The claim that 'variable interest rates will definitely rise' was correct—yet why it should not be used as a basis

Revised September 2026

‘I was told that variable interest rates wouldn’t change, so I took out a variable-rate loan. Now that they are rising, I’m honestly anxious…’

Currently, more and more people are voicing similar anxieties.
No one can accurately predict how interest rates will move in the future.
However, what I am about to write is not about how ‘predictions are impossible’.

【Predictions may trigger a decision, but they should not be used as a basis】

I will explain why I can say this by reviewing my own article from that time.

Do you remember when people said, ‘Variable interest rates won’t change’?

This article was originally written on September 18, 2023.

At that time, the view that ‘Japanese variable interest rates will not change in the future’
was common even within the industry.

Low interest rates had continued for many years, so
it likely felt like the norm.

At that time, I intentionally took the opposite stance.

Variable interest rates will eventually change.

That is what I believed.

Furthermore, the structural discussion of ‘why variable interest rates were difficult to change’ was covered in detail in
The Mechanism of Variable Interest Rates—Why They Don’t Fall After Borrowing but Rise When They Do (Variable Interest Rates: Part 1), and Does Investing the Surplus from a Variable-Rate Loan Work?—Variable Interest Rates Do Not Vary (Part 2).

Also, eight days later, I wrote a separate article on the very topic that ‘one should not rely on predictions’—
‘Can You Rely on Mortgage Interest Rate Forecasts?—Why You Shouldn’t Rely on Predictions‘.

This time, I will not be discussing either of those, but rather verifying my own prediction from that time.

Was that assessment correct?

In conclusion, variable interest rates did indeed rise afterward.

On September 18, 2026, the Bank of Japan
raised its policy interest rate from 1.00% to 1.25%.

The DH Mortgage Index (variable rate, APR-based)
has risen to 1.37% as of September 2026.
(Source: DH Mortgage Index, as of September 2026)

In terms of direction, it turns out I was correct.

Should you really base your decisions on a prediction that happened to be correct?

Here, I will verify the figures I used at the time.
Back then, I touched on the concept of the neutral interest rate as a benchmark for how far interest rates might rise in the future.

According to a review published by the Bank of Japan in March 2026,
the estimate for the real natural rate of interest is between approximately -0.9% and +0.5%.
(as of the third quarter of 2025)

When adding the 2% inflation target for a trial calculation, the nominal range becomes quite broad, roughly between 1.1% and 2.5% (this nominal conversion is my own estimate).

The Bank of Japan itself can only provide a range of this magnitude.

The same can be said for economic forecasting experts.

In the ‘ESP Forecast Survey’ by the Japan Center for Economic Research (February 2026 survey, 36 respondents), the difference between the lowest and highest responses for the policy interest rate at the end of 2027 reached approximately 1.25%.

I will also look back at the bank’s expense ratio, which I touched upon in the article at the time.

Back then, I anticipated that if inflation continued for 10 years, the expense ratio would exceed 1.2%.
Over four years, the expense ratio has barely moved, from 0.63% to 0.65%.
On the other hand, the funding cost has risen significantly, from 0.55% to 0.84%.

It is true that costs are rising, but it was different from what I assumed at the time.

Even if the direction was correct, the figures used to justify it were sometimes off and sometimes accurate.
The fact that the direction was correct does not mean the reasoning behind it was correct.

Can you really maintain a one-year interest rate forecast for 35 years?

A mortgage is a product with a long duration of 35 years.
On the other hand, interest rate forecasts are only discussed for at most one to two years ahead.

I do not think interest rates are completely whimsical.
I personally believe there is a certain degree of continuity.

However, even so, it is hard to believe that a one-year forecast will continue to be correct for 35 years.

If you choose a mortgage based on a forecast, you would ideally need to keep updating that forecast every year and act correctly each time.

In reality, no one can do that.

Even if a prediction turns out to be correct, are there actually any measures you can take?

This is the most important point.

Even if you could read next year’s interest rates to some extent, the tools available to an individual for a mortgage are limited.
All you can choose is the fixed period, the interest rate type, and the loan amount.
You cannot use futures contracts, interest rate swaps, or options for a mortgage.

The tools that financial institutions and institutional investors use to control interest rate risk are not available to individuals taking out home loans.

Even if you build a precise strategy based on predictions, it is meaningless if you do not have the actual means to execute it.

Can you recover by refinancing?

Some people think, “Even if my prediction is wrong, I can just refinance later.” However, refinancing involves

significant costs, such as administrative fees and registration costs. It is safer to anticipate around 3% of your outstanding loan balance.

[Refinancing is a poor strategy for recovery]

If you are currently being encouraged to refinance to a low variable interest rate,
please pause and consider it carefully.

Unless the interest rate difference is enough to cover the costs, you will only lose money by refinancing. Moreover, that does not mitigate the risk of interest rate fluctuations.

A situation where you can flexibly switch the interest rate type you have chosen multiple times
does not happen easily.

How should you deal with predictions?

Let me summarize what we have discussed so far.

Predictions about which way interest rates might move
are fine to use as a starting point for thinking.
However, you cannot use them as the basis for your decisions.

There are three reasons for this.

1. Even if the direction is correct, the specific figures may not be.

2. It is not realistic to accurately predict a one-year outlook for 35 years.

3. Even if you are right, there are almost no means available to an individual to act on it.

Then, what should you base your decisions on?

There is a method of basing it on the tolerance of your own household budget,
specifically, how much you can endure if your predictions turn out to be wrong.

How to specifically prepare for this will be covered in another article.

Read also

“Variable interest rates do not vary (Part 1) ~Light and Shadow~” (August 23, 2026)

“Variable interest rates do not vary (Part 2) ~Where does the light gather?~” (August 25, 2026)

“Are mortgage interest rate forecasts reliable? —Reasons not to rely on predictions” (September 26, 2023)

Magazine “How to choose a mortgage without failing”

Regarding sources and calculation conditions

Variable interest rate: DH Mortgage Index (as of September 2026, APR-based)
Expense ratio/Funding cost: Japanese Bankers Association “Attached Tables” FY2025 Financial Results, Table 28 (Domestic business segment, nationwide bank basis)
Neutral interest rate: Author’s estimate adding the 2% price stability target to the real natural rate of interest estimated in the Bank of Japan Review “Trends in the Natural Rate of Interest and Evaluation of the Degree of Monetary Easing” (published March 27, 2026) (approx. -0.9% to +0.5%, as of Q3 2025)
Dispersion of policy interest rate forecasts by economists: Japan Center for Economic Research “ESP Forecast Survey” February 2026 survey
Noriaki Ogou | Mortgage Consultant. Over 6,000 consultations. 5 books authored.
#InterestRate #Mortgage #HouseholdBudget #VariableInterestRate #FixedInterestRate #Variable