The Bank of Japan’s green loans meet rising interest rates

A bank can earn more when its central bank raises interest rates, even on a position funded by a loan from that same central bank. The borrowing cost need only stay fixed while the return on reserve balances rises. Two different pricing rules create a gap that a policy-rate decision does not automatically close.

The Bank of Japan has changed the rules that allowed this mismatch in its climate-response financing programme. New loans will be priced at the average deposit-facility rate over their life. Existing loans will retain their previous terms until maturity. The reform therefore takes effect before the whole stock of lending has moved to the new regime. S02

It accompanies the rate increase announced on 18 September 2026: the target for the overnight interbank rate will rise to 1.25% on 24 September. At this article’s 22 September cutoff, that increase has not yet taken effect. The climate-loan amendment, by contrast, became effective on 18 September. The two dates matter because they determine which side of a bank’s balance sheet reprices first. S01 S06

A one-year loan does not reset at every policy meeting

The climate facility provides eligible financial institutions with collateralised BoJ funding. Access depends in part on their portfolio of investments and loans supporting Japan’s climate response. Each central-bank loan normally runs for one year. Another may replace it at maturity, but successive loans are not an indefinite commitment at the original interest rate. S03

Under the previous formula, the loan rate was the complementary deposit-facility rate on the date of disbursement. Once the money had been advanced, a later monetary-policy decision did not change that loan’s price. The operating outline dated 7 May 2025 still describes this earlier rule. The September 2026 amendment now governs new operations. S04

The asset side can move differently. Reserves are balances that banks hold at their central bank. In Japan, balances above required reserves earn interest under the complementary deposit facility. Their remuneration follows the prevailing facility rate rather than the rate on the day the balances were acquired. S05

The borrowing rate and the return on reserves can consequently start at the same level and then diverge. No borrower has to default, and no bond has to appreciate, for that gap to appear. It comes from applying different reset rules to the two sides of a balance sheet.

¥100 billion and two interest calculations

Consider an entirely fictional example. A bank borrows ¥100 billion for one year at 1%. Throughout the year, it holds an equal amount in fully remunerated reserves. The deposit-facility rate stays at 1% for the first half of the year and rises to 1.25% for the second half.

Annual interest on the loan is ¥1 billion. The reserve balances earn ¥500 million in the first half and ¥625 million in the second: ¥1.125 billion in all. The difference is ¥125 million.

The same result can be calculated as ¥100 billion multiplied by a 0.25 percentage-point spread for half a year. Applying the spread for a full year would overstate the result, because it exists only during the second half.

This is a gross interest differential, before collateral costs, balance-sheet constraints, fees and hedging. The example assumes equal half-years and simple interest; actual operations follow their own daily accrual and payment conventions.

Now reverse the rate movement. If reserve remuneration falls to 0.75% for the second half, annual interest income is only ¥875 million. The loan still costs ¥1 billion, leaving a negative ¥125 million differential.

The original contract leaves the bank exposed to the future path of rates. An increase helps in this particular balance-sheet configuration; a decline can hurt. Actual remunerated reserve holdings depend on each institution’s balance sheet.

The new formula removes a pricing mismatch

New climate loans will be priced at the average of the deposit-facility rates applying during their term. Instead of fixing the borrowing cost using a snapshot taken at disbursement, the rule makes it depend on the rate path over the contract’s life. S02

In the rising-rate example, that average is 1.125%. The loan costs ¥1.125 billion, matching reserve income under the stated assumptions. The gross differential is zero. It is also zero in the falling-rate example: both income and borrowing costs decline to ¥875 million.

That cancellation assumes matching principal amounts, exposure periods and interest conventions. It does not mean every institution using the facility will earn exactly zero. A bank may deploy its funding differently, pay interest to its own depositors, hedge its exposures or attach value to the liquidity it obtains. The model isolates the rate gap removed by the revised pricing clause.

Nor is the average-rate formula new to the BoJ. In its 31 July 2024 decision, it chose this approach for its bank-lending support programme while linking climate loans to the rate at disbursement. Programmes run by the same institution could therefore respond differently to later rate increases. S10

The BoJ presents the reform as an adjustment to support the smooth conduct of market operations. Its effect on the interest differential is inferred here from the contractual terms. S02

July’s loans keep their earlier terms

On 17 July 2026, the BoJ announced ¥13,981.2 billion of loans for disbursement on 21 July, with a maturity of 21 July 2027. That term extends well beyond the reform date. S08

The deposit-facility rate then stood at 1%, following the June decision that took effect on 17 June. Combining that rate with the disbursement-date pricing rule gives a reconstructed contractual rate of 1% for the July tranche. S07 S04

The transitional provision preserves earlier terms until maturity. The increase in reserve remuneration to 1.25%, scheduled for 24 September, can therefore raise the return on matching reserve balances without changing the borrowing rate on those loans. Any subsequent differential will depend on later policy decisions and on the positions banks actually retain. S02 S06

Measuring a rate increase’s effect requires the origination date and maturity of each borrowing. Those details distinguish loans that can reprice from tranches that remain fixed.

At maturity, any replacement would be a new loan. Preserving the old contract does not permanently attach its earlier rate to the institution’s climate-finance portfolio. S03

Nearly ¥25 trillion in announced outstanding loans

July’s announcement distinguishes the gross amount of the operation from a reference outstanding balance of ¥24,936.1 billion at 21 July 2026. It was published before that date: this was the balance announced for the disbursement date, not a daily observation at 22 September. The Japanese version confirms the units and the prospective status of the figure. S08 S15

The preceding release reported a reference balance of ¥21,114.6 billion at 30 January 2026. The difference between the two balances is ¥3,821.5 billion, much less than the ¥13,981.2 billion of gross loans announced for July. A lending operation may replace maturing funding; its gross size does not measure the programme’s net expansion. S09 S08

The principal has to be repaid. An interest advantage depends on the rate attached to each tranche, the remunerated balances held against it, their holding period and the alternative funding the institution would otherwise have used.

The reform also introduces a ¥50,000 billion overall ceiling and a ¥10,000 billion limit per counterparty for its own use. Separate rules distinguish cooperative central organisations’ own borrowing from funds borrowed for their members. These are capacity limits, not fresh allocations already paid out or forecasts of take-up. S02

Climate lending and central-bank reserves can coexist

A bank loan creates a deposit in the customer’s account. Reserves serve, among other purposes, to settle payments between banks. When the customer pays a business at another bank, the deposit and the corresponding reserve transfer can move to that other institution. There is no identifiable bundle of green-labelled cash to track from the BoJ to a construction site. S14

The Japanese programme uses a stock of eligible financing to determine access to central-bank funding. A bank can therefore hold climate-related loans, have financed its customers’ expenditure and also hold interest-earning reserves. These positions are not mutually exclusive.

Additionality means extra lending or improved financing terms attributable to the programme. Assessing it requires estimating which loans would have existed without the facility and how much they would have cost. The eligible loan stock alone does not supply that counterfactual.

In its original 2021 strategy, the BoJ explained that it wanted to support private financing while avoiding direct involvement in allocating resources to individual projects as far as possible. Assessing environmental outcomes requires data on investments and emissions. S12

Floating-rate funding can still provide valuable certainty

The programme has two separate gates. An eligible climate portfolio establishes potential access to refinancing; collateral actually pledged secures the BoJ’s claim. Borrowing remains constrained by the eligible portfolio, applicable ceilings and unused collateral value. Counterparties must also meet financial-soundness and disclosure requirements. S03 S04

Pledged collateral has an opportunity cost: it cannot simply be assumed to remain freely available for every other transaction. One-year funding is also different from borrowing that must be renewed daily. These considerations remain relevant even when the loan’s rate follows current policy rates.

In its report covering April 2025 to March 2026, published on 12 August 2026, the BoJ says some counterparties valued one-year fixed-rate climate funding for its cost, stability and contribution to their liquidity ratios. These qualitative assessments come from participating institutions. S11 S16

Moving to a floating rate removes protection against a future increase in the cost of that loan. It preserves the ability to secure a funding maturity.

The ECB changed the price of outstanding loans

Europe offers a useful comparison. On 27 October 2022, the ECB recalibrated TLTRO III, its targeted longer-term refinancing operations. From 23 November, interest on the remaining outstanding loans was to follow the average of the applicable key rates over the remaining period. Additional early-repayment opportunities accompanied the change. S13

That altered the future price of loans already granted. It did not retroactively reclaim all interest calculated before the change. TLTRO III was not a climate programme, and its pricing depended in part on lending performance. The relevant comparison is the treatment of existing contracts, not an assertion that the programmes were identical.

One approach changes the price of outstanding borrowing sooner. The other lets previous terms expire with the loans. Preserving a contract provides visibility to the borrower but allows the mismatch to continue. Repricing its future interest can accelerate monetary transmission while changing the terms on which institutions based earlier decisions. Neither property, considered alone, establishes which policy is preferable.

Contracts determine the speed of transmission

The reform brings the cost of new climate loans closer to the rate paid on reserves. Existing contracts retain their own repricing timetable. Monetary normalisation also depends on when a contract permits its price to change.

Calculating banks’ net gains would require matching loan tranches and contractual rates with remunerated balances, hedges and other costs. It would also require choosing a counterfactual: no central-bank borrowing, market funding or another source of finance. The spread against an old BoJ loan can differ from the saving against the best available alternative.

The rate charged to a business is another link in the chain. A funding advantage can be retained in a bank’s margin, shared with a borrower or absorbed by other expenses. Measuring that pass-through requires comparable lending terms before and after the reform.

Limits

Evidence cutoff: 22 September 2026, before the announced 24 September implementation of the 1.25% rate. September’s new rules are distinguished from earlier operating documents. Disbursement announcements are not a complete daily record of outstanding loans. Their amounts include financing for members of cooperative central organisations, as specified in the BoJ’s notes.

The ¥100 billion calculations are fictional: simple interest, two equal half-years, constant loan principal and remunerated reserves, and no fees, hedging, balance-sheet constraints or compounding. They reproduce neither a particular contract nor an official daily accrual convention. No sector-wide profit, causal lending effect, environmental outcome or market impact is estimated. Institutions’ feedback is reported by the BoJ; no independent interviews were conducted for this article.

Original text, figures and calculations: l0g, CC BY 4.0. Third-party documents remain subject to their owners’ rights.

Further reading: the Bank of England’s QE exit and reserve remuneration in the euro area.

Sources

1. S01 · Change in the Guideline for Money Market OperationsBank of Japan · 2026-09-18 · p. 1, rate decision, climate funding and effective dates.
2. S02 · Amendment to the climate-response financing operationsBank of Japan · 2026-09-18 · Attachment 1, PDF pp. 2-3: paragraphs 6 and 8 and supplementary provision; Attachment 2, p. 4.
3. S03 · Principal terms and conditions: Climate Response Financing OperationsBank of Japan · 2026-09-18 · Paragraphs 3, 5, 6, 8-11 and 13; supplementary provision. Adopted 22 September 2021, revised 18 September 2026.
4. S04 · Outline of Transactions: Climate Response Financing OperationsBank of Japan · 2025-05-07 · II.C, II.E-H; III; IV.A-B. Displayed version date: 7 May 2025.
5. S05 · Principal Terms and Conditions of Complementary Deposit FacilityBank of Japan · 2026-06-16 · Paragraphs 4 and 5: annual rate and excess-reserve interest base.
6. S06 · Amendment to Principal Terms and Conditions of Complementary Deposit FacilityBank of Japan · 2026-09-18 · Attachment, PDF p. 2: paragraph 4 and effective-date provision.
7. S07 · Change in the Guideline for Money Market OperationsBank of Japan · 2026-06-16 · p. 1: policy target, deposit facility and 17 June 2026 effective date.
8. S08 · Loan disbursement under Climate Response Financing OperationsBank of Japan · 2026-07-17 · p. 1: amount to be disbursed, reference outstanding balance, dates and scope notes. Unit: billion yen.
9. S09 · Loan disbursement under Climate Response Financing OperationsBank of Japan · 2026-01-29 · p. 1: 30 January 2026 disbursement, 1 February 2027 maturity and reference outstanding balance.
10. S10 · Amendment to Principal Terms and Conditions of Complementary Deposit Facility and related operationsBank of Japan · 2024-07-31 · Attachments 3 and 5, PDF pp. 4 and 6: pricing of the lending-support and climate operations.
11. S11 · Market Operations in Fiscal 2025Bank of Japan, Financial Markets Department · 2026-08-12 · Box 4, pp. 46-47: counterparties’ feedback on climate funding. Fiscal period: April 2025 to March 2026.
12. S12 · The Bank of Japan’s Strategy on Climate ChangeBank of Japan · 2021-07-16 · “Monetary Policy” section: financing support and neutrality in allocation.
13. S13 · ECB recalibrates targeted lending operations to help restore price stability over the medium termEuropean Central Bank · 2022-10-27 · Paragraphs on pricing from 23 November 2022 and additional early-repayment windows.
14. S14 · How is money created?Bank of England · 2019-10-01 · “How does it work?” and “Can banks create as much money as they like?”.
15. S15 · 気候変動対応を支援するための資金供給オペレーションの実施結果Bank of Japan · 2026-07-17 · p. 1: 139,812 and 249,361, in hundred-million yen units; expected balance at 21 July.
16. S16 · Market Operations in Fiscal 2025: publication pageBank of Japan · 2026-08-12 · Publication date and link to report S11.