Chief Strategy Officer / Japan
Four commitments, each with what it cost us
- Revenue, year to March 2026
- ¥486.0bn
- Condominiums, rental housing for the managed fund, reconstruction and renewal, detached housing and property management
- Return on equity, year to March 2026
- 8.9%
- Up from 6.1 per cent in the year to March 2022
- Rental homes held by the fund our subsidiary manages
- ¥238bn
- 5,900 homes recycled since 2019; Mikurane keeps a 15 per cent stake
Decisions & consequences
01Leaving ten suburban estates and moving the capital to three station districts
In April 2022, my first month as chief strategy officer, I reviewed our detached-housing land bank: 2,100 unsold lots across fourteen suburban estates in eleven municipalities in northern Kanto, Tohoku and Hokuriku, carried at ¥31.2 billion. Sales had averaged 160 lots a year since 2019, so the bank held about thirteen years of selling, in towns whose household numbers were projected to fall by a fifth or more by 2040. Four estates lay within walking distance of a station. Ten depended on cars and thinning bus routes.
Holding the land meant carrying costs, maintaining roads and parks we had promised, and an impairment that would arrive eventually. Selling meant recognising the loss at once, in the first year of reporting returns by business.
I recommended selling the 1,650 lots in the ten car-dependent estates over two years, keeping the four station estates and moving the capital into station districts of regional core cities. The head of our detached-housing business disagreed: the estates carried our name, employed 120 sales and maintenance staff and served families who had trusted us to finish what we started. Two outside directors asked whether an impairment of that size would undermine a plan in its first year. The board approved the exit in November 2022.
We recognised an ¥8.6 billion impairment in the year to March 2023. Lots were offered first to local builders and municipal housing programmes, then in blocks to a regional housebuilder. The 120 staff moved into our rental and renewal businesses, with no redundancies. We also committed ¥40 million a year for ten years to maintain parks and meeting halls in the estates we left.
By March 2025 the lots had sold for ¥16.3 billion, slightly above their written-down value. We invested ¥64 billion in station districts in Sendai, Kanazawa and Kumamoto: 1,180 homes with two clinics, a childcare centre and shops. The Kanazawa building was completed in January 2026 and is 94 per cent sold. Detached-housing revenue fell from ¥52 billion to ¥38 billion. Every land proposal now carries a household projection and a walking-distance map.
What I would do differently: residents of one estate learned of our exit from the local press before our letter reached them. I would tell existing residents first, and negotiate the maintenance commitment with their associations instead of announcing it to them.
Accepting half our usual margin to rebuild a 1971 estate with its owners
In 2012 the management association of Sasamegaoka Heights, five five-storey walk-up blocks with 312 homes built in 1971 in western Tokyo, asked Mikurane to study reconstruction. A seismic diagnosis had found the blocks below current standards, there were no lifts, and the median owner was 71. The site could hold a fourteen-storey building of 520 homes, leaving 208 to sell to new buyers to fund the rebuild.
The economics were thin. As construction costs rose in 2013 and 2014, our projected margin fell to 6.5 per cent against the 12 per cent our investment committee required, and owners faced an average contribution of ¥8.4 million.
The investment committee, chaired by our then executive vice-president, recommended withdrawing. I argued that we could not build a reconstruction business without a completed reference project, and that withdrawing after two years of meetings would end our credibility with other associations. I proposed accepting the lower margin, buying out at a fixed price any owner who did not want to return, and paying for temporary housing in the same district for owners over 75. Those measures added ¥2.3 billion of cost. The committee approved, on condition that I report the margin to the board quarterly.
We held 46 owner meetings and visited every household that missed them. A group of owners opposed the plan throughout, arguing that repair would be cheaper; they were right about the next ten years and wrong about the next thirty. In October 2015, 89 per cent of owners and of voting rights approved reconstruction, above the supermajority the law requires.
Seventy-four owners took the buyout, and nine who had voted against later sold to us by agreement; we never used the compulsory purchase route the law allows after a valid resolution. Construction ran from 2016 to March 2019, and 229 original households returned. Contributions averaged ¥9.1 million rather than ¥8.4 million, and our realised margin was 5.4 per cent. Sasamegaoka became the reference for six further reconstruction mandates and for a renewal service for associations whose buildings cannot be rebuilt.
What I would do differently: the owners' consultants were paid by Mikurane, which the opposing group fairly called a conflict. I would fund independent advisers through the association from the outset. I would also have shown owners a range for their contribution, not a single figure.
Buying 35 per cent of a Texas homebuilder rather than all of it
In 2017 our board wanted growth outside Japan's shrinking housing market, and the medium-term plan I was drafting aimed to earn 15 per cent of profit overseas by the year to March 2026. An intermediary offered us Pellridge Homes, a Dallas-based builder closing about 1,400 homes a year in the Dallas-Fort Worth and Houston areas, whose founders wanted to realise part of their investment. The whole company was valued at about US$300 million.
Buying it outright would have committed about a fifth of our net assets to a market where we had no operating experience, and where results depend on land options, local subcontractors and mortgage rates we could not influence.
Our then president favoured a full acquisition, so that Pellridge would be consolidated and its revenue would show in ours at once. I recommended 35 per cent, with two of seven board seats, full information rights and an option to reach 51 per cent after five years at a valuation to be agreed then. The founders preferred a minority investor because it left them in charge. The costs were plain: no consolidated revenue, less control and a higher price if we later bought more. The board approved the minority stake in November 2017.
We paid US$105 million in December 2017. I sat on Pellridge's board and its audit committee from 2018 to 2021, and we seconded two Mikurane engineers to learn its land-option and subcontracting practices rather than to impose ours. Pellridge entered the Raleigh area in 2020.
When mortgage rates rose in 2022, the founders cut land commitments early and Pellridge stayed profitable throughout. In 2023 we raised our stake to 45 per cent for US$88 million, at nearly three times our entry valuation. We chose 45 rather than the 51 per cent our option allowed, because keeping the founders' incentive mattered more than consolidation. Closings reached 2,600 in 2025, and our equity-method profit was ¥5.8 billion in the year to March 2026, about 17 per cent of our net profit. Mikurane now enters unfamiliar markets through minority partnerships with a defined route to more.
What I would do differently: I would fix a formula for the price of later stakes in the original agreement. Leaving it to negotiation at a strong point in the housing cycle cost us an estimated US$15 million.
Committing to three contractors before buying more land
By the year to March 2024 our tender prices for mid-rise condominiums in the Tokyo area were 31 per cent above those of three years earlier, and nine of our projects received no acceptable bid: either no contractor offered, or the lowest price exceeded budget by more than 15 per cent. Sites bought on earlier cost assumptions sat idle, and overtime limits for construction from April 2024 threatened longer programmes.
Our land bank had been planned for about 4,300 launches a year. Idle sites were costing ¥3.4 billion a year, and our land team was still buying at a pace that assumed construction would follow.
I proposed reversing the order: secure construction capacity first and buy only the land it could build. That meant four-year framework agreements with three mid-sized general contractors for about fifteen buildings a year, on standard designs priced against an open-book index for materials and labour, and cutting land purchases from about ¥120 billion a year to ¥75 billion. The head of our condominium business opposed it: launches would fall, and rivals would take central Tokyo sites we had always competed for. An outside director warned that relying on three contractors concentrated risk. The board adopted it in the 2024 to 2026 medium-term plan, capping any one contractor at 40 per cent of volume.
Our design team and the contractors agreed two structural grids and a catalogue of prefabricated bathroom, kitchen and balcony elements. Savings and overruns against the index are shared. We let four central Tokyo sites go to competitors in 2024.
In the year to March 2026 one tender went unawarded. Standard buildings took about 22 months against 26 for bespoke ones, and holding costs on idle sites fell to about ¥0.7 billion a year. Launches fell to about 3,400 homes, and gross margin on condominiums from 23.6 to 22.4 per cent. Land acquisition now begins from contractor slots, and the board sees which pipeline projects have one.
What I would do differently: in the first year we applied standard designs to every site, including a sloping one in Bunkyo where they did not fit, and lost a year redesigning. I would reserve about a fifth of volume for bespoke design from the start, and bring the contractors' site managers into the design catalogue earlier, because they found problems our designers had missed.