Japan’s interest rate normalisation: Gradual and measured

The Bank of Japan began to step back from its ultra-accommodative policies in 2024, which has placed the spotlight on how interest rates may affect the country’s property market. At the September 2025 meeting, the central bank kept the short-term policy rate at 0.5%. We do not expect the Bank of Japan to shift to a strongly hawkish stance. The path ahead is likely to be gradual and guided by data.

One reason is the inflation outlook. Price growth is expected to ease below target, and the Bank of Japan’s own projections reinforce the case for patience. In its July forecast, the central bank estimated core CPI would remain in a range of 1.5% to 2.0% in fiscal 2026 and hover near 2% in fiscal 2027. That outlook does not justify rapid tightening and instead points to a slow process of normalisation that supports growth while anchoring expectations.

External dynamics also influence the outlook. The Federal Reserve and the European Central Bank are expected to adopt a more aggressive easing stance over the coming year. As rate differentials narrow, the yen should strengthen against the dollar and the euro. A stronger currency reduces imported inflationary pressures and lessens the urgency for the Bank of Japan to accelerate policy normalisation.

Fiscal constraints add another layer. The Ministry of Finance has estimated that every one percentage point rise in interest rates increases annual debt service by about 3.7 trillion yen (approximately USD 25 billion) within three years. With public debt exceeding 250% of GDP, even small increases in borrowing costs can worsen the fiscal burden. This reality makes it difficult to justify a rapid and large increase in policy rates, as the government needs to maintain market confidence while limiting fiscal strain.

Taken together, these considerations suggest a clear baseline scenario. Policy normalisation is likely to continue, but at a measured pace. Inflation is converging toward target, currency dynamics should ease near-term price pressures, and the scale of Japan’s debt makes aggressive tightening difficult to sustain. For real estate markets, this implies that upward pressure on discount rates will remain moderate compared with shifts in long-term Japanese government bond yields. Cap rates are still expected to move higher over time, but the adjustment should be limited given resilient income profiles across key property sectors.

Complex relationship between rates and property pricing

Conventional wisdom informs that higher risk-free rates translate directly into higher cap rates and, consequently, lower property values. In practice, however, this neat linkage has proven somewhat unreliable. Historical data across markets show periods where bond yields and cap rates rise together, but also long stretches where they move independently, and at times even in opposite directions. The relationship is indeed much more complex.

One important reason lies in the distinction between real and nominal interest rates. When inflation pushes nominal yields higher, it often supports rental growth and property income at the same time. The increase in cash flows offsets the effect of higher discount rates, leaving cap rates broadly unchanged. Growth expectations add yet another dimension. If investors believe property income will grow, they are more willing to accept tighter yields, rationalising that future cash flow will compensate for a lower entry return. Conversely, when growth prospects weaken, cap rates widen regardless of whether the cost of capital is rising.

To add, appraisal-based valuation methods and the illiquidity of direct property transactions can create a natural lag between movements in financial markets and adjustments in property pricing. Beyond these valuation mechanics, broader capital market conditions exert considerable influence. The trajectory of cap rates often reflects shifts in credit availability, investor risk appetite, and the flow of capital into real estate. When liquidity is strong and investors are confident, yields can stay tight even as borrowing costs rise. Conversely, when risk tolerance diminishes or lenders tighten their covenants, cap rates may expand even if government bond yields are inching down.

Japan property market displaying inertia

The rise in government bond yields has passed quickly through to higher cap rates in many markets post-pandemic.

In the United States, the sharp rise in Treasury yields since 2022 fed quickly into higher discount rates, with values repricing rapidly in a deep and liquid market. Europe followed a similar path, with cap rates in the UK, Germany and France rising broadly in line with bond yields, while some Southern European markets saw income growth soften the adjustment. Australia real estate also repriced fairly quickly, as higher bond yields and tighter credit conditions pushed office and retail cap rates upward.

Japan, while arguably in the initial phase of the rate normalisation cycle, has seen a far slower adjustment. For example, in the multifamily sector, structural demographic patterns, urban concentration, and preferences for rental housing underpin steady occupancy and income visibility. Even as borrowing costs edge up, the durability of cashflow provides a counterweight that limits upward pressure on cap rates.

We also observe the persistence of wide yield spreads relative to government bonds in Japan. This buffer provides room for interest rates to normalise somewhat without forcing a knee-jerk repricing of real estate assets. Investors remain willing to accept lower spreads in exchange for stability and predictable income, particularly in a low-volatility market like Japan.

The funding environment also contributes to this stickiness in property pricing. Japanese lenders have remained broadly accommodative, with credit availability preserved despite higher policy rates. This stands in contrast to other regions, where tightening bank lending standards typically compound the effect of rising rates on property values.

This divergence underscores the need for a more structured framework to explain what drives cap rates beyond the risk-free rate alone. It is useful to move beyond anecdotes and apply the lens of financial theory. In this paper, we turn to a fundamental valuation model in finance that helps decompose the drivers of cap rates.

The financial theory perspective

The Gordon Growth Model¹, also known as the dividend discount model, provides a simple but powerful framework for valuing income-generating assets. It rests on the principle that the value of an asset equals the present value of its future cash flows, assuming those flows grow at a constant rate into perpetuity. When investment professionals say that real estate must trade at a spread over bonds, they are really pointing back to the intuition at the heart of the Gordon Growth Model.

The basic Gordon Growth Model applied to real estate is:

Where:

• p = property value
• NOI = net operating income
• r = required rate of return (essentially the discount rate)
• g = expected long term NOI growth rate

Now, the required rate of return represents the minimum return an investor demands for holding an asset. The starting point is the risk-free rate, which is the baseline return that any rational investor would demand. Real estate, however, carries risks that go beyond those embedded in a plain vanilla government bond. To be compensated for these uncertainties, investors require an additional return above the risk-free rate. This additional margin is the risk premium.

As such,

Where:

• rf = risk-free rate
• RP = risk premium (the extra return above the risk-free rate that investors demand for real estate’s risk)

Within this framework, three factors determine the level of a cap rate: the risk-free rate, the risk premium associated with real estate, and the expected growth of net operating income. If both the risk premium and growth assumptions are held constant, an increase in the risk-free rate feeds directly into a one-for-one rise in the cap rate. This is the assumption often made in practice. When bond yields rise, property yields are expected to follow by roughly the same margin to preserve the spread.

The Gordon Growth Model, however, makes clear that the adjustment is not always mechanical. Cap rates may remain steady, or even compress, when investors are prepared to accept a narrower premium or when they anticipate stronger income growth. In Japan, this nuance has been especially evident. For much of the past fifteen years, ultra-low bond yields left real estate trading at historically wide spreads. As interest rates have begun to rise, we posit that investors have allowed the risk spread to compress rather than demand a full repricing of property yields. The result has been a degree of stickiness in Japanese cap rates, as the starting risk premium buffer has absorbed much of the recent upward pressure from higher government bond yields.

Extending the Gordon Growth Model: Endogenous risk premium

Investors do not set their return targets in a vacuum. Market realities such as investor preferences, capital structures and availability of assets shape how much excess return investors require. When rates rise, investors could accept a smaller risk premium if they still feel confident about the income and the asset’s quality. But this willingness has limits. The premium cannot shrink to nothing, and if bond yields were to rise substantially, investors would eventually require compensation.

We postulate that the risk premium should be treated as endogenous, responding to changes in the risk-free rate rather than standing as an exogenous factor. Cap rates are most sensitive when spreads are compressed, but additional widening has diminishing marginal impact. At higher bond yields beyond a certain point, the spread is narrow, and investors will demand that property yields adjust with government bonds. This creates what we refer to as “diminishing flexibility” in required risk premiums. The dynamic is captured more realistically with an exponential form rather than a linear one.

We model the risk premium as an exponential decay function of the risk-free rate.

Where:

• ϕ = the asymptotic floor that investors require minimally
• A = the headroom available to compress when rates are low,
• λ = the compressibility factor, or the headroom can be used up

Essentially, as the risk-free rate (rf) goes up, the compressible portion (Ae^-λrf) shrinks, and the risk premium gets smaller and moves toward the floor ϕ.

We make this compressibility explicit and test the model on the Tokyo office and multifamily sectors. The dataset spans the last fifteen years with a quarterly frequency. For each market segment we carry a single consistent cap rate series and match it to the local ten-year government bond yield in each period. NOI growth information is retained for robustness. Cap rates and risk-free rates were aligned by lagging cap rates one quarter relative to the 10-year government bond yield. This reflects the reality that value adjustments typically occur with a delay.

The regression outputs are as below:

Pass through into cap rate for every 100bps increase in risk-free rate at different risk-free levels:

Analysis and interpretation of outputs for the multifamily sector:

The pass-through profile for Tokyo multifamily shows a clear transition from cushioning at very low interest rates to meaningful rate sensitivity as the risk-free rate rises. At a 0.5% starting point for instance, a 100 bp increase in the risk-free rate is associated with a tightening in the modelled cap rate. This counter-intuitive result reflects how much “excess” spread is embedded when rates are near zero for a long time. Investors are willing to give up on the risk premium quickly as policy normalises at very low levels, and the compression in the spread more than offsets the rise in the base rate. As the risk-free rate moves into the 2.0% zone, that cushion thins out and the pass-through climbs to the 20bps to 60bps range.

The economic rationale is straightforward for Japan’s institutional rental housing. Income streams are granular and resilient, vacancy volatility is low, and long-horizon capital continues to anchor asset demand. In the Japan multifamily sector, risk-free rates will have to rise to around 3.8% before the sector exhibits a close to one-for-one movement in cap rates.

Analysis and interpretation of outputs for the office sector:

The Tokyo office exhibits the same pattern but with its own nuances. At low levels of risk-free rates, the pass-through of a rate increase is similar to the multifamily segment, in which investors surrender the initial spread. The combination of limited prime CBD stock, leases with strong tenant covenants, and a deep domestic core buyer base allows early rate moves to be absorbed by spread compression. As the risk-free rate moves toward 2.0%, pass-through remains modest (about 49 bps per 100 bps increase in bond yields), then rises steadily into a materially rate-sensitive range at 3.0% to 4.0%. Once the easiest spread compression is used up, cap rates share more of the subsequent rate moves.

The consequence of strong corporate performance and resilient labour market is the increased demand for and support on occupier performance in the office sector. Indeed, occupier sentiments are improving across the real estate markets. For instance, Tokyo’s office rents appear to have found the floor in 2023, seeing a positive rental growth recorded in 2024 and into 2025. Vacancy rate also dropped to a three-year low to 3.4% as at June 2025. In our model, risk-free rates will have to rise to around 5.0% before the office sector exhibits a close to one-for-one movement in cap rates.

Conclusion

In summary, the practical takeaway is that Japanese real estate can be relatively resilient in a lift-off from ultra-low rates, provided rate increases are gradual and accompanied by improving fundamentals.

Japanese institutional portfolios have long been calibrated to thin returns. A modest rise in government bond yields is unlikely to reset required returns overnight, especially while other asset classes remain relatively low yielding. Many investors have liability-matching mandates and long investment horizons, so they continue to favour stable, income-producing real estate even as spreads compress. In addition, structural features of Japan’s property market, including a limited supply of tradeable core assets, deep domestic demand, and relatively low volatility, support a persistently lower risk premium than in more cyclical real estate markets globally.

Our theoretical extension of the risk premium compressing as rates rise has been borne out by the market’s behaviour so far. Fears of a sudden crash in Japan’s property prices due to the BoJ tightening are overblown. Instead, what we are seeing (and expect to continue seeing) is a measured adjustment. Government bond yields will move up slowly and very marginally while spreads narrow with cap rates inching only slightly. Indeed, property values in Japan have continued to hold firm in the past year, reflecting confidence in Japan’s long-term real estate outlook and structural demand drivers.

Japan’s case illustrates that interest rate sensitivity in real estate is very much about investor psychology and market structure. We have argued and shown that investors adjust their return expectations in ways that buffer the Japan real estate market from rate shocks. This does not mean cap rates are completely divorced from interest rates. Indeed, over a longer horizon, cap rates must eventually reflect higher base rates, but the journey is slow and nonlinear.

As Japan enters a new chapter of slightly higher interest rates, the real estate sector appears poised to navigate it without severe dislocation. Barring an unforeseen shock, we expect that interest rates will likely normalise gradually and minimally, and the real estate sector has the tools (and the investor mindset) to adapt. For current and prospective investors, understanding this dynamic provides confidence that Japan can remain a solid income play and portfolio diversifier.

Appendix: Assumptions and limitations

While our analysis provides a coherent framework for understanding cap rate sensitivity in Japan, it is a simplified representation of a complex market.

It is important to acknowledge assumptions and limitations in this framework. These caveats do not undermine the usefulness of the findings, but they do underscore the need for caution in extrapolating them beyond the specific data and contexts examined.

Reliance on the Gordon Growth Model: We relied heavily on the Gordon Growth Model. In practice, cap rates are shaped by a much wider set of factors, including rental growth expectations, liquidity conditions, the availability and cost of credit, and investor sentiment. Also, the growth rate can vary with macroeconomic conditions, asset-specific factors, and inflation. For Japan, we assumed low and stable long-term growth, which is a reasonable simplification. Future research should model time-varying growth expectations and consider how other drivers feed through to cap rates.

Simplified risk premium function: We proposed an exponential function for the risk premium as a function of r_f. This was a theoretical construct to illustrate variable sensitivity. For researchers with sufficient historical data on cap rates and bond yields, it would be worthwhile fitting a nonlinear model to determine if indeed an exponential (or other functional form) describes Japan’s spread behaviour.

No explicit role for credit conditions: Credit availability and market liquidity can influence cap rates, and we did not include them in the model. In Japan, lending attitudes among major city banks and regional banks can potentially affect pricing. If banks tighten credit because of regulatory guidance or market stress, investors may demand higher cap rates regardless of Bank of Japan policy, since both the cost of capital and refinancing risk rise. If policy changes are paired with measures that keep credit flowing, for example through BoJ liquidity facilities, cap rates can remain lower than fundamentals based on rates alone would suggest.

Behavioural factors: Our model treats investors as largely homogeneous and fully rational. In practice, behavioural biases matter. One common bias is anchoring to nominal yield levels. For instance, Japanese investors may view a 3% cap rate as the norm and adjust only slowly when conditions change. Herding can also play a role. If large players such as J-REITs keep transacting at low cap rates, others may follow, which sustains pricing. These effects are difficult to measure.

1-Gordon Growth Model

Where

• P₀ = current stock price
• D₁ = dividend expected to be paid in the next period
• r = required rate of return
• g = dividend’s constant growth rate