Why Banks Lend So Little to Business

Lending

More and more questions are being raised about lending, because lending to the real sector in our country is very low. The balance is heavily tilted toward consumer lending: in 2021, loans to legal entities grew by only 4.5%, below inflation, while retail lending grew by more than 30%. Why is this happening, and what can be done? How can banks be made to lend, and at low rates, so that the economy grows?

This is a very complex and contested question.

Much has been written on the subject, but unfortunately the trend has not yet turned toward lending.

I would divide lending into four segments: consumer, mortgage, SME and large corporate. Each has its own specifics and its own problems.

Of the four, only consumer lending is growing, because it is currently the most competitive segment. Unfortunately, consumer lending stimulates imports more than demand for domestic goods.

Unfortunately, competition in Kazakhstan's banking sector is narrowing every year. Over the past 10 years the number of banks has only fallen, and foreign banks are in no hurry to enter the market or even to open branches. Meanwhile, it is mainly domestic banks that are leaving the market.

Kazakhstan now has 22 banks. Of these, 12 are subsidiaries of foreign banks, three more are majority-owned by foreign nationals, and one is state-owned (Otbasy). That leaves just six domestic banks: Halyk, Forte, Zhusan, BCC, RBK and Nurbank. The top five banks account for 65% of all assets, and the top ten for almost 90%. This is a high level of concentration, and it prevents competition from developing.

Because the foreign subsidiary banks have high ratings, a very large volume of government liquidity is held in foreign banks. In the long term this hurts the country's balance of payments, since part of the profit earned on government funds placed with them is paid out abroad as dividends. All this is the result of an overreaction to the non-market placement of funds after a number of Kazakhstani banks went bankrupt.

The main problem holding banks back from lending to legal entities is excessive regulation and the high risks involved in assessing the solvency of businesses. All the risks that have materialized in the banking sector since 2008 have come from large corporate business, and overregulation is the response to that.

That is why banks now take a very conservative approach to assessing businesses.

The main problems in lending to legal entities are as follows:

The AQR. After the AQR was carried out, collateral and credit policies tightened many times over, out of proportion to the risks. In effect, the AQR was conducted as a formality, without looking into individual cases and without adaptation to Kazakhstan's realities. As a result, many good companies were classified as Stage 3 (defaulted), which prevents banks from lending to them in full. Introducing an annual AQR as a system of regular supervision places a heavy burden on banks and keeps them from taking risks. The regulator does not understand the real creditworthiness of businesses, while banks have no room for maneuver, because they must constantly meet AQR requirements that are stricter than IFRS. In addition, there are problems with the tax authorities, which do not accept the larger regulatory provisions and allow only IFRS provisions as deductible. The State Revenue Committee and the ARDFM need a single, consistent policy.

No mechanism for holding the owners of large companies accountable. The legislation is built in such a way that it is practically impossible to recover anything by lawful means from companies beyond the pledged collateral. Collateral enforcement takes so long that by the time it is completed, the quality and value of the collateral no longer cover the losses on the loans issued. Entrepreneurs use many schemes to avoid their obligations. These range from building up accounts payable through companies that are formally unaffiliated but in fact controlled, to stripping out the main assets through sham sales. Priority claims are inflated, rehabilitation schemes are used, and so on. All this prevents banks from lending normally to large businesses, because the risks are out of proportion to the potential gain. We need tougher legislation and a tougher bankruptcy regime, which at present works more for business than for banks.

Stress testing. Tighter requirements through stress testing push banks to build up capital buffers in order to comply. In effect, this strengthens the position of large banks and prevents small and medium-sized banks from developing. For example, risk weighting of enterprises in the SME segment is calculated as a percentage of equity. In practice, the weighting is lower for large banks: if the threshold is 0.02% of equity, a small bank must weight a KZT 100 million exposure at 100%, while a large bank weights it at 50%. This creates unequal conditions.

Oligopolization of GovTech. Large banks' dominance of GovTech prevents the other banks from developing in this area. We need uniform standards for all banks to connect to government databases through an open API. For example, Halyk and Kaspi get priority when connecting to government services, while the other banks wait in line or are not allowed to integrate at all, because the key resources simply move on to another project for these large banks.

Scoring. We are now investing heavily in developing scoring models for SME lending, so that credit decisions can be issued within a pipeline of anywhere from 20 minutes to a maximum of 3 days, depending on the size of the loan. The better the scoring, the higher the approval rate and the lower the risk. Lower risk is reflected in the final interest rate. For this we need integrations with various government databases so that decisions can be made in real time. I wrote to you about this project in a letter that R. Batalov passed on.

Subsidized rates. Subsidizing rates for legal entities raises a great many questions. The share of subsidized loans in total issuance is now larger than that of unsubsidized ones. This prevents the National Bank of Kazakhstan from fully carrying out its monetary policy. The subsidy policy clearly needs to change. I propose the following:

• Link the final rate for the borrower to the NBK base rate. If the base rate rises, the rate on business loans rises, and vice versa. In effect, this would synchronize the movement of NBK rates and business loan rates. At the same time, if subsidies now amount to almost 9% of the rate, we would cut them to 5%, which also reduces the overall burden on the budget.

• Set a single criterion for subsidies: value added for the country. If value added exceeds 50%, the project qualifies for subsidy. If not, it does not. The methodology for calculating value added should be defined for each economic activity classifier (OKED) code, down to the product nomenclature, and then approved. Every new product to be included in the program should be reviewed by the Ministry of National Economy and the National Chamber of Entrepreneurs (Atameken).

Mortgage lending. To develop mortgage lending, I would suggest drawing on Russia's experience with DOM.RF. Otbasy Bank would act as a sandbox and would later work on government programs alongside second-tier banks in an equal competitive environment.

Debt write-offs. A major problem is the write-off of part of a debt in order to settle the obligations of troubled borrowers. For example, if a company cannot service the full amount of its debt, it becomes a problem borrower, and the bank has no right to write off part of the debt and restructure the remainder, because the write-off carries tax consequences. The only option is bankruptcy, but that liquidates the company, which often ceases to operate and drops out of economic circulation. All that is needed is a tax relief for cases where the bank makes a write-off. The country's economy as a whole benefits from this. In such a case the bank has no incentive to avoid taxes.