NPLs and Provisions in Plain Language: Why Banks Should Not Be Given Money Blindly

Banks through the eyes of the layperson

Banks are institutions that borrow money from some and lend it to others at a higher interest rate, earning their income from the difference.

Banks also earn money from various additional services, first of all operational ones: transfers, payments, cash collection, payment cards, and so on. As the stock market develops, the share of operational income in banks' revenue grows, because in countries with developed stock markets people lend to companies and invest in their capital directly, bypassing banks. But we are still very far from that.

For now, banks earn a significant part of their income from the interest rate spread.

Here I would like to explain two things: what the difference is between provisions and non-performing loans (NPLs).

Let's start with NPLs. NPLs are loans that generate no income, usually loans with payments overdue by more than 90 days, because with such a long delay it is unlikely that the borrower will repay the loan in full with interest. This classification is fairly arbitrary, but that is not the point.

Suppose a bank has a total loan portfolio of KZT 1 trillion and has raised this money at 8% per annum.

Its target is to earn a margin of 4 percentage points on top of that 8%. This means the average lending rate is 12% per annum.

Suppose 30% of the loans are reclassified as NPLs. The bank's average rate on loans then falls to 8.4% (12 × 0.7). To keep a 4-percentage-point margin on the KZT 1 trillion it has raised at 8%, the bank must raise the rate on the remaining performing portfolio to 17.1% (12 / 0.7). This is the first problem with NPLs: loans become more expensive for subsequent borrowers, as the bank seeks to compensate for the non-performing ones.

The second problem with NPLs is a liquidity gap. When borrowers become problem borrowers, they disrupt the bank's forecast of incoming cash, while the bank has its own obligations to its creditors. The bank is then forced to look urgently for money to cover cash gaps. Kazkommertsbank is a classic example.

Impairment reserves. Banks now use IFRS to create impairment reserves on loans, and IFRS takes a fairly strict approach to the concept of reserves. A reserve is a liability created by a charge against profit to cover potential losses. For example, a bank issues an unsecured loan with a 50% risk of non-repayment; 50% of the loan must then immediately be booked as a loss. Or a loan is classified as non-performing and a review of the collateral shows that, after selling costs, it does not cover the loan; provisions are then also created against such a loan and charged to losses. These losses create a serious problem for shareholders in the future, because they affect net profit and, therefore, the bank's capital. When a bank's capital shrinks, various regulatory restrictions imposed by the National Bank of Kazakhstan come into play: limits on the size of the loan portfolio, on the amount per borrower, and so on. For this reason, bank shareholders are not interested in creating provisions, since this prevents the bank from growing or requires shareholders to recapitalize it.

The current problem is that all banks are under-provisioned.

If all banks were now forced to create provisions in line with IFRS, half of them would simply have negative capital.

There are many reasons for this:

1. Poor corporate governance. Issuing loans that later become non-performing is partly the result of weak risk management.
2. The external environment. In a falling market, the quality of borrowers objectively deteriorates, as does the condition of collateral. This leads to problems on banks' balance sheets.
3. The absence of a collateral market, proper valuation, and mechanisms for foreclosing on collateral. In this area the situation is simply dismal: if a bank holds a plant as collateral, it can later recover at most 20–30% of its value, because selling such collateral also requires expertise.

There are many problems, and they are simply not being solved. So when I hear that KZT 2 trillion is going to be given to banks to buy out problem assets, I immediately ask myself the following questions:

1. Will the BTA Bank situation repeat itself? After all, this does not solve the problems of the banking system; it solves the problems of the banks' owners.
2. Does the state have the expertise to deal with these problem assets afterward, so that it does not end up selling them for even less than it paid for them?
3. Might it turn out that the money goes precisely to the banks with the most problems, and that the problems are not solved but papered over and made even worse?

In short, there are many questions, and before any money is given, it is necessary to:

1. Ensure that financial statements are reported correctly
2. Fix risk management in banks
3. Fix corporate governance
4. Fix asset valuation
5. Fix the market for collateral

Only then should we decide on providing money and on what principles. I think money should be given precisely to the banks with the best management and the best corporate governance.

PS: I am not a banker, so if bankers correct me on anything written above, I will be very glad.