The strategy work behind fewer, longer commitments

Japan's population has been shrinking for over a decade, but housing demand turns on households, and households decline later and less evenly than the headcount. I built Mikurane's demand model on household projections by municipality and type: elderly people living alone, couples whose children have left, young families. We then overlay walking distance to a station, a hospital and a supermarket. Roughly one dwelling in seven in Japan now stands vacant, and those vacancies cluster away from stations, so a municipal average tells a board very little. The model drove our withdrawal from ten suburban estates in 2022 and the choice of three regional station districts for new capital. It also changed how we phase projects: we sell the homes nearest the station first and hold back outer phases until absorption is proven. Many municipalities intend to concentrate development around their stations; I describe that intention in general terms and look for sites where their plans and ours point the same way.

Japan has a large stock of condominiums built before the earthquake-resistance standard introduced in 1981, and many of their owners are older than the buildings. Rebuilding one needs a resolution by a large supermajority of owners under the unit-ownership law, finance for owners who cannot afford a contribution, and somewhere to live during construction. The traditional model paid for all of it with surplus floor area sold to new buyers, which works only where planning rules allow a much larger building and land values are high. I founded our reconstruction business in 2012 and have since steered it towards renewal: long-term repair plans priced at current costs, stepped increases in repair reserves, seismic retrofitting and lifts for walk-up blocks. Of the 73 management associations we advised in the year to March 2026, six are pursuing reconstruction and the rest renewal. The segment earned ¥41 billion of revenue, and its margins depend on our willingness to tell owners which buildings cannot be rebuilt.

Our tender prices for mid-rise condominiums in the Tokyo area rose 31 per cent between the years to March 2021 and March 2024, and in the second of those years nine of our projects received no acceptable bid. The overtime limits that began to apply to construction work in April 2024 lengthened programmes further. I responded by treating contractor capacity as the first resource to secure. Since 2024 we have held four-year framework agreements with three mid-sized general contractors for about fifteen buildings a year, built on two standard structural grids with prefabricated bathroom, kitchen and balcony elements and priced against an open-book index for materials and labour. We cut land purchases from about ¥120 billion a year to ¥75 billion to fit that capacity. The arrangement cost us four central Tokyo sites and about a point of gross margin. In return, only one tender went unawarded in the year to March 2026, and standard buildings are finished in about 22 months against 26 for bespoke ones.

A developer that keeps every rental building it completes becomes a landlord by accumulation, with equity earning less than it costs. In 2019 I set up the Mikurane Rental Residence Fund, with a life insurer and a corporate pension fund as co-investors and our subsidiary Mikurane Asset Management as manager. We sold it 2,400 rental homes for ¥96 billion in the first tranche and keep a 15 per cent stake. Three further tranches have taken it to 5,900 homes valued at ¥238 billion. The structure contains an obvious conflict, since we sell buildings to a fund our own subsidiary manages, so every purchase from Mikurane needs two independent appraisals, a price no higher than the lower of them and approval by an investor advisory committee on which we have no vote. Recycling has released about ¥170 billion of capital for development and brings fee income of ¥3.1 billion a year. It also means giving up rental growth, and I report that cost to the board every year.

Visitor numbers to Japan recovered strongly after the pandemic, and the demand concentrates in a few cities and districts. In 2018 our executive committee considered five hotels in Kyoto and Osaka. I recommended two serviced-apartment buildings in Osaka instead, designed so that floors could move between monthly lettings and short stays, subject to the licences each use requires. When travel stopped in 2020 both buildings switched to monthly residents and occupancy held at 88 per cent; a dedicated hotel would have had no such fallback. Since 2021 our capital-allocation plan has capped hospitality-related assets at 5 per cent of invested capital. Any proposal in a tourist district must also show demand from residents and workers, because rents and land prices driven by visitors can make a neighbourhood harder for the people who live there, and the patience of residents and municipalities with that has limits. I would rather earn less from the boom than own buildings that depend on it.

In 2023 the Tokyo Stock Exchange asked companies on its Prime and Standard markets to show management that is conscious of their cost of capital and share price. Our price-to-book ratio was then 0.68. I led the analysis we published: an estimated cost of equity of about 8 per cent, return on equity by business and the capital each business employs. It showed that detached housing in declining regions and our cross-shareholdings were holding capital without earning its cost. Our 2024 to 2026 medium-term plan set a 9 per cent return on equity for the year to March 2027, reduced cross-shareholdings from 14 per cent of net assets to 6 per cent and moved capital towards renewal, the fund and our US partner. Return on equity reached 8.9 per cent in the year to March 2026 and the price-to-book ratio 0.94. I am careful about what the plan cannot show: much of our return still depends on condominium prices in the Tokyo area, which no plan controls.