Drugs and NPLs: a view from Moscow

Published: July 7, 2009

Evgeny Gavrilenkov, chief economist at Troika Dialog, explains why the non-perfoming loan 'crisis' in Russia may not be such a crisis after all, and uses a new poll on drug addiction in Russia to support his intriguing argument.

The Central Bank recently reported that banks' assets shrank 2.3% in May, while their capital decreased 0.5% over the same period. The shrinking of assets continued for almost a fourth month, whereas capital has never contracted in recent years. Even though banks' capital was up overall in 5m09 (by 9.5%), many experts believe that it will keep declining in the coming months and quarters, thereby heightening the risk of non-performing loans (NPLs) exploding at some point. However, we at Troika offer a slightly different interpretation as well as an outlook.

The Levada-Center [a well-known Russian polling institute] has released rather interesting results of a poll asking Russians whether their relatives or those with whom they are closely acquainted are (or were) addicted to drugs (call it variable X). The results suggested that around one fifth of Russians believe that their close friends and relatives use drugs - a rather scary figure. At the same time, only 5% of Russians responded positively when asked if they themselves had tried drugs at least once in their life (variable Y).

The gap between the percentages of those who responded affirmatively to these two questions is too large to be ignored. Of course, one can imagine a society consisting of, say, twenty individuals who have full information about each others' habits and lifestyles (this assumption is important) one of whom, with four relatives or close friends, is a drug addict. In this case, the poll would perfectly reflect the situation described in our 20-person model. In reality, however, with nearly 142mn people living in Russia (who know little about each other), variables X and Y should not differ that much. In fact, they should be nearly equal, assuming, of course, that the samples are representative (which we believe is the case with Levada-Center's methodology).

Why such a difference in the poll results? There are two possible explanations - the respondents are either lying about themselves, or they simply enjoy gossiping about others and are exaggerating their problems. A combination of the two reasons is also possible, but believing that nearly 20% of society is comprised of drug addicts implies that in a each group of people consisting of two parents, one child and, say, two grandparents, one is a drug user.

Or, to put in another way, every second family (consisting of two parents and one child) contains one drug addict (i.e. 25-30mn Russians, including children and elderly people), which is hard to believe. That said, the poll almost certainly indicates that people are too suspicious of others and tend to exaggerate their problems. Does this not actually explain why people are so fond of detective stories and soap operas, which offer easy reading and viewing while discouraging thinking?

The same is probably largely true of the problems in Russia's banking sector. We do not mean to imply that they are non-existent, but their complexity and the rapid evolution of the economic system these days are such that almost every month brings change, whereas people psychologically would prefer to have a very straightforward and conservative view.

Experts have been speculating about a potential surge in non-performing loans and a second wave of crisis enveloping Russia's financial system (and the economy in general) since last autumn. Some experts now suggest that NPLs may rise to 20-30% by year end. Even though Central Bank data indicates overdue loans at around 2.8% of total loans as of June 1, 2009 (up from 2.5% as of May 1), the financial system remains afloat. We think that despite its obvious problems, the system has enough potential to weather the storm without major government intervention.

Actually, amid decelerating inflation and falling overnight rates (which are now below 7%), a relatively stable ruble (which appreciated from R41.0/US$EU1 to Ru36.3/US$-EU1 and then depreciated slightly to around R37/US$-EU1) and an expanding money supply (which increased by more than 4% in May - and probably even faster in June), the NPL problem should look different than, say, half a year ago, when overnight rates climbed to nearly 30% and the ruble was falling 1% and even more each day.

We believe that the above-mentioned positive trends on the money markets in recent months may help banks and companies roll over their debt, so that at some point NPLs will stop rising and huge capital injections into the banking sector may not be needed. Meanwhile, various experts have estimated that between US$22bn and US$130bn must be injected into Russia's financial system in order to achieve stabilisation. The latter figure would represent nearly a third of the current money supply, which is far too high a number.

It seems quite natural, however, that banks have been reluctant to increase lending to companies and households, despite the relative improvement on the money market. Central Bank data indicate that in 5m09, lending activity did not expand at all, so that the total stock of bank lending to the rest of the economy fluctuated around R20tn. It must be noted that over the same period, banks repaid around Ru1trn in loans to the Central Bank, while outstanding debt is still around Ru0.7trn (excluding repo operations).

This debt is represented mostly by uncollateralised loans given by the Central Bank at end 2008 and in January 2009 to increase system liquidity. At the same time, banks increased their own reserves by around R0.43tn in 5m09 - quite a natural reaction in anticipation of a second wave of the financial crisis. From this standpoint, the lack of credit expansion and even the contraction of capital seemed quite natural.

Meanwhile, in recent notes we have mentioned that interest rates on deposits that the Central Bank offers to commercial banks (8%) look quite attractive, especially for those banks with clients that receive money directly from the budget (i.e. at 0%). Why take any risks and lend money to other banks or the real economy if you can enjoy an 8% margin by lending to the Central Bank without assuming any risk?

Unsurprisingly, as of July 7, banks held R0.76tn on deposit with the Central Bank and R0.36tn on correspondent accounts. The sum of these two numbers (R1.1tn, or US$36bn) accounts for voluntary reserves of the banking sector, which may also be utilised should the situation deteriorate. That said, money is available and banks have accumulated some reserves, which may be used to either cover potential losses or increase lending.

As said, we believe that the deceleration of inflation amid ongoing re-monetisation and falling interest rates accompanied by gradual recovery in economic activity will encourage banks to increase lending in 2H09. We would not be surprised to see lending rates fall to around 12-14% within a few months. What the government must do is refrain for a while from injecting massive amounts of money; rather, it is now more important to suppress inflation and restore a more normal circulation of money amid lower interest rates.

If so, then the supposed threat of a 20% NPL rate emerging by year end may turn out to be the same sort of phantom as in the case of a 20% nationwide rate of drug addiction. Would this not also imply that NPLs will not increase significantly above 5%, as suggested by the abovementioned Levada-Center poll?