Holding the line: Monetary policy under asymmetric uncertainty

Bangladesh is at a moment when stability can no longer be taken for granted. People feel it in the price of essentials, the value of the taka they earn and save, and confidence that their deposits will be there when needed. Macroeconomic and financial stability are lived realities that shape how households earn, spend and save, and how firms invest.

These two dimensions of stability are closely related, but distinct. That distinction matters for those responsible for delivering both.

The issue is especially pertinent now. Bangladesh Bank has just issued its first quarterly Monetary Policy Statement and kept the policy rate at 9.5% against a difficult backdrop of weak growth and renewed inflation risks. The decision brings into focus how monetary policy should respond.

That response cannot be considered in isolation. Fiscal choices, administered prices, external shocks and the health of the banking system shape both the pressures monetary policy confronts and how effectively it can address them. What works therefore depends on the state of the economy.

New risks to disinflation

BB's latest quarterly MPS sees the recent decline in headline inflation as fragile. Administered energy-price increases and the prospective national pay scale could interrupt disinflation even as economic activity remains subdued.

The combination matters. Higher public pay adds to demand, while energy-price adjustments raise production and transport costs. Subsidy savings may offset part of the additional wage bill and contain the deficit, but the package can still be inflationary: fiscal neutrality does not imply inflation neutrality.

An examination of Bangladesh's monthly macroeconomic data, generally covering 2010–2026, helps gauge these risks and how they have played out under different economic conditions. Past energy-price adjustments show a striking asymmetry: electricity tariff increases are associated mainly with food inflation, with the strongest response after one month; diesel-price increases mainly with an immediate rise in non-food inflation.

Applied to the recent adjustments, these historical relationships imply that the 14% electricity tariff increase could add about 2 percentage points to food-inflation acceleration over the following two months. The roughly 17% diesel-price increase could add about 1 percentage point to non-food-inflation, mostly around the time of the adjustment.

Full implementation of the new pay scale would add about Tk1.06 lakh crore to annual recurrent spending, equivalent to roughly 17% of operating expenditure and 1.6% of GDP. Comparable pay adjustments in 2009 and 2015 were followed by a temporary acceleration in inflation, predominantly in urban non-food services. The additional inflationary impulse faded, but this does not mean the earlier increase in prices was reversed. The fiscal shock is larger this time, though the transmission channel is similar.

These estimates indicate the potential scale and timing of inflation pressures based on past experience, not forecasts of actual outcomes. The broader point is that containing the fiscal impact of the pay increase does not necessarily contain its inflationary consequences: demand pressure from higher pay is coinciding with cost pressure from subsidy-reducing energy-price adjustments. For monetary policy, the question is then whether these first-round pressures fade or begin to propagate.

What can the policy rate do?

The policy rate cannot undo the first-round impact of higher energy prices. The policy choice turns instead on two questions: can a restrictive stance reduce the risk that these pressures propagate, and what does it cost in terms of forgone growth?

On the growth side, policy-rate changes pass fairly quickly to bank lending rates, but transmission weakens thereafter. Lending rates have little reliable relationship with private-credit growth, while credit growth itself has little systematic relationship with industrial or aggregate growth, in either direction. When investment is constrained by poor credit allocation and uncertainty, cheaper money does not ensure that additional finance translates into additional productive activity.

The evidence does not identify a precise policy rate that balances these risks. A lower rate could support activity, but the transmission from lending rates to credit and growth is weak. If easing nevertheless stimulates rapid credit and spending, it could allow current price pressures to spread and increase demand for foreign exchange. Historically, unusually rapid credit growth is more clearly associated with inflation and depreciation pressure, while a weaker taka subsequently adds to food inflation.

A higher rate presents the opposite uncertainty. It may help contain second-round effects and support the currency, but the recent tightening cycle provides limited evidence that higher rates alone accelerate disinflation, particularly when inflation originates in food and administered energy prices. Higher borrowing costs could also weigh on already weak investment. The question is therefore not whether 9.5 or 10% is inherently the right rate, but which error is more costly under current conditions.

Asymmetric uncertainty

The asymmetry lies in what happens when policy gets it wrong. Ease too soon, and the growth payoff may be modest while faster credit and demand can amplify inflation and exchange-rate pressure. Keep rates too high for too long, and some growth may be sacrificed, but the weak transmission from rates to activity suggests that cost is less certain. The two errors therefore carry different risks: the potential cost of premature easing is more immediate and visible than the potential cost of waiting. That tilts the balance toward caution.

Bangladesh is hardly alone in facing this dilemma. Other emerging-market central banks confronting renewed energy shocks, persistent inflation and exchange-rate risks have also been reluctant to cut rates simply because growth is weak. Pakistan and Sri Lanka have recently held rates. Indonesia has kept its policy rate focused on inflation and currency stability while using other instruments to support credit and growth. 

The asymmetry is particularly acute in Bangladesh because transmission beyond lending rates is weak and financial intermediation is impaired. The shift to quarterly Monetary Policy Statements is useful: it can keep caution from becoming inertia by asking more often whether the balance of risks still supports the existing stance.

Matching instruments to problems

BB's latest quarterly MPS gets much of the diagnosis right: support activity without derailing disinflation, address supply-side constraints, strengthen monetary transmission through bank solvency and credit discipline, and preserve external stability through exchange-rate flexibility. The evidence examined here broadly supports that diagnosis.

The emphasis on stimulus and targeted refinancing deserves more qualification. The weak relationship between credit and growth may partly reflect misallocation: the exceptionally high NPL ratio suggests that past lending has not always financed productive uses. But even well-allocated credit need not translate immediately into additional investment or output. It may finance working capital, inventories or the higher nominal cost of maintaining existing production. The test of refinancing is therefore not how much credit is disbursed, but whether it generates additional productive activity.

This points to a policy mix in which instruments do different jobs. The policy rate can anchor disinflation; exchange-rate flexibility can absorb external pressure and protect reserves; and targeted refinancing can address specific financing failures. Financial regulation and resolution have a deeper role: sound intermediation supports growth by allocating capital and helping households and firms manage risk. When the financial system itself becomes a source of vulnerability, it can instead impede growth. 

The immediate case is for caution rather than another rate move. That does not establish 9.5% as the right policy rate, nor retrospectively validate the recent cut from 10%. It argues instead for allowing the evidence on inflation propagation, credit and exchange-rate pressure to accumulate before reassessing the stance. The challenge is to make the new quarterly framework a credible calendar for that reassessment. Policy must respond when circumstances change, but departures from the announced review cycle should be exceptional and clearly explained. Otherwise, more frequent MPS publication will do little to make the policy reaction function more predictable.

The author is a former chief economist of the World Bank, Dhaka office.