Inside the BOJ Revolt: Minutes Reveal the Battle Over Negative Interest Rates

The room where minus 0.1% nearly failed

On January 28 and 29, 2016, the Bank of Japan’s nine-member Policy Board met for three hours and 17 minutes across two days. The public decision sounded controlled: Japan would add an interest-rate dimension to its enormous quantitative and qualitative easing program. Behind that language was a near-revolt. Five members voted yes. Four voted no.

The newly available full account shows that several members had not been brought into the preparation of the negative-rate plan. Sayuri Shirai regarded the design as rushed and insufficiently developed. Takahide Kiuchi questioned whether such a major measure could be implemented responsibly without much deeper study. Koji Ishida doubted that pushing already low yields lower would produce lending or investment. Takehiro Sato feared a damaging rate-cut competition with Europe and stress on the banking system.

Kuroda had also told parliament only days earlier that the BOJ was not considering negative rates. A surprise can make policy powerful by forcing investors to rewrite their assumptions. But contradiction can weaken trust. The argument inside the room was therefore about two things at once: whether negative rates would work, and whether the central bank was deciding and communicating in a way worthy of its independence.

Negative rates did not mean charging every saver

The simplest misunderstanding must be removed first. The BOJ did not order commercial banks to subtract 0.1% from every household deposit. It charged minus 0.1% on only one portion of the current-account balances that financial institutions held at the central bank. Retail deposit rates remained decisions for banks, which generally avoided passing an explicit negative rate to ordinary customers.

Why do banks hold money at the BOJ? They need reserves to settle payments with one another and to meet regulatory requirements. Under quantitative easing, the BOJ bought assets from the market and paid by creating more reserves. The banking system as a whole could not make those reserves disappear; a payment simply moved them from one bank’s BOJ account to another’s.

The negative rate changed the marginal price. A bank receiving an additional reserve balance beyond protected allowances could face a small annual charge. That encouraged it to buy securities, lend or accept lower money-market returns rather than willingly hold the marginal balance. Because market prices are set at the margin, a small negatively charged tier could pull short-term rates below zero without taxing the entire stock.

This three-tier design was both engineering and confession. The BOJ believed the negative marginal rate could influence markets. It also knew that charging the whole reserve stock would crush bank earnings and weaken the very financial intermediation it wanted to stimulate.

Why cross below zero at all?

Conventional central banking lowers the short-term nominal interest rate during weakness. Cheaper borrowing should encourage investment and durable purchases; higher asset prices can support demand; a lower exchange rate can lift exports and import prices. But what happens when the policy rate reaches zero and inflation is still too low?

Economists once treated zero as a hard lower bound because paper currency pays a nominal return of zero. If a bank charged deeply negative rates, depositors could hold banknotes instead. Storage, insurance, transport and payment inconvenience make the practical lower bound somewhat below zero, however. The European Central Bank, Denmark, Sweden and Switzerland had already crossed it.

The BOJ’s intended chain was: push overnight and short-term market rates below zero; pull down yields across the curve alongside massive bond purchases; reduce borrowing costs; encourage investors to move from safe assets into loans, stocks and foreign securities; weaken deflation expectations; and lower the real interest rate.

The real interest rate is approximately the nominal rate minus expected inflation. If a loan costs 0.5% but prices are expected to fall 1%, its real burden is about 1.5%. If the nominal rate is −0.1% and expected inflation is 1%, the real rate is roughly −1.1%. The BOJ’s target was not negative numbers for their own sake. It wanted spending today to become more attractive than holding money whose purchasing power was expected to rise.

The long road from the bubble to Kuroda’s surprise

Japan’s problem began long before 2016. After the late-1980s asset bubble collapsed, falling land and share prices left banks with bad loans and companies focused on repaying debt. Weak demand and delayed financial repair fed deflation. When people expect prices and wages to fall, postponing purchases seems rational, companies hesitate to invest and the real weight of debt rises.

The BOJ cut rates to near zero in 1999, briefly exited in 2000, then reversed as the economy weakened. In 2001 it pioneered modern quantitative easing, targeting the quantity of current-account balances rather than only the overnight rate. It exited in 2006, cut again after the global financial crisis, and added broader asset purchases in 2010.

Kuroda arrived in 2013 with a more forceful promise. QQE would double the monetary base in about two years, buy huge quantities of long-dated government bonds and change expectations decisively enough to achieve 2% inflation. The BOJ expanded the program in October 2014. Yet by early 2016, oil prices had collapsed, China’s slowdown was shaking world markets, the yen was strengthening and measured inflation remained near zero. The danger, in the majority’s view, was that the public would conclude the 2% promise was not credible.

The four dissenters were not making the same argument

It is tempting to call the four opponents “hawks,” but their reasons differed. Shirai worried that introducing a complex tool immediately after supplementary QQE measures could tell markets that asset purchases had reached their limit. Complexity might create confusion instead of confidence. Her objection combined timing, communication and policy architecture.

Ishida, a former commercial banker, focused on transmission. Companies do not borrow simply because rates are a few basis points lower. They borrow when they see profitable demand. If firms already have cash and banks already want customers, lower government-bond yields may squeeze intermediation without creating capital expenditure.

Sato feared an international dynamic. When several central banks cut below zero, exchange rates could become the hidden battlefield. Each country might try to weaken its currency, only to provoke the others into further cuts. Japan could absorb damage to banks without gaining a lasting exchange-rate advantage.

Kiuchi emphasized proportionality and process. A tool that could disrupt JGB purchases and financial-market functioning should be reserved for a true crisis and studied thoroughly. Temporary surprise in stocks, bonds or the yen was not the policy objective; medium-term economic and price improvement was.

The five supporters—Kuroda, Deputy Governors Kikuo Iwata and Hiroshi Nakaso, Yutaka Harada and Yukitoshi Funo—accepted those risks but judged inaction more dangerous. Oil and China-related volatility threatened business confidence and the conversion of Japan’s deflationary mindset. Waiting until expectations fell could make recovery still harder.

What the first market reaction proved—and did not prove

The announcement initially lifted global shares, weakened the yen and drove sovereign bond yields lower. That showed the surprise was powerful. It did not show the policy would raise wages, investment and inflation. Over the following months the yen strengthened again, bank shares suffered and progressively more of the JGB curve traded at negative yields.

Market reaction is a transmission channel, not the final score. A currency can reverse because foreign growth, risk aversion and other central banks change. A lower mortgage rate can help households while a flatter yield curve squeezes lenders. Rising stocks can support confidence while lower pension and insurance returns hurt savers. Monetary policy always redistributes before its aggregate effect becomes clear.

It is also difficult to identify the counterfactual. Inflation remained below the BOJ’s target for years, so the policy plainly did not deliver the promised result on schedule. But that does not prove it had no effect. Without it, the yen might have been stronger, borrowing costs higher and deflation expectations worse. Serious evaluation must compare a messy reality with an unknowable alternative, not only with the original promise.

Why banks objected so strongly

Banks earn much of their income from the spread between lending returns and funding costs. In Japan, deposit rates were already near zero and difficult to push much lower. When market and loan rates fell further, the asset side declined while the cost of deposits could not follow equally. Net interest margins compressed.

Large banks could seek fees, overseas lending and securities income. Regional banks in shrinking prefectures faced a harder equation: fewer borrowers, aging populations, intense competition and limited capacity to pass negative rates to depositors. Negative policy was meant to encourage lending, yet if it damaged profits and capital too far it could reduce willingness to lend—the reversal rate problem.

There were wider effects. Money-market trading became less attractive. Insurers and pensions searched abroad or moved into longer and riskier assets. Government-bond market liquidity weakened as the BOJ became the dominant buyer. Cheap credit could keep low-productivity companies alive and delay restructuring. On the other hand, borrowers refinanced cheaply, the government’s interest burden stayed low and employment conditions tightened. The balance sheet has two sides; so did the policy.

September 2016: from buying a quantity to controlling a price

Only eight months later, the BOJ conducted a comprehensive assessment and introduced QQE with yield-curve control. It continued the −0.1% short rate but aimed to keep the 10-year JGB yield around zero. This was a tacit recognition that driving the entire curve lower was not always desirable. Banks and insurers need some slope between short and long rates, while borrowers need stable financing.

Quantitative easing specifies how many assets the central bank intends to buy. Yield-curve control specifies the interest rate it wants purchases to produce. The BOJ could buy more or less depending on market pressure. It also promised to expand the monetary base until observed inflation exceeded 2% and stayed above it.

YCC stabilized borrowing costs, but prolonged control weakened price discovery and made exit difficult. When global inflation surged after the pandemic and Russia’s invasion of Ukraine, markets repeatedly tested the cap. The BOJ widened its tolerance before ending YCC and negative rates together in March 2024.

Did negative rates end deflation?

The honest answer is: not by themselves, and not on the promised timetable. Japan did not achieve a stable wage-price cycle during most of the negative-rate period. Inflation rose sustainably only after the pandemic disrupted supply, global commodity prices increased, the yen weakened and labor scarcity helped produce stronger wage settlements.

Yet the policy era cannot be reduced to failure. Employment was strong, corporate finance remained exceptionally easy and the fear of a deflationary collapse receded. The BOJ’s determination may have prevented expectations from falling further. But prolonged easing also made the financial system and public finances accustomed to near-zero funding, expanded the central bank’s bond and ETF holdings and contributed to an exchange-rate environment that made imports painful.

In March 2024, Governor Kazuo Ueda’s board judged that a virtuous cycle between wages and prices had come into sight. By a 7–2 vote it guided the overnight rate to around zero to 0.1%, paid +0.1% on eligible reserves, ended YCC and stopped new ETF and J-REIT purchases. The experiment did not end with an admission that it had been meaningless. It ended with the judgment that its role had been fulfilled.

2026: the mirror-image argument

Japan now confronts the opposite risk. The policy rate reached 1% in June 2026 as Middle East conflict raised energy costs and the weak yen amplified import inflation. In April, three board members wanted an immediate increase to 1%; in June, one member resisted the hike. Dissent has returned, this time over how quickly to tighten.

Nearly half of companies in a July Reuters survey said recent increases had hurt operations, largely through borrowing costs and weaker investment. Yet 55% said the weak yen was negative for earnings. Keeping rates low can help borrowers but worsen imported inflation; raising them can support the currency and price stability but burden firms, mortgages and government finance.

A nominal rate of 1% does not necessarily mean money is tight. If expected inflation is above 1%, the real rate remains negative. The deeper question is where the neutral rate lies—the rate consistent with stable inflation and an economy at potential. After decades near zero, Japan has unusually little recent evidence with which to estimate it.

What the ten-year release teaches about central-bank power

The BOJ publishes summaries soon after meetings but releases fuller records only after roughly a decade. The delay protects candor and prevents old internal debate from destabilizing current policy. Its cost is democratic: the public learns the true intensity of a decision only after the officials have left and the policy has run most of its life.

The 2016 record shows why dissent is valuable. A 5–4 decision is not necessarily weaker than a unanimous one. Recorded opposition forces the majority to confront transmission, financial stability and communication. It leaves future policymakers a map of the risks they inherited. Consensus obtained by withholding preparation, however, is not healthy consensus. Surprise directed at markets is one thing; surprise inside the decision-making body is another.

It also teaches humility about instruments. A central bank can set the price of reserves and buy vast quantities of bonds. It cannot compel a cautious company to borrow, create productive investment opportunities, reform an aging economy or determine global oil prices. Monetary policy can change financial conditions; fiscal, labor, competition, energy and social policy determine much of what happens next.

The historical meaning: policy at the edge of the possible

January 2016 belongs to the history of what governments do when familiar instruments appear exhausted. Japan had lived with stagnation and deflation long enough that zero interest no longer looked extraordinary. Kuroda crossed below zero because he believed hesitation would allow the deflationary mindset to harden again. The dissenters resisted because a central bank can damage the system through which its policy must travel.

Both sides saw real dangers. The majority was right that deflation expectations could become self-reinforcing and that zero was not an absolute technical floor. The minority was right that bank profitability, market function, communication and diminishing returns could not be treated as footnotes. The later move to yield-curve control and the carefully tiered reserve system implicitly acknowledged several of those objections.

For 2026, the central lesson is not “never use negative rates” or “Kuroda saved Japan.” It is that unconventional policy is an emergency bridge, not a complete growth model. Its success depends on the health of banks, the credibility of communication, complementary government action and a clear path out.

Ten years ago, the board argued over whether minus 0.1% could make households and companies believe prices would rise. Today it argues over whether 1% is enough to contain inflation without breaking investment. The sign has changed; the responsibility has not. A central bank must make decisions under uncertainty, explain who bears the cost, preserve the machinery of finance and remain willing to change course when the evidence changes.