🏯 Investing in Japan, Part 3

Konnichiwa! 👋 Welcome to the Alts Sunday Edition, and Part 3 in our series on Investing in Japan.

In ​​Part 1​​, we discussed Japan’s tourism supercycle, the culture of craftsmanship & longevity, and niche manufacturing.

​Part 2​ covered the generational shift remaking corporate Japan, now the world’s second-largest PE engagement market.

Today in Part 3, the subject is buying and renovating akiya homes, or empty houses.

Japan now has ​9 million empty houses​.

These vacant homes represent ​14% of the national housing stock​. It’s estimated that one in three homes will be vacant by 2038.

The internet is full of videos of foreigners buying homes for a few thousand dollars and lovingly restoring them.

Adding to the opportunity is the fact that Japan’s yen is ​down 35% in three years​​, currently at its weakest level against the dollar since 1986.​​​

Municipalities are adding ​​renovation grants​ up to ¥3 million​​($20,000) because they’d rather subsidize an outsider than manage another derelict property.

It’s all become a bit of a fantasy; abundant cheap homes, full of character and history, patiently waiting for someone to bring them back.

But what these renovation videos leave out is everything that happens after you buy. Everyone focuses on the low prices. I want to write about what they leave out.

To be clear, I did not come to this topic through real estate. I came through bonsai. See, the bonsai industry actually has a similar pattern: a shrinking population, nobody left to take over, trading inside a closed circle, and a long history accumulating anyway. The idea for this issue began when I saw that same pattern sitting under Japanese houses.

But I’m not an expert in this area, so I reached out to fellow Altea member ​Terrie Lloyd​ to help me understand this world.

Terrie has spent 41 years in Japan, and now buys up old houses, revitalizes them, and runs them as ​HINOKI Stays​ under his Akiya 2.0 project.

Terrie also founded Japan Travel KK, who is helping plan and organize our upcoming Investor Trip.

He answered my questions in writing, using the real figures from his own deals. His numbers run through the rest of this piece.

Let’s go!

The prices in those videos are real. What goes unsaid is why they are so low.

Japan has ​9 million vacant homes​. That’s one house in seven, or a full 13.8% of the housing stock, the highest on record. And the number keeps climbing as Japan’s population falls.

So it’s more accurate to read that cheapness as a description of where Japanese housing stands, than as proof you have found a bargain.

One week in Japan. A lifetime of stories.

Japan’s empty houses are a big theme for October’s investor trip.

Five days in Japan. A private distillery tour, snow monkeys in Nagano, small-cap deep value, and the akiya thesis straight from the guy building it. ​Come with us →​

In 1971, a Yale student named Alex Kerr was hitchhiking around Japan reached the Iya Valley in Tokushima, and made it the subject of his senior thesis.

Alex looked at more than a hundred abandoned houses before ​buying one in 1973 for ¥380,000​ ($2,400), borrowed from a family friend.

Three hundred years old, and empty for seventeen. ​Chiiori, Kerr’s farmhouse in Iya. Photo: Motohiro Sunouchi, CC BY 2.0

What happened next matters most here. While land prices rose across Japan through the bubble, Iya kept falling. He had bought land with no next buyer.

He re-thatched the roof, brought the house back, and wrote a famous book called ​Lost Japan​, whose Japanese edition won the Shincho Gakugei prize in 1994.

​Romance, he said, “is like bread or rice.” People need a bit of it just to live, and that demand drives this asset class.

Lost Japan explores Kerr’s 30-year journey through Japan, serving as both a love letter to traditional Japanese culture and a critical lament over its architectural destruction. It was the first time a foreign writer had taken the Shincho Gakugei prize.

His record carries the warning too. ​Chiiori still operates as a guesthouse​, while Iori, the Kyoto machiya business he founded in 2004 and later left under separate management, ​closed in January 2024​.

One person, fifty years, both outcomes.

Unlike Mexico, Thailand, Bali, or the visa-linked schemes of southern Europe. Japan places no restriction on foreign ownership.

There’s no visa, no residency, and buying without setting foot in the country is possible under the rules, which is unusual.

​A traditional Japanese family residence. It literally has it’s own emoji: 🏯 Photo: TANAKA Juuyoh, CC BY 2.0

Most sellers are not property companies. They are aging innkeepers, grandchildren who live three prefectures away, titles nobody has updated in two generations, and heirs facing an ​inheritance tax as high as 55%​.

For them this is less selling an asset, and more like disposing of a burden. It’s also often the moment a family history quietly ends. The entrance is open because of what is happening on the other side of it.

For most buyers this is an asset for living rather than for income, and pricing it that way from the start saves the disappointment.

Most of that money is the restoration, not the house. The chart below shows the split.

Non-residents rarely get a Japanese mortgage, so assume cash. And YouTube videos often leave out property tax of up to 1.7% a year, plus a 20.42% withholding on any rent you earn.

Another cost which never appears on an estimate is craftsmen. Like so much else, the number of people who can actually restore these houses keeps falling, and money no longer guarantees the work gets done.

The heaviest cost arrives twenty years out, when you sell. Central Tokyo will always have a next buyer, but a random valley may not. What separates them is liquidity rather than charm or yield.

Let’s look at one of Terrie’s own deals: Amazaki.

HINOKI Stays “Tea House,” before renovation. Photo: Akiya 2.0

The house, including land, cost ¥2-4M (roughly $12,000-$25,000). It was a small traditional building of about 120 sqm, in Terrie’s words, the property was “lower but not inferior quality.”

Renovation, furniture and fittings at came to about ¥16M. Landscaping, which he insists on counting separately, another ¥2M, so ¥22M all in, or about $135,000.

The plan assumed 70% occupancy across a 180-day season

Terrie estimated 120 nights at ¥27,000 per night ($170/night) with cleaning charged on top,

This would yield ¥3.24M a year at the low end ($20,000) and roughly ¥10M of profit after costs over ten years ($63,000)

Depreciate the building straight-line over 22 years and about ¥10M of value is still standing

Over ten years, gross income and the home’s residual value return about ¥42M on the ¥22M that went in, or roughly $258,000.

So far, reality has run ahead of the plan.

The actual nightly rate is closer to ¥40,000 ($250/night), because he put a good sofa bed in the living room and can now take 4-6 guests.

“Japanese groups in particular are very happy to rent at ¥10-12K per person per night if it means they get a kitchen and some luxury.” — Terrie Lloyd

Net of operating costs, and before income tax on either side, that unit is looking to yield roughly 6.5-10% on the ¥22M invested.

Set that against central Tokyo, where a small unit at ¥50M, about $305,000, gives ​3.5% gross and 2% net​, liquid and professionally managed. Roughly 3-5x the net yield, which is what the countryside pays you for doing the hard renovation work and taking on a house you may never have a buyer for.

Last year, the average for a used 70sqm unit in Tokyo ​passed ¥100M​ ($633k) for the first time since records began in 1997. Occupancy sits at ~97%, and the weak yen has cut the foreign entry price by roughly a third against 2020-21 rates. Photo of Azabudai Hills by Alex Pagnotta / Unsplash

So the countryside wins on yield and loses on liquidity: a Tokyo unit has a next buyer, a beautiful rural valley home may not. The yield helps, but it’s not the only reason to do this. The other reasons are diversification, currency, lifestyle, or perhaps the simple fact that you can operate it yourself.

The trip of a lifetime is happening in 70 days.

This October you can hear Terrie lay out the whole thing face to face, then go taste cask whisky in the Nagano mountains and stand in a thousand-year-old bonsai collection.

One week. 4 spots left. ​Claim your spot now →​

I find it hard to decide on a house by looking only at the house. Look at the town first.

Walk around one and you can feel whether the place still supports itself. If the shops are open, people are moving with somewhere to be, and the houses are maintained, a single empty house among them has room to come back.

Where the whole town has emptied and the functions holding it together have gone, fixing one house will not carry.

What I check is the axis the town was built around. Hot springs, a castle, a sport, or the functions of daily life kept compactly connected. The longer that axis has held as history, the harder the foundation is to move.

Something Japan built over centuries is still working as the mechanism of the town, and the value of one house comes down to whether it sits inside that. The projects running a village’s empty houses as a single inn work for the same reason.

Inbound tourism reached a record 42.7m visitors in 2025, against a government target of 60M by 2030. But those are national figures, and they may not apply to the town you’re buying in.

Yes, you can buy property site unseen. But it’s wise to feel the air of a town before you look at the property. What you pick up there becomes the standard you read the documents against.

Two rules decide whether your plan works. The rest is paperwork.

The first is how many nights you can operate, and that is set locally. The ​180-night annual cap​ under the Private Lodging Business Act has not changed, but in June 2026 the Japan Tourism Agency accepted that municipalities may write ordinances taking it to ​zero​.

Shibuya ​widened​ its restricted zones from July 1st, and ended the 180-night exemption for externally managed properties, leaving about 63 nights a year inside them.

Other wards have restricted weekday operation where no manager lives on site. And Osaka ​closed​ new special-zone minpaku applications in May 2026. A plan built on the national number will not survive the ward office.

The second is the renovation. Since April 2025 two-storey wooden houses ​no longer skip structural review,​ and replacing more than half of a wall, column, floor, beam, roof, or staircase requires a building confirmation application.

Re-roofing down to the sheathing and rebuilding a floor from the structural members are exactly what a kominka restoration involves. Interiors and finishes fall outside it. Whether the application is needed changes both the budget and the timeline.

Sannenzaka, Kyoto. Photo: Aporon999, CC BY-SA 3.0

The rest is filings. From October 5 2026, anyone registering as an owner ​supplies official proof of nationality​, foreign and Japanese alike. Forest land and larger plots carry their own reports, and land near defense facilities or on border islands needs notification before you sign.

Physical first.

The 30-year probability of a Nankai Trough megaquake was recently ​raised to about 80%​. And the September 2025 revision gives 60-90% or higher and 20-50% depending on method.

Anything predating the 1981 seismic code drops out first, which covers almost every kominka. Check the tsunami and landslide maps before the listings, because insurers decline designated zones outright and earthquake cover is ​capped​ at 30-50% of the fire policy and ¥50M on the building.

Wajima, February 2024. Photo: Hurohukidaikon, CC BY 4.0

Then there’s population.

Japan had 126 million people in 2020 and that’s projected to fall to 87 million by 2070.

In April 2024 a private council classified 744 of 1,729 municipalities, (43%!), as ​at risk of disappearing​. When a municipality shrinks, water rates rise, schools merge, road maintenance thins, and liquidity follows. A house is never more solvent than the town it stands in.

Financially, the weak yen lowered your entry price and a reversal takes it back, and as the Bank of Japan normalizes, required yields rise.

The countryside is exposed to disasters and demographics, the city to price and rates.

Two answers have come from opposite ends of the market.

​NOT A HOTEL​ works the new-build side, selling architect-designed homes in shares as small as 1/36. They buy back the nights you don’t use, and the exit handled for you.

Tea House,” after renovation. Photo: Akiya 2.0

​Akiya 2.0​ works the old-house side, and Terrie san’s answers are the more instructive, because he takes the constraints above one at a time.

“Renovating a traditional home is in fact a little cheaper or about the same price as a new building. Tthe timbers and foundations stay, and wood in a house 60 years or older sets very hard and is very durable.” — Terrie Lloyd

He actually buys entire villages rather than houses, though “you can seldom buy a cluster right from the beginning.”

It starts with the neighbors, showing them you will be responsible and friendly, and the opportunities follow. Terrie chose Mizuba, where more than half the homes are empty and, apart from two families, everyone is 85 or older, with a target of ten houses over the next five years. Through Akiya 2.0, Terrie plans to launch fractional investing later this year.

Scale is the point. Economies of scale arrive above six houses or so, and a rural build needs a project manager on site that one house cannot justify. The contractors are local, and “dangling the prospect of more builds is a great way to keep budgets in line.”

The difficulties are often mundane stuff, like finding a rural laundry that can handle oversize CalKing sheets.

Demand connects too. Japan Travel is about to promote self-guided tours on the ​Shimanami Kaido​, where 360,000 cyclists pass through Omishima each year.

Profitability needs beds for 24, four or five houses, beyond which his own travel company becomes a significant off-season user of his own clusters. His first attempt was Noto (stopped by the January 2024 earthquake) with the thinking carried over to Omishima.

The Shimanami Kaido. Photo: Fabimaru, CC BY-SA 4.0

His advice to investors is direct:

“If you want to own their own stuff, they need to have at least 3-4 properties in a given area to make money. My strong advice is that you pick locations where there is tourist traffic and not enough supply of good quality accommodation.” — Terrie Lloyd

If you do go in, three things before you shop.:

Be honest about whether you are buying for lifestyle or income.

Line up your cash first, because sellers move first come, first served.

And most importantly, build the team before the purchase. A bilingual judicial scrivener, the tax representative non-residents are required to have, and, for lodging, the local rules checked in advance (Kyoto is strictest and licenses do not transfer with the building.

Japanese real estate is not a market you buy just because it’s cheap. It is a market you buy when you can design the exit.

I live in Tokyo and meet both sides of this market every week. What began as one man’s romance in a valley is being turned into an asset class by operators who design exits.

The objection that foreign money is buying Japanese heritage cheaply is fair. But remember: a house that does not sell is demolished rather than preserved, and the houses that survive are the ones somebody keeps using.

I watch the same thing happen to bonsai gardens every year. And I find huge value in preservation.

That’s it for today.

I hope you enjoyed Part 3 in this series on investing in Japan. A big thanks to Terrie San for sharing some early numbers and lending his expertise.

If you are seriously considering this, talk to Terrie san who is an expert in this space.

I look forward to meeting everyone who is joining the ​Altea trip to Japan​ October 18-22!

Until next time, Hayato

This issue was written by Hayato Takahashi with help from Terrie Lloyd. It was edited by Stefan von Imhof

Alt Assets, Inc has no current holdings in any companies mentioned in this issue.

We exploring creating an SPV with Terrie Lloyd and Akiya 2.0 in late 2026.