The Monetary Policy Statement (MPS) for FY24 announced a few steps to transition from monetary base to interest rate targeting by mid-2026. The conduct of monetary policy so far in FY24 shows glimmers of hope in modernising the monetary policy framework while at the same time reminding of Mark Twain's adage "history doesn't repeat itself, but it often rhymes".
Monetary policy is supposed to start wedding the policy rate and the interbank call money rate through a symmetric Interest Rate Corridor (IRC). The targeted call money rate can vary within 200 basis points of the 6.5% repo rate. Banks and non-bank financial institutions can borrow from the BB at 8.5% from the Standing Lending Facility. They can deposit cash with BB at an interest rate of 4.5% in a Standing Deposit Facility. BB renamed the repo rate, special repo rate and reverse repo rate as repo policy rate, standing lending facility (SLF) rate and standing deposit facility (SDF) rate respectively from July 1, 2023.
Actions and outcomes are more precious than words. We now have data for the first two months of FY24 when the new MPS took effect. We can learn how BB has implemented the IRC from this data. How has BB fared in targeting the call money rates? What is the progress on bringing flexibility to bank lending rates?
Money market outcomes
Consider first the state of the money markets. We have two active ones. The interbank call money market deals with very short term loans from one financial institution to another as does the interbank repo market but with agreements involving the sale and repurchase of securities. Temporary funding needs such as meeting BB's cash reserve requirements drive financial trades in both markets. Figure-1 on recent daily transaction data shows the call money market is deeper.
The volume of lending in these markets are influenced by similar factors such as monetary policy, risk perceptions, domestic and overseas market conditions and idiosyncratic forces. The lending volumes in the two markets do not necessarily move together on a daily basis. The July-August data manifest a negative correlation between the money market rates and volume, especially in the interbank repo market, indicating the dominance of supply side conditions in determining market movements. The conduct of monetary policy is one major mover on the supply side.
Reserve (high powered) money, the total amount of BB's net foreign and domestic assets, contracted in July-August. Foreign exchange sold to banks mopped up Tk 25,149.5 crores cash from the excess reserves of the banking system. BB's domestic assets declined as the government decreased its monetary debt by Tk 23,374.43 crores (July-August 26).
Since the increase of the BB's repo rate from 6 to 6.5% from July 1, the effective call money rate followed the effective interbank repo rate 20 to 25 basis points lower than the target rate after increasing by about 20 basis points in early July. Since then both rates have been sluggish. The effective interbank repo rate appears well anchored around the 6.5% policy rate and above the effective call money rate all the time (Figure-2). The effective rates are the volume weighted average of the rates in each market.
You cannot be faulted for thinking BB may actually be targeting the interbank repo rate rather than the call money rate envisaged in the July-December 2023 MPS (p.35). The rates at which banks have traded funds in either of the two markets have not tested the corridor boundaries.
What these data say about the conduct of monetary policy depends on how we interpret the undershooting of the call rate target vis-à-vis BB interventions. Here is how the IRC is expected to work eventually. BB will use the LAF and SDF to keep the effective call money rate in the target range. In theory, then, if the call money rate is below target, BB is supposed to mop up liquidity and vice versa. Assuming the policy intent is to let the rates float within the boundaries, there is no prima facie case to intervene when the deviations from the target do not cross the symmetric 200 basis points.
The fact that SLF and DSF have not been used is consistent with the float within the corridor assumption. However, BB provided liquidity support in the repo market at 6.5 to 7% on a discretionary basis from the pre-existing liquidity support facilities. Outstanding repo has increased (Figure-3) even though the deviation of the effective call money rate from the target has entirely been on the down side. BB could argue that the effective call rate averaging 22 basis points below the policy rate in July-August is close enough. After all, keeping "close" is all BB committed.
What is beyond debate is money remains cheap. Short-term risk free nominal interest rates range between 6.5% (policy rate) to 7.4% (182 day T-bill rate). Taking account of 8.2% (12-month moving average through July) inflation, these are equivalent to -2.7% to -0.8% real rate. They have been negative for a considerable period of time, not just currently. We do not know the unobservable short-term real interest rate (called the neutral interest rate) that would prevail when the economy is operating at its potential and inflation is stable. But surely it is not negative.
Second decimal place flexible lending rate cap
The money market rates are supposed to weigh on the rates banks charge their customers, influencing business and consumer spending.
BB introduced a new method of setting the lending rates for all types of bank loans, replacing the previous 9% cap on most commercial lending. A 7.1% Six Month Moving Average Rate of Treasury bills' (SMART) as a reference rate was announced on the first working day of July with a commitment to update the rates at the beginning of each month. Bank lending rates are limited to a maximum of 3% above SMART and non-bank financial institutions to a maximum of 5%.
It is patently wrong to think that the interest rate is no longer capped. The interest rate cap for the largest segment of lending gained the potential to rise from 9% to the SMART rate plus 3% cap. BB is using the government debt market to make sure the lending rate is where it wants it to be. BB's absorption of Tk 10,343 crores Bangladesh Government Treasury Bonds during July-August and pausing the auction of long term bonds suggests a preference for cooling the short term rates as surplus liquidity flocks to the short term T-bill market.
The SMART rate does not have a history of flexibility. BB's backward calculation of SMART rate shows it was 7.04% in February, rising to 7.13% in May, and then falling to 7.1% in June-July. The 182-day T-bill rate rose from 7.13% in the first week to 7.42% in the second week of July. It declined to 7.36% in the second half of August. The SMART rate for September has risen by 4 basis points to 7.14%.
The flat behaviour of the 182-day Treasury rate (Figure-2) on which SMART is based shows we are still miles away from a market based, flexible bank lending rate. Monetary policy transmission mechanism through bank lending rates thus remains muted. July-August data show a high correlation (0.92) between the interbank repo and call money rates. But the pairwise correlation coefficients with the 182 day T-bill rate are low—0.38 with interbank repo rates and 0.33 with call money rates.
David Altig, Executive Vice President at the US Federal Reserve Bank of Atlanta, constructed a long rope analogy as a simple way of thinking about how monetary policy generally works. The short-term, relatively riskless interest rates are at one end followed by yields on relatively safe longer-term assets farther along. A whole slew of lending rates at the other end of the rope drive decisions by businesses and consumers. BB's repo policy rate is at the short end of this rope. The motion from the snap it gives the repo rate ripples down to call money rates but cannot reverberate throughout the economy because of the rigid SMART rate.
Forward guidance on transition could help
There is therefore only a peripheral difference as of now between the operation of the monetary policy before and after July 1 2023. This is not necessarily unintended. Moving completely from reserve money to interest rate targeting is projected by the end of the IMF program period in mid-2026. Neither the MPS nor the IMF program document provide any clue on the exit strategy from the quantitative targeting and the lending rate ceiling regime. Surely, an abrupt exit in June 2026 is not on the cards.
Gearing interventions in the money markets to let the rates float within the corridor is supposedly the end game in the monetary policy reform already put in place. This may elude if secondary markets for government bills and bonds do not develop in tandem. SMART is a transitional arrangement. BB is committed to removing all interest rate caps by mid-2026. It can be dismantled in the 36 months (counting from July) by adjusting either the six month calculation period or the spread or both with every MPS going forward. An appropriate dynamic transition formula needs to be worked out and made public.
The modernisation plane has at best ignited the engine. BB can help with taxiing and takeoff by communicating their change management plans. It needs to signal the transition by making interest rates increasingly flexible. Interest rates will then be in a position to play the role of automatic shock absorbers and set the stage for switching monetary policy to inflation targeting.