Recent policy actions by the central bank have reduced the buffer Sri Lanka had against the next default, with the latest being the last rate cut, which came as deposit rates were beginning to rise. It can be argued that since budget deficits are decreasing, interest rates can be lowered. On the other hand, private credit is also gaining momentum, and people are withdrawing cash from banks to purchase cars and make investments.
Lower rates in any case will reduce the flow of cash from loans to banks to some extent. In addition, when the gap between lending rates and Treasury bill rates narrows, banks tend to reduce their holdings of Treasuries and shift to private loans or refrain from buying a large number of new gilts. All of this cannot be calculated in a socialist planning style, but it will be reflected in market rates.
Single Policy Rate, Excess Liquidity
The central bank’s most egregious policy error is the single policy rate, or to put it more correctly, the mid-corridor rate. The pursuit of the mid corridor rate from around 2015, with excess liquidity, ignoring domestic credit, was the proximate reason for Sri Lanka’s sovereign default.
Chasing the single policy rate involves injecting large volumes of money to push down rates by 25 to 50 basis points. The money from open market operations then leaks into import credit and results in reserve losses, as interventions then have to be made to stop the rupee depreciating and triggering social unrest.
The rupee will not only depreciate against foreign currencies but also against commodities like vegetables. If they are not imported, the pressure will drive their price up, and the excess liquidity will be absorbed into notes and coins in circulation to pay for them. David Hume explained this as far back as 1752.
Rate Signal Without Printing Money
Even if large volumes of money are not printed and no reserves are lost to intervene, it is possible to run BOP deficits due to lower collections of dollars by the central bank. The best example is 2019. In 2018 the central bank cut rates with printed money to target potential output, when the unfortunate Finance Minister Mangala Samaraweera raised taxes and market-priced fuel.
As a result, the currency came under pressure. Then interventions were made, losing reserves for imports and more money was printed to sterilise the interventions and target the policy rate. The excuse was that fiscal policy was tight and monetary policy should be loose. In addition, all this was done because inflation was low and there was ‘space’ to cut rates. And another statistical doctrine was that output was below ‘potential’ as well.
All of these are spurious doctrines The applicable principles are as follows: (a) Inflationary open market operations are impossible if there is a need to collect reserves. The corollary is that the central bank must run a deflationary policy to collect reserves. (b) This other principle is that the interest rates must be sufficiently high to curb domestic investments and persuade banks to buy central bank-held securities.
The second principle was seen in 2019. In that year, central bank credit to the government fell (there was deflationary policy), but a balance of payments deficit started to emerge from around July as rates were suppressed through several means (See Graph 1). The path to the currency crisis that ended in default did not start in 2020 with the Gotabaya Rajapaksa administration. It started in mid-2019. There was some type of operation twist at work where longer-term gilt yields were suppressed.
Due to lower than required rates, the central bank was unable to collect sufficient reserves to repay debt, and there was a balance of payments deficit. The central bank, therefore, missed the IMF target for net international reserves. Among other things, some of the swaps the central bank had with domestic banks were also terminated in that year.
As Bad As June 2019?
The situation in June 2025 is probably not as bad as it was in mid-2019. The central bank is no longer publishing monthly overall BOP data to get an idea of what is happening. In May, the central bank bought $260 million from the interbank market, after almost nothing was bought in December, January and February after printing money in the last quarter. Collecting $260 million a month is pretty good. Any number around $200 million, give or take, is good.
This column itself pointed out shortly before the last rate cut, that around that time the overnight rate was signalled up and there was no printing. However, what was more concerning was that the rates were cut as deposit and lending rates were starting to move up. Rising deposit rates indicate that banks are raising more resources to give loans. There are already indications that some finance companies are unable to raise deposits at the rates controlled by the central bank. The printing in the last quarter also took place as rates were starting to move up.
That was a big no-no. Cutting rates when interbank rates were slightly higher than credit demand does not do any harm. All the rate cuts in 2023 and 2024 were on that basis. But both the printing in the last quarter of 2024 and the printing in 2025 happened as deposit rates were starting to move up.
THE BUFFER
Since private credit can only go up, this is a danger signal. The IMF is wrong in saying that collected reserves are a buffer. Reserves are based on past performance. One cannot rest on past laurels. Debt repayments, including interest payments, come in the current and the future. The buffer is the interest rate gap between domestic credit (which is required to keep the exchange rate stable and neither buy nor sell reserves) and the need to build reserves and repay debt.
Collecting central bank reserves is simply an exercise in getting commercial banks to sell deposits (basically rupee reserves) to the central bank instead of giving an import-generating loan to a customer. To avoid forex risk, the central bank has to buy dollars outright. Swaps are not good enough. That is the Lebanon collapse, the Bretton Woods collapse and the Soros swaps of the East Asian crisis. Swaps are a type of self-deception.
The total swaps with domestic counter parties have gone up over the past year. That is partly how gross reserves were boosted. Without NIR data, it is difficult to say whether this rate cut will take the country to default.
Gross reserves will go down when the central bank’s loan to India is paid down, and net reserves will improve. The NFA data is growing. The problem is not that. In the past year, the central bank managed to improve both NFA and gross reserves. The problem is that it no longer appears to be able to do it now.
Slowing Budget Support Loans
One reason why the central bank is not able to grow gross and also grow net foreign assets is undoubtedly the recovering private credit. It must also be noted that debt repayment was easier in the past two years because a lot of budget support loans were available. These will now be reduced.
This is why late-cycle rate cuts, as private credit is recovering, are dangerous and can take Sri Lanka closer to default. The IMF has never been able to fix Argentina. Argentina collects massive amounts of reserves through deflationary policy and squanders everything through rate cuts in about a year.
The fiscal fixes are good, but have not been able to solve external defaults. IMF’s statistical models go against economic principles. Their technical assistance rejects economic principles. That is why programmes are suspended halfway after missing reserve targets, and defaulting countries default again.
If the exchange rate is kept steady, Sri Lanka’s interest rate will move to the levels seen in the US, as it was before 1978. But that takes a bit of time. In the intervening period, rates may move up from time to time. This was clearly seen in countries like China when they moved from depreciating currencies to a fixed peg.
When they broke the peg and ran a flexible exchange rate, interest rates again moved up due to anchor conflict. Now they are operating at a more or less clean floating rate. Meanwhile, resisting short-term interest rate rises, under the single policy rate or for any other reason, will be suicidal.
That was the case in 2012, it was the case in 2015, it was the case in 2018 and from late 2019 to 2022. And it will be the case in the future, unless this country stops rejecting economics for statistics and discretionary policy. A recovery in private credit should be something to be happy about. But under flexible inflation targeting or potential output targeting or a 5% inflation target, it will lay the seeds for a second sovereign default.