Research & Investment Strategy

Bank of Japan wary of deeper negative interest rates

Key points


  • The Bank of Japan (BoJ) will hold its monetary policy meeting on 19 September. Amid rising expectations for rate cuts by the US Federal Reserve and European Central Bank, we do not expect the BoJ to lower its key policy rate further, seeing a less than 20% chance of a rate cut this year
  • We believe the BoJ will strive to delay rate cuts for as long as possible but will maintain an easing bias by changing its forward guidance and possibly changing the parameters around its yield curve steepening (less purchases on long-term Japanese Government Bonds over 5Y and more).
  • We consider the potentially negative impact of further rate cuts on financial institutions and inflation expectations
  • We also provide an historical overview of BoJ policy measures since 2013

Rate cut or not?

The Bank of Japan (BoJ) will hold its monetary policy meeting on 19 September, with rising expectations that it will join the Federal Reserve (Fed) and European Central Banks (ECB) in easing monetary policy further. Markets currently suggest a 70% probability that a rate cut occurs before the end of the year, against 22% six-months ago.

Against such an outlook the BoJ is likely to be keen to maintain an easing bias, if only to avoid further unwanted yen currency appreciation as a side effect of monetary easing in other jurisdictions. However, we believe that the BoJ will strive to delay rate cuts for as long as possible, concerned about the impact that further easing could have on financial institutions specifically and uncertain of the benefits that such a move would have on inflation expectations, amongst additional broader concerns. Moreover, we note that no serious concerns were raised about the pace of price inflation in the minutes of July’s BoJ meeting, which would have been a clear signal of a more substantial shift in the policy outlook. We consider the likelihood of rate cuts this year as less than 20%.

In the absence of fresh rate cuts, the BoJ is likely to consider alternative policy changes at its September meeting to demonstrate an ongoing accommodative policy bias. First, the BoJ could extend its forward guidance (currently “as long as it is necessary for maintaining that target in a stable manner”). Second, it may consider more technical adjustments for yield curve steepening, albeit we do not believe it would be particularly effective.


Reversal rate

A paper by Brunnermeier & Koby (2018) [1] defined the reversal interest rate as the rate below which policy easing does not provide more policy accommodation, but ‘reverses’ and becomes contractionary for lending. They explain this as a function of reduced net worth. Specifically, interest rate reductions increase capital gains on bank assets holdings but lower their net interest income (NII) on new businesses. This is partially the case as banks have difficulty in passing on associated rate reductions to their liabilities i.e. deposits, particularly in the household sector. If the decline in NII exceeds capital gains, then the bank’s net worth declines. Due to increasing capital constraints – i.e. Basel III – banks are then forced to adjust their risk-weighted assets, or in other words loans reducing their ability to lend.

Japan has faced a prolonged low rate environment since the end of the 1990s and negative interest rates since January 2016. According to S. Shirai[2], the NIRP has also raised a number of broader concerns beyond the possible perverse impact on bank lending:

A decline in the profitability of the financial sector and potential financial instability risk. NIRP has squeezed the spreads between lending and deposit interest rates in a number of financial institutions beyond banks. Moreover, Japanese Government Bond (JGB) interest income has also decreased. This has contributed to relatively radical decisions to ensure the viability of insurance companies, including the suspension of providing savings type insurance plans and/or raise premiums for new clients.

NIRP contributes to keeping unproductive firms alive. The “zombification” of Japanese firms is not new, but has not completely ended in Japan, with ever easier monetary policy allowing unproductive firms to survive. This impairs the usual “creative destruction” process which should ultimately raise the overall productivity of the country. It may also add to financial stability concerns.

Cash substitution. We observed an increase in notes in circulation since 2013 but this has accelerated in 2016, rising 6.5% year-on-year (yoy). Since 2017, the growth rate has stabilised around 4%yoy. This increase is due to conversion from deposits into cash holdings in safety boxes, reflecting households’ attempts to protect the nominal value of their deposits, with a guaranteed zero percent nominal return. This increasing proportion of cash in the Japanese economy makes it even more complicated for BoJ policy to transmit the desired effect.

Reduced JGB market liquidity. Massive purchases of JGBs under Quantitative and Qualitative Monetary Easing (QQE) have led to a deterioration in the liquidity and functioning of JGB markets. The NIRP amplified these conditions further and constrained market participants to transact actively to avoid negative territory. Moreover, there is a growing scarcity of JGBs, with the BoJ currently holding 42% of total government debt (Exhibit 1) – in turn, this has led to a shrinkage of related monetary market activities.

BoJ operational challenges and balance sheet risk. NIRP is not completely consistent with the QQE objective. The banking sector is dominant in Japan, with ample deposits from retail customers, so the size of securities markets like corporate bonds and asset-backed securities are relatively small. As a result, JGBs are the main financial assets the BoJ purchases from banks to expand the monetary base, though banks need to hold them. Large banks could sell JGBs to avoid potential interest rate risk and earn revenue from investing abroad. But regional banks have limited alternatives: holding JGBs until maturity to earn a small, but positive, coupon rate or selling JGBs to earn the +0.1% on excess reserves. This has created a top tier system[3] – suggesting the tiering structure has created incentives for banks to sell JGBs.

Inflation expectations key to policy rate

Inflation expectations are also crucial to monetary policy and the level of the policy rate. A central bank has to consider both its influence on setting those expectations, and the impact of these expectations on the scale of policy accommodation in the economy.

Despite their central importance for monetary policy, measuring inflation expectations is not easy and tends to diverge depending on who you ask. In Japan, the percentage of households expecting inflation above 0% in 12 months has reached a peak in recent months. Even allowing for an impact from the next consumption tax hike, household inflation expectations appear to be slowly increasing (Exhibit 2).

Financial markets are less optimistic. According to markets, the 5Y5Y inflation swap rate – a measure of what five-year inflation expectations will be in five years’ time – has retraced to its lowest level since 2016 (+0.1%). However, we note that due to the scarcity of Treasury Inflation-Protected Securities (TIPS) in Japan, illiquidity may be distorting breakeven inflation measures, creating a downward bias to the implied inflation expectations measure.  Finally, the Japan Center of Economic Research publishes professional forecasts of ‘core’ inflation. These suggest a more modest rise, which we associate with the consumption tax hike, before reverting to a level similar to 2016-17 (Exhibit 3).

Given variations in each measure, we provide an additional estimate of inflation expectations. We construct our own state space model[4] and estimate an inflation expectations term by assuming that inflation still reacts to the economic cycle. In such a model, core inflation adjusted from taxes will depend on inflation expectations – an unobserved variable – and lagged output gap.

Exhibit 3 illustrates our estimates and suggests that inflation expectations fell in the 2000s and only returned into positive territory in 2014, one year after the QQE announcement, before reaching a peak in 2016 and declining – interestingly, after the implementation of NIRP. On this measure, expectations appear to have stabilised around 0.2%, a low level compared to the BoJ’s target of 2%, but still towards the top end of the range over the past two decades.

The immediate decline in inflation expectations after the announcement of the NIRP has also been observed in a recent paper published by the Federal Reserve Bank of San Francisco[5]. According to the authors, “the reaction stressed the uncertainty surrounding the effectiveness of negative policy rates as expansionary tools when inflation expectations are anchored at low levels”. However, this paper referred only to the short-term market. The paper also referred to similar perverse reactions following unanticipated conventional US policy easings. Moreover, it is likely that inflation expectations and monetary policy were both influenced by exogenous factors, namely a decline in activity and fall in oil prices. While we do not see solid evidence to suggest that NIRP leads to a decline in inflation expectations there are clearly doubts over the efficacy of the policy to deliver a rise in inflation expectations.

Inflation expectations alone are insufficient to analyse the BoJ’s monetary policy. The Taylor rule is a classic tool to indicate appropriate policy levels. However, due to an extremely long low interest rate regime – the BoJ reached the zero-lower bound in 1999 with few major changes in the policy rate since – this is difficult to identify. Again, applying a state space model to the Taylor rule, we estimate a ‘shadow’ rate, rather than policy rate (Exhibit 4). The shadow rate, estimated by Krippner[6] from shadow/lower bound term structure models, is useful for monitoring the stance of unconventional monetary policy and quantitative analysis. It depends on the difference between the estimated inflation expectations with the inflation target, the output gap (estimated by the Cabinet Office) and the equilibrium nominal rate of interest  which is the unobserved variable4. As explained in one of our previous comment[7], i* can be defined as a “more pragmatic version of neutral rate, allowing for additional cyclical factors […]. These may include changing regulation, de-leveraging, sentiment and animal spirits. This makes i* more volatile and more susceptible to short run shifts”.

Using this formulation, we note that the estimate of coefficient  is significant and equal to 0.87, suggesting the BoJ is prioritising its inflation mandate.

Moreover, from a Taylor rule perspective, if we consider a plausible scenario where the output gap continues to slowly converge towards zero, i* should remain constant. This implies that the BoJ policy rate could only stay at the current level if inflation expectations were rising. If this is not the case, the implication is that the BoJ will likely have to consider easing policy further.

Medium term risk of a cut to persist

Of course, a Taylor Rule approach only affords a broad approximation of the policy rate over the medium term. However, if inflation expectations remain stable, or retrace further, this would add to pressure for further BoJ policy easing. In the shorter term – that is, after this month’s meeting – we would also consider an escalation of trade disputes between the United States and China and yen appreciation as a possible trigger for further policy easing. We estimate that ¥100 per US dollar would be a trigger for rate cut.

In such a scenario, the BoJ would likely implement measures to mitigate the side effects on financial institutions. Among the likely options would be:

  1. A Japanese version of the ECB’s Targeted longer-term refinancing operations programme. The idea is to provide funds to financial institutions at negative rates, as opposed to 0% currently. Specifically, banks could borrow from the BoJ at a negative rate and credit the reserve to the zero-rate Macro Add-on Balance[8]. Such a measure, associated with a depo rate cut, should be neutral in the short run, but in the medium term would depend on the bank’s ability to boost loan growth above current demand. Boosting lending is of course a key feature of monetary policy, but also runs the risks of creating bubbles 
  2. Increasing the amount accepted in the zero-interest rate tier system
  3. Raising the rate on the positive tier system (from 0.1% to 0.2%)
  4. Shortening the target duration for long-term rates under the yield curve control from 10 years to five years.

Nevertheless, such measures risk neutralising any stimulative impact from lower policy rates in the first place and may be politically difficult to implement.



[1] Brunnermeier, M.K. and Koby, Y. “The Reversal Interest Rate”, March 2018

[2] Shirai, S., “The bank of Japan’s super-easy monetary policy from 2013-2018”, ADB Institute, Nov 2018

[3] cf. Section on Negative Interest Rate Policy in Appendix, page 5

[4] There are two main benefits to representing a dynamic system in state space form. First, the state space allows unobserved variables (known as the state variables) to be incorporated into, and estimated along with, the observable model (i.e. Phillips curve). Second, state space models can be analysed using a powerful recursive algorithm (i.e. Kalman filter). State space models have been applied in the econometrics literature to model unobserved variables: (rational) expectations, missing observations, the non-accelerating rate of unemployment etc.

[5] Christensen, J.H.E and& Spiegel, M., “Negative Interest Rates and Inflation Expectations in Japan”, FRBSF Economic Letter, 26 August 2019

[6] Krippner’s shadow rate is useful but constraining. The fall in shadow rate helps to interpret accommodative monetary policy but it gives a misleading interpretation about i*. In fact, it falls down with the shadow rate in the opposite direction of the BoJ’s balance sheet expansion. This is what we must have in mind when interpreting i*.

[7] Page, D., “The best guide for US Treasury yields points upwards’, AXA IM Research & Strategy Insights, 22 March 2018

[8] Cf. Appendix section on NIRP where we detail the three-tier system, page 5


This document is for informational purposes only and does not constitute investment research or financial analysis relating to transactions in financial instruments as per MIF Directive (2014/65/EU), nor does it constitute on the part of AXA Investment Managers or its affiliated companies an offer to buy or sell any investments, products or services, and should not be considered as solicitation or investment, legal or tax advice, a recommendation for an investment strategy or a personalized recommendation to buy or sell securities.

It has been established on the basis of data, projections, forecasts, anticipations and hypothesis which are subjective. Its analysis and conclusions are the expression of an opinion, based on available data at a specific date. All information in this document is established on data made public by official providers of economic and market statistics. AXA Investment Managers disclaims any and all liability relating to a decision based on or for reliance on this document. All exhibits included in this document, unless stated otherwise, are as of the publication date of this document. Furthermore, due to the subjective nature of these opinions and analysis, these data, projections, forecasts, anticipations, hypothesis, etc. are not necessary used or followed by AXA IM’s portfolio management teams or its affiliates, who may act based on their own opinions. Any reproduction of this information, in whole or in part is, unless otherwise authorised by AXA IM, prohibited.

This document has been edited by AXA INVESTMENT MANAGERS SA, a company incorporated under the laws of France, having its registered office located at Tour Majunga, 6 place de la Pyramide, 92800 Puteaux, registered with the Nanterre Trade and Companies Register under number 393 051 826. In other jurisdictions, this document is issued by AXA Investment Managers SA’s affiliates in those countries.

In the UK, this document is intended exclusively for professional investors, as defined in Annex II to the Markets in Financial Instruments Directive 2014/65/EU (“MiFID”). Circulation must be restricted accordingly. 

© AXA Investment Managers 2019. All rights reserved