What Is the Banking "MRI" and Why Do We Need It

Everyone has been hearing about the "MRI" lately, though not the medical one, but the banking one. Let's look at what it is and what it is for. The banking "MRI" is minimum reserve requirements (MRR), which the National Bank uses to carry out monetary policy in the country.

Minimum reserve requirements oblige second-tier banks to hold a certain reserve in accounts at the National Bank to support liquidity. Worldwide, MRR are used both to reduce risks in the banking sector and to manage the money supply.

First, let's see how MRR work using an example. Then we will analyze why the NBK is raising reserve requirements and what this will affect.

MRR are calculated as a percentage of a bank's liabilities, which it must hold in correspondent accounts at the NBK. The reserve percentage depends on the maturity of the liabilities and on the currency. As a rule, the shorter the maturity, the higher the requirement, and foreign-currency liabilities carry higher requirements as well.

Let's calculate the aggregate level of reserves across our banking system and see how it will affect the market.

Total liabilities in the banking system come to KZT 55 trillion. Of this, KZT 45 trillion can be classed as long-term liabilities (although all our deposits are in fact short-term) and KZT 10 trillion as short-term liabilities (mainly current account balances). Tenge-denominated liabilities make up roughly 75% of long-term liabilities and 58% of short-term ones. That gives us:

Long-term liabilities in tenge: KZT 33.75 trillion

Short-term liabilities in tenge: KZT 5.8 trillion

Long-term foreign-currency liabilities (tenge equivalent): KZT 11.25 trillion

Short-term foreign-currency liabilities (tenge equivalent): KZT 4.2 trillion

Before 1 September 2025, the MRR for these categories were 0%, 2%, 1% and 3% respectively. Multiplying through, we get:

33.75 × 0% = 0

5.8 × 2% = KZT 0.116 trillion

11.25 × 1% = KZT 0.1125 trillion

4.2 × 3% = KZT 0.126 trillion

Total: KZT 0.3545 trillion. Half of this amount could be held in the form of vault cash. Since banks' vault cash exceeds the MRR, the actual reserves of second-tier banks came to about KZT 177 billion, or 0.3% of total liabilities.

From 1 September 2025, the NBK raised the MRR to 3.5% and 5% for tenge liabilities and to 10% and 15% for foreign-currency liabilities.

Calculating under the new MRR methodology, we get the following:

33.75 × 3.5% = KZT 1.18 trillion

5.8 × 5% = KZT 0.29 trillion

11.25 × 10% = KZT 1.125 trillion

4.2 × 15% = KZT 0.63 trillion

Total: KZT 3.225 trillion.

That is an almost tenfold increase, but from a low base, so it is nothing to fear. The banking system holds roughly KZT 1 trillion in vault cash, so the effective reserve requirement will be around KZT 2.225 trillion (unless the NBK has removed the offset for vault cash), or about 4%.

Now let's consider how all this will affect the market as a whole.

MRR affect inflation by curbing the capacity to lend and to place money in financial instruments, which in turn also reach the market, mainly through government securities. More expensive liabilities raise the cost of borrowing and, accordingly, reduce demand for loans.

In our case, it is important to understand that liquidity in second-tier banks is currently very high, and according to the NBK, at least KZT 8 trillion is available for lending in the system. So the MRR will have no effect on lending through liquidity. The effect will come through the cost of lending. Consumer lending rates are not very sensitive to the expected price increase, so the new requirement will most likely affect the cost of funding for mortgages and business loans, where sensitivity is fairly high.

The reduction in liquidity will have a greater effect on banks' ability to buy government securities. That will make borrowing more expensive for the state and increase debt servicing costs. Whether this will automatically lead to lower government spending is debatable, since changing budget expenditures is a rather complex process.

That is why I think the main goal was not to fight inflation but to rein in the high profitability of the banking sector. Because money is withdrawn through the MRR, banks cannot earn income on this liquidity, including through risk-free securities. Unlike in the US or the EU, for example, no interest is paid on MRR balances in Kazakhstan.

Let's calculate, however, how much income second-tier banks will lose from this operation.

The amount of tenge liquidity withdrawn will be KZT 1.47 trillion (new MRR) minus KZT 0.116 trillion (old MRR) = KZT 1.354 trillion. Half of this will be covered by vault cash and will not affect banks' financial results. That leaves KZT 677 billion of additional tenge liquidity withdrawn, which at a 17% risk-free rate reduces second-tier banks' profit by KZT 115 billion.

The amount of foreign-currency liquidity withdrawn will be KZT 1.755 trillion (new MRR) minus KZT 0.2385 trillion (old MRR) = KZT 1.5165 trillion. Given that half will be covered by vault cash, the total additional withdrawal will be KZT 758.25 billion. Banks can most likely put foreign-currency liquidity to risk-free use at an average of 3.5%, which gives KZT 26.53 billion.

Result: on an annual basis, the MRR changes will reduce the banking sector's profit by KZT 0.141 trillion.

Free balances on current tenge accounts and the restrictions on foreign-currency deposits bring second-tier banks an additional KZT 1.2 trillion or so before taxes. Against that, the MRR will reduce the banks' excess returns by 12%.

But the mechanism of taking away part of second-tier banks' profit through the MRR has certain imbalances:

• Small and medium-sized second-tier banks do not have as much capital as large banks, and liabilities make up a larger share of their balance sheets, so the MRR burden on them will be much heavier. This will further encourage concentration in the sector.

• Second-tier banks with extensive branch networks and large income from cash operations are less affected by the new MRR, which also puts pressure on smaller and mid-sized banks.

• On the whole, the MRR put more pressure on foreign-currency balances in the banking sector. However, given the restrictions on foreign-currency lending, the effect on second-tier banks will be fairly neutral, except for banks that borrow heavily abroad. For them, external borrowing will become more expensive and harder to redeploy, which effectively creates a barrier to attracting investment into the country through the banking system. I have written before that we should, on the contrary, use the banking channel to attract investment from abroad, and doing so has now become much harder.