mistake was due to a major bug in Microsoft Excel”.
HDFC Bank’s famed consistency inspires both awe and disbelief. When in 2009 I had told a US investor that the bank’s high valuation is primarily attributable to its earnings consistency, his response was “Well Sir, Lehman also faced only one bad quarter, which was its last”.
Consistency in earnings growth may not necessarily be an outcome of smoothing. Unfortunately, the term income smoothing has an unduly negative connotation. Perhaps this feeling was bolstered by the only and fleeting reference to it that the Reserve Bank of India (RBI) made when they effectively banned the use of floating (”unallocated”) provisions in 2005, suggesting that banks were using them to manage earnings. But it need not be that way. There is a lot of research on income smoothing that has linked it to counter-cyclicality. This is precisely what regulators are struggling to introduce after the 2008 financial debacle.
Income smoothing can be done through two partly discretionary items in the P&L account: occasionally treasury profits and mainly loan loss provisions (LLPs). The early 2000s period was a prominent example of state-owned banks income doing smoothing by using large treasury gains to bump up LLPs, a link that was, more generally, found to be statistically insignificant by Kiridaran Kanagaretnam and Gerald J. Lobo (2001). Analysis thereafter has largely focussed on the use of LLPs for income smoothing.
Smoothing is controversial because bank regulators and managements are on one side, and auditors, securities’ regulators and investors on the other. Even in India, auditors have disapproved some banks’ making floating provisions masquerading as specific (i.e. against actual NPLs) provisions. In the late ‘90s, within an overall investigation of earnings management, the US Securities and Exchange Commission had asked Suntrust Bank to REDUCE provisions made in three earlier years. But as Heba Abou El Sood (2011) observed, the 2007 financial crisis seriously challenged the backward-looking norms for LLPs, and diminished the significance attached to “transparency” demanded by accounting standard-setters.
The author also documented evidence that i) effect of income smoothing on LLPs is amplified when banks hit the regulatory minimum capital target, or are more profitable ii) banks have income smoothing incentives to delay the provisioning process during recessionary periods (”procyclicality”) iii) bank internally-set capital targets are more significant triggers of income smoothing than the regulatory-set minimum ratios iv) during pre-crisis periods, banks tend to accelerate LLPs to smooth income downward, and the reverse during crisis periods.
In a study covering only OECD banks, Neila Boulila Taktak et al (2010) observed that a significant proportion of banks resorted to income smoothing by “manipulating” LLPs or treasury profits. Curative and prudential banking regulations appear to have an impact on smoothing practices. Banks poorly or highly capitalized or with insufficient primary equity capital get further involved in discretionary practices.
The controversial subject of managerial incentives to undertake income smoothing was dealt with by Kiridaran Kanagaretnam et al (2001). Their study showed that that bank managers do save earnings through LLP in good times and borrow earnings using LLP in bad times. They also provide evidence that the need to obtain external financing is an important variable in explaining differences in the extent of income smoothing.
It may be interesting to note that Indian banks have been remarkably more counter-cyclical than their regional peers (Frank Packer & Haibin Zhu, BIS, 2012). This has less to do with bank managements and more with evolving regulation - “LLP did not ... become more conservative at all points in time..., but actively leaned in a fashion that ameliorated swings in earnings...”. RBI, particularly since 2004, has been driving banks towards creating a buffer in good times, and drawing on them in bad times (not just through LLPs). Very few banks have gone a step further and institutionalized it, providing credence to the authors’ observation that “the degree to which policy initiatives were responsible for this, as opposed to ... prescient behaviour ... of banks, remains a subject for ... investigation”.
As banks in India have not gone through many cycles, it will be important for RBI to carry on with their counter-cyclical approach, and not object to any resultant income smoothing unless it reeks of undesirable practices. It is in this context that their stance on restructuring has somewhat wavered from that goal. They set up a committee only after the volume of restructuring went out of hand. So the income smoothing that they themselves could have guided did not take place, and additional provisions may hit at the wrong time.
Dipankar Choudhury was director of Indian Financial Services Research at Deutsche Bank, and is currently an independent consultant focusing on banks and financial services.