How to Calculate IRR on a Tokyo Apartment (10-Year Hold, Full Worked Example)

TL;DR: Net yield tells you annual income return. IRR tells you total return — income plus capital gain or loss — on every yen invested, accounting for when it moves. For a Tokyo apartment, the difference between a 4.2% net yield and a 6.8% IRR can mean you’re actually doing well even when the monthly cash flow looks thin. And the moment you borrow, a third number takes over: cash-on-cash return (CoC), which is annual pre-tax cash flow divided by total cash invested — what your equity is actually earning after debt service. The math isn’t hard. Most investors just never run it.

Inputs you can check (primary data, 2026): the 23-ward average resale condo price was ¥1,326,900 per ㎡ in August 2026 (REINS) and the average condo asking rent was ¥5,170 per ㎡ per month (Tokyo Kantei, +6.4% YoY). Our arithmetic on those two published averages gives a gross yield of 4.68% before any costs — the starting point for the IRR below. Sources: Tokyo 23 Wards Property Data 2026.

---

I closed on a 1LDK in Meguro-ku three years into my portfolio. Net yield at acquisition: 3.9%. Every Tokyo-focused investor I mentioned it to raised an eyebrow. “That’s low.”

On income alone, they were right. But I had a 10-year IRR projection of 7.4% based on my expected exit, and after seven years that scenario is tracking. Net yield is a snapshot. IRR is the whole film.

What Is IRR and Why Does It Matter for Tokyo Property?

IRR — internal rate of return — is the discount rate that makes the net present value of all your cash flows (in and out) equal to zero. Translation: it’s the annualized return that accounts for the timing and size of every cash movement over the life of the investment.

For a property investment, IRR captures:

• Initial cash out (purchase + acquisition costs)
• Net operating income each year (after all costs)
• Debt service if you’re financing
• Proceeds from sale at exit
• Sale costs (agent commission, taxes on gain)

Net yield ignores the sale. Cap rate ignores financing and the sale. IRR ignores nothing.

In Tokyo specifically, IRR is critical because capital values move. Central ward properties have appreciated meaningfully over the past decade. Regional properties have often depreciated. A 4% net yield in Minato-ku with 2% annual appreciation gives you a very different IRR than a 7% net yield in a depreciating market.

From the desk — In a decade of closings, the clients I’ve watched do best are the ones who treated the exit assumption as the whole argument and stress-tested it before signing. The ones who got burned almost always backsolved a number they wanted to see, then called it analysis. When a low headline yield makes a buyer flinch, I’ve learned the real conversation is never about the rent today, it’s about what they honestly believe the unit sells for in year ten, and most have never once written that number down. The leveraged buyers I watch hesitate are almost always staring at a near-breakeven first year, and the ones who do well are those who understand the principal paydown is quietly building equity even when the cash flow reads slightly red.

Related reading: The Gross Yield Trap: How Japanese Listing Yields Are Calculated to Flatter Sellers.

Setting Up the 10-Year IRR Model: The Inputs

All figures below are illustrative — representative of this property type in this submarket, not a specific audited transaction.

A 1K in Nakameguro — walking distance to the station, 2005 build, 25 sqm.

Purchase:

• Asking price: ¥18,000,000
• Acquisition costs (agent fee, registration, taxes, scrivener): ¥1,300,000
• Total cash out at Year 0: ¥19,300,000 (assume all-cash for clarity; leverage covered below and in the FAQ)

Rental income:

• Current rent: ¥110,000/month
• Gross annual rent: ¥1,320,000
• Vacancy allowance: 8% → effective income: ¥1,214,400

Annual operating costs:

• Management fee (6%): ¥72,864
• HOA fee + building repair reserve: ¥168,000 (¥14,000/month)
• Fixed-asset tax + city planning tax: ¥90,000
• Insurance: ¥22,000
• Maintenance/repair reserve: ¥36,000

Net operating income (NOI): ¥1,214,400 − ¥388,864 = ¥825,536/year

Net yield on total invested: ¥825,536 ÷ ¥19,300,000 = 4.28%

The 10-Year Cash Flow Table

IRR needs year-by-year cash flows. Conservative model: rent grows 0.5%/year from Year 3 (when current tenant likely turns over), costs grow 1.5%/year due to rising repair reserves.

Year	Gross Rent	Vacancy	NOI (approx)	Net Cash Flow
1	¥1,320,000	−¥105,600	¥826,000	¥826,000
2	¥1,320,000	−¥105,600	¥820,000	¥820,000
3	¥1,326,600	−¥106,128	¥821,000	¥821,000
4	¥1,333,233	−¥106,659	¥818,000	¥818,000
5	¥1,339,899	−¥107,192	¥815,000	¥815,000
6	¥1,346,599	−¥107,728	¥812,000	¥812,000
7	¥1,353,332	−¥108,267	¥809,000	¥809,000
8	¥1,360,099	−¥108,808	¥806,000	¥806,000
9	¥1,366,899	−¥109,352	¥803,000	¥803,000
10	¥1,373,734	−¥109,899	¥800,000	¥800,000

Small year-on-year NOI decline reflects rising repair reserves and costs growing faster than rents. Realistic for a 2005 build entering its second major maintenance cycle.

Related reading: Vacancy and Tenant Risk in Tokyo Rental: Underwriting the Downside.

The Exit: Sale Price and Net Proceeds

Year 10 exit. The Nakameguro 1K market doesn’t collapse — but it doesn’t boom either. Conservative model: 1% nominal annual capital appreciation on a starting value of ¥18,000,000.

¥18,000,000 × (1.01)^10 = ¥19,876,000 (rounded: ¥19,900,000)

Sale costs:

• Agent commission: 3% + ¥60,000 + tax = approx ¥720,000
• Other closing admin: ¥50,000

Net sale proceeds: ¥19,900,000 − ¥770,000 = ¥19,130,000

Year 10 combined cash flow (NOI + net sale): ¥800,000 + ¥19,130,000 = ¥19,930,000

Calculating the IRR

Cash flow series for IRR:

• Year 0: −¥19,300,000
• Years 1–9: +¥826,000 down to +¥803,000 (as above)
• Year 10: +¥19,930,000

Using Excel/Google Sheets =IRR() on this series: approximately 5.3% IRR

That’s the all-cash, pre-tax, 10-year IRR on this illustrative Nakameguro deal. 5.3%. Not thrilling. But compare it to:

• Japanese savings account: 0.02%
• 10-year Japanese government bond: ~1.5% (as of writing)
• Tokyo REIT index average distribution yield: ~3.5%

And the property has a built-in leverage option, is denominated in yen (no FX risk if you’re already earning yen), and gives you operational control.

Now run the same model with 2% annual capital appreciation instead of 1%:

Exit price: ¥18,000,000 × (1.02)^10 = ¥21,946,000 Net proceeds: ¥21,946,000 − ¥770,000 = ¥21,176,000 Year 10 combined: ¥800,000 + ¥21,176,000 = ¥21,976,000

IRR: approximately 6.5%

The rental income barely changed. The IRR jumped 1.2 percentage points because capital appreciation at exit compounds powerfully. A “low yield” central Tokyo property can still be a good total return investment. Exit assumption carries most of the weight.

Adding a Loan: Cash-on-Cash Return

The model above is all-cash. The moment you borrow, net yield and cap rate stop telling the full story — they are asset-level metrics that ignore financing. Cash-on-cash return (CoC) is the yield number that survives a loan. It answers one question: what is my equity actually earning?

CoC = Annual pre-tax cash flow ÷ Total cash invested

Annual pre-tax cash flow = NOI − Annual debt service (principal + interest)

Total cash invested = Down payment + Acquisition costs

This is the after-financing, before-tax income your equity is generating. It’s the number that tells you whether borrowing made sense.

A Tokyo property with a 4.2% net yield can generate a 7.8% cash-on-cash return for a leveraged buyer. Same building. Same rent. Same costs. Different return — because cheap leverage amplifies equity returns. Japanese mortgage rates have been extraordinarily low for a long time. When borrowed money costs 1.5–2.5% and the asset earns 4.2% net, the spread goes to your equity. That’s positive leverage. When borrowed money costs more than the asset earns, CoC drops below cap rate. Right now in Japan, positive leverage is still accessible for most buyers. That won’t last forever. The math turns quickly when it changes.

Use cap rate when you’re comparing two assets independently of how you finance them, discussing valuation with a seller, or benchmarking against other asset classes. Use CoC when you’re borrowing, need to know actual annual cash in/out, are deciding between financing options, or are stress-testing whether you can service debt during a vacancy. Cap rate tells you asset quality. CoC tells you whether your financing structure makes it viable.

A Worked Example: ¥20M Edogawa 1LDK

Figures below are illustrative — representative deal structure, not a specific property or guaranteed outcome.

Assumptions: ¥95,000/month rent, 10-year-old RC building, financed with 30% down + 70% loan.

Property details:

• Purchase price: ¥20,000,000
• Down payment (30%): ¥6,000,000
• Acquisition costs (6.5%): ¥1,300,000
• Total cash invested: ¥7,300,000
• Loan amount: ¥14,000,000
• Loan terms: 35 years, 2.0% (illustrative; actual terms vary by lender and borrower profile)
• Annual debt service: ~¥560,000 (principal + interest, year 1)

NOI calculation:

Item	Annual Amount
Gross potential rent	¥1,140,000
Vacancy (8%)	−¥91,200
Effective gross income	¥1,048,800
PM fee (5.5%)	−¥57,684
HOA fee (¥10,000/month)	−¥120,000
Building repair reserve (¥7,500/month)	−¥90,000
Fixed-asset tax + city planning	−¥90,000
Insurance	−¥18,000
Maintenance	−¥18,000
NOI	¥554,116

Cap rate: ¥554,116 ÷ ¥20,000,000 = 2.77%

Now subtract debt service:

Annual pre-tax cash flow: ¥554,116 − ¥560,000 = −¥5,884

Negative in year 1. Slightly, but negative.

CoC return: −¥5,884 ÷ ¥7,300,000 = −0.08%

Is this a bad deal? Not necessarily. The principal repayment is building equity. In year 1 on a 35-year, 2.0% loan, roughly ¥200,000 goes to principal and ¥280,000 to interest. Economic position:

• Cash flow: −¥5,884
• Principal paid down: ~¥200,000
• Net economic position: slightly positive, just not as cash

By year 5, if rents hold, cash flow turns positive. By year 10, CoC on the original equity invested can be meaningfully positive.

When Positive Leverage Works in Japan

The leverage arbitrage only works when the loan rate is below the cap rate. Let’s rerun with different loan terms to show the sensitivity — all illustrative:

Same property (¥20M, NOI = ¥554,116, cap rate 2.77%)

Loan Rate	70% LTV CoC	Comment
1.5%	+1.2%	Positive cash flow immediately
2.0%	−0.1%	Near breakeven, principal-building
2.5%	−1.4%	Negative cash flow, harder to justify
3.0%	−2.7%	Clear negative leverage

At 2.0%, you’re essentially at the tipping point. This is why the direction of Japanese interest rates matters so much to leveraged investors — a 1 percentage point rate increase on a 2.0% loan cuts across the entire cash flow positive/negative threshold.

For foreign buyers, loan access in Japan is already constrained — most major Japanese banks lend only to permanent residents or Japanese nationals. Regional banks, credit union banks (shinkin), and some specialized non-bank lenders do finance foreigners, but typically at higher rates (2.5–4.0%) and lower LTVs (50–70%). At 3.5% on a 2.77% cap rate asset, you have significant negative leverage. CoC is deeply negative. That’s a capital appreciation bet, not an income strategy.

Related reading: American Buying Property in Japan: A US Citizen’s Tax & Finance Guide.

Withholding Tax: The CoC Killer for Non-Residents

Foreign investors who are not Japanese residents face withholding tax on rental income. Under Japan’s tax code, a non-resident landlord’s Japanese-sourced rental income is subject to:

• 20.42% withholding on gross rental income if no tax agent is appointed
• Alternatively, appoint a Japanese tax management agent to file properly, paying income tax at progressive rates on net income

The gross withholding hits CoC hard. On ¥1,048,800 effective gross income, 20.42% withholding = ¥214,165/year off the top. That takes the already-thin cash flow and makes it structurally negative.

The fix: appoint a Japanese tax agent, file properly, and pay tax on net income (after deductions). Your effective tax rate on net income is lower than 20.42% on gross, and you can use costs against income. But this requires a Japanese accountant (zeirishi) familiar with non-resident taxation. Budget ¥50,000–¥150,000/year for this. It shows up in your CoC calculation — and in your after-tax IRR.

Where This Goes Wrong

IRR is sensitive to the exit assumption. Most of your return is in Year 10. Change the assumed exit price by 10% and IRR shifts 0.8–1.2 percentage points. Investors who backsolve a high IRR by plugging in optimistic exit prices are running a confidence exercise, not an analysis.

Japan adds a specific risk: depreciation of the building component. Japanese accounting standards depreciate wooden buildings over 22 years, reinforced concrete over 47 years. At resale, buyers recalculate based on remaining depreciation life. A 2005 RC build is fine for a 10-year hold. A 1990 build might face buyer resistance in 2034 due to remaining life concerns.

Also, capital gains tax in Japan for non-residents on property sale: if held under five years, 30.63% on the gain. Held over five years: 15.315%. Your after-tax IRR depends heavily on hold period. Model both.

On the leveraged side:

• Not modeling debt service changes over time. Some Japanese loans are fixed-rate; many are variable. A 35-year variable rate loan that resets every few years can significantly alter CoC if rates move. Run sensitivity.
• Ignoring principal repayment as equity building. CoC looks bad in early years of a long amortization loan because you’re paying mostly interest. But principal repayment is building asset value. Total return is CoC + equity accumulation + appreciation.
• Not accounting for withholding tax as a non-resident. This alone can turn a marginally positive CoC deal into a cash flow negative.
• Underestimating acquisition cost in the denominator. Some investors put just the down payment in the CoC denominator, forgetting ¥1M+ in acquisition costs. That’s cash out the door. It belongs in the denominator.
• Assuming the same CoC in year 1 as year 10. As the loan amortizes, interest payments fall, principal payments rise (same total outflow), but cash flow improves — CoC gets better over time on a fixed-rate loan with stable rents.

FAQ

Q: How does leverage change the IRR? Say you financed 60% (¥10,800,000 loan at 2.5% over 30 years). Your Year 0 cash out drops to about ¥8,500,000 (40% equity + acquisition costs). Annual debt service is approximately ¥511,000. Your net annual cash flow drops sharply — maybe to ¥300,000 early years — but your Year 10 equity payout rises because you’ve been amortizing the loan. On a positive-appreciation scenario, leveraged IRR on equity beats unlevered IRR significantly. On a flat or declining market, leverage destroys equity IRR.

Q: Can I use IRR to compare properties in different cities? Yes, and you should. A 7% gross yield in Sendai might have a much lower IRR than a 4.5% gross yield in Minato-ku if the Sendai property depreciates and the Minato property appreciates. IRR is the only metric that forces you to make your capital appreciation assumption explicit.

Q: What IRR should I target for Japanese residential property? Unlevered, pre-tax, as a rough directional benchmark: 5–7% is typical for quality Tokyo stock. Below 4%, you’re taking real estate illiquidity risk for bond-like returns. Above 8% unlevered, be suspicious of the exit assumption or the true cost structure.

Q: What CoC return should I target for Tokyo property? For a leveraged deal, even 1–2% positive CoC in year 1 is functional if you believe in capital appreciation. Some investors accept breakeven CoC for the first few years in central wards. If you need cash flow from day one, target 4%+ net yield assets in outer wards and use conservative LTV.

Q: Is a zero-cash-flow deal in Tokyo worth doing? Possibly, if three conditions hold: you believe in long-term Tokyo appreciation, you can fund the cash shortfall without stress, and the principal paydown is building meaningful equity. It’s a total return play, not an income play.

Q: How does CoC change if I use a shorter amortization (e.g., 20 years vs 35 years)? Shorter amortization means higher annual debt service, lower CoC in early years, but faster equity building. On the same ¥14M loan at 2.0%: 20-year annual service ≈ ¥848,000 vs. 35-year ≈ ¥560,000. Cash flow is worse early, but you own the asset outright in 20 years.

Q: Does IRR account for tax? Not automatically. Build it in by adjusting cash flows — subtract income tax on rental profits each year, adjust the exit for capital gains tax. The after-tax IRR is the number that actually matters. Few investors calculate it correctly. The same goes for CoC: before-tax CoC shows operating performance, after-tax CoC shows real return. For a non-resident filing properly on net income, Japanese income tax on rental profits at lower brackets can be 5–20% of net income. Model your specific situation with a zeirishi.

Q: What software should I use? Google Sheets with =IRR() is enough. Put Year 0 outflow in cell A1, Years 1–10 inflows in A2:A11, and =IRR(A1:A11) gives you the answer. If cash flows vary (they always do), use =XIRR() with actual dates for more precision.