Central-bank balance sheets without the money-printing shortcut

Start with the transaction

When a central bank purchases a domestic government bond, it changes the composition of financial assets held outside the central bank. The balance sheet acquires the bond and creates a corresponding liability, commonly reserves within the banking system. This accounting description is more precise than imagining that the purchase immediately becomes household spending or a mechanically fixed quantity of foreign-currency demand.

Quantitative tightening reverses part of that process through maturing assets or sales, subject to operational arrangements. The speed, predictability and assets involved matter. A runoff that was announced months earlier may contain little new information at the moment a bond matures. Separate a balance-sheet level from a surprise change in the expected policy.

The seller matters for the accounting. If a central bank buys a bond from a non-bank investor, the investor's bank can receive reserves while the investor receives a bank deposit. If the seller is a bank, the transaction more directly replaces that bank's bond asset with reserves. Neither case means reserves are handed to households as spendable banknotes. Reserves are claims held by eligible institutions at the central bank; customer deposits sit on commercial-bank balance sheets.

Several channels can pull differently

Asset purchases can influence longer-term yields by removing duration risk, signal an intended policy stance and affect financial-market functioning. Any currency response depends on those channels relative to developments abroad. A lower domestic yield may reduce relative attraction, while improved market functioning may attract capital. There is no accounting identity forcing a particular exchange-rate change.

For analysis, label the proposed mechanism. Is the claim about short-rate expectations, a term premium, bank liquidity or risk appetite? These channels call for different evidence. A chart of the balance-sheet total next to a currency chart cannot distinguish them, especially when both respond to the same recession, inflation shock or financial disruption.

Removing duration is not the same as removing every kind of risk. When the central bank buys a long fixed-rate bond and issues reserve liabilities remunerated at a short-term rate, the consolidated public sector has changed the maturity structure of its liabilities to the private sector. Investors hold less of that bond's duration, while the central bank's income becomes sensitive to short rates and the asset yield. This balance-sheet perspective is more informative than counting the purchase as an undifferentiated injection of money.

Liquidity is not the same as a policy stance

The quantity of reserves and the interest rate paid on reserves belong to the operational framework. A large reserve supply can coexist with a restrictive administered rate. Conversely, a smaller balance sheet does not automatically imply that banks are unable to fund themselves. Distribution, collateral, payment timing and access to facilities can matter as much as the aggregate amount.

Treat claims of a universal reserve threshold cautiously. The amount considered ample may vary with regulation, banks' preferences and payment-system behavior. For a currency study, unexpected funding stress may be more informative than a predetermined weekly decline in total assets. Look for a change in the cost or reliability of obtaining the relevant currency, not just a smaller published number.

Reserve supply can move for reasons unrelated to a new monetary-policy decision. Government payments into and out of its central-bank account, demand for banknotes and changes in other central-bank liabilities can shift reserves even with the asset portfolio unchanged. Conversely, bond runoff need not translate into an identical same-day decline in reserves. To understand a funding squeeze, inspect both sides of the central-bank balance sheet and the settlement calendar rather than treating total assets as a complete liquidity measure.

Avoid incompatible comparisons

Two central banks can report balance sheets in different currencies, accounting categories and valuation conventions. Exchange-rate movements can alter the reported domestic value of foreign assets without any purchase or sale. Comparing raw totals also ignores the size of each economy and financial system. Even normalization by GDP does not establish that the institutions' tools have identical effects.

Create a reconciliation table for each observation: purchases, maturities, lending operations, valuation changes and other adjustments. If only a total is available, state that limitation. An apparent expansion during a crisis might be emergency lending rather than renewed purchases of long-duration bonds. Those actions can have different intended purposes and different market implications.

A stock-flow reconciliation also needs a duration distinction. Two purchase programs of equal cash size can remove different amounts of interest-rate exposure if one buys short bills and the other long bonds. Likewise, a portfolio can shrink in market value because yields rise even when the number of securities held is unchanged, if that reporting framework uses market valuation. State whether the published figure is nominal, book or market value before interpreting a decline as a sale.

Build scenarios, not a mechanical signal

A useful practice scenario holds the expected short-rate path constant while changing the pace of asset runoff. A second scenario changes both. A third introduces temporary liquidity lending without a change in the inflation objective. Comparing them helps distinguish the variables that a simple balance-sheet narrative merges together.

Record what would weaken your interpretation. If the proposed channel is rising domestic term premia, stable relative long yields challenge that account. If the channel is funding stress, calm funding markets challenge it. Evidence can remain inconclusive, which is a legitimate research outcome. The purpose is a transparent explanation of risks, not a promise that balance-sheet charts predict currency returns.

For an accounting exercise, split the same total balance-sheet change into three different combinations of runoff, lending and valuation. Explain which combination best matches each proposed market mechanism. If the currency conclusion is identical in all three cases, the argument may rely on the headline total rather than the economically relevant details.

The FX comparison should involve the other central bank too. An unexpected domestic runoff acceleration could tighten relative financial conditions, but the relative effect may be small if the foreign authority simultaneously makes a larger change. There is also no universal conversion from a billion units of QT into a number of policy-rate basis points. The assets, market conditions and signaling content differ. Treat such equivalences as model estimates with assumptions, not as accounting facts.