Sell residence rights by the apartment, not by the premium
Chinese life insurers attach rights to live in their retirement communities to large savings policies. Issued without a count of the apartments behind them, those rights become promises that cannot be kept. They should be capped by capacity and reported to the board as commitments.
Over the past decade many Chinese life insurers have moved into elderly care. The most visible form is the insurer-owned retirement community: apartments for independent living, assisted living and nursing care on one site, usually near a large city and close to hospitals. The commercial logic joins two things a saver in their fifties wants, an income in retirement and a credible place to receive care. The insurer contributes long-term capital suited to long-lived property, and the community makes its savings products easier to sell.
The two are usually joined by a residence right. A customer who buys a policy above a certain premium receives a letter giving priority to apply for a place in the insurer's community after reaching a qualifying age. At Jiangheng Life we began issuing such letters for Songhe Court, our community in Wuhan, in 2020, two years before it opened. My argument is that residence rights should be sold by the apartment and not by the premium: capped at the number an insurer can honour, recorded in a register and reported to the board as a commitment, rather than counted only as a sales aid.
A right with a waiting list
The difficulty is plain once drawn. A community has a fixed number of places. Letters are issued at the pace of sales, which for an attractive community can be fast. Holders apply at a pace set by their age and health, slow at first and then rising. If the letters issued exceed the places available when holders apply, someone waits. If the wait is long enough, the letter that helped sell a policy worth several million yuan becomes a source of complaints, and possibly of claims that the customer was misled.
Most insurers understand this in principle. In practice the incentives run the other way. A letter costs nothing to issue, it helps close large sales, and its cost arrives fifteen or twenty years later, long after the sales plan, the agent's commission and perhaps the executives who approved it have gone. When I first modelled Songhe's position in 2023, we had issued 4,950 letters for 780 apartments, and letter-linked policies had produced RMB 6.4 billion of premium the year before.
A ratio of six letters to each apartment is not alarming in itself. Many holders will never apply: some will prefer to stay at home, some will move nearer their children, and some will die before they need care. Apartments also turn over as residents move into nursing care or die. The question is not the ratio but the timing.
Counting apartments across twenty years
Our model was not sophisticated, but it was honest about what it did not know. For each letter holder we projected the probability of applying in each future year, using age, sex, the policy's design and the early experience of communities opened by other insurers. Against that we set the apartments expected to become free, using initial occupancy and expected turnover. Our chief actuary's team challenged the mortality and morbidity assumptions, and my first proposal changed as a result.
The model showed applications exceeding available places from about 2034, with waits of several years by 2040 if letters continued at the 2023 pace. We could not know that this would happen. We could know that the outcome depended on sales decisions being taken that year.
I recommended a cap of 6,200 letters for Songhe, the number at which projected waiting stayed within a year in our central case. Further letters would be issued only for partner communities where apartments had been contracted. The cap was reached in February 2025. Letter-linked premium fell to RMB 3.8 billion in 2025, about RMB 3.1 billion below plan across two years, and some of our agents' wealthiest clients moved to insurers that had not capped.
I will not pretend that capacity discipline is free. But consider the alternative. An insurer that issues twelve or fifteen letters for every apartment, without a second community funded and under construction, is selling a promise whose fulfilment depends on capital decisions its board has not yet taken. That is a liability, even if no accounting standard yet records it as one.
What a residence letter should say
A residence right will never be as precise as a savings policy. It depends on the holder's health, on the community's admission rules and on choices made decades later. That is an argument for more disclosure, not less.
I would want every letter to name the community or communities it covers, the number of letters issued for that community at the date of issue and the insurer's projected waiting time for a holder applying at the qualifying age. I would want insurers to tell existing holders when those projections change, as we did in 2024. And I would want a clear statement of what happens if the holder cannot be accommodated: whether the right transfers to another community, whether any associated fee is returned, and over what period.
Inside the insurer, the letters belong in a register reviewed by the board's risk committee, as ours has been every quarter since September 2023. The register should show letters issued, projected applications by year, places available and the gap, community by community. It should be read beside the capital plan, because the honest remedy for a projected shortage is often another community, and that is a decision about capital.
Partner communities help, and we now have contracts with eleven operators in eight cities for 1,350 reserved apartments. They are not a substitute for counting. A reservation contract lasts for a fixed term, and the operator's own residents compete for the same places. We apply the same projection to partner capacity as to our own and stop issuing letters for a city when the projection says so.
China's population is ageing quickly, and the gradual rise in the statutory retirement age that began in 2025 does not change the arithmetic of care for people now in their sixties. Insurers have a real contribution to make with long-term capital, health insurance and the operating experience they are building. They will make it more credibly if the promises they sell are counted against the apartments they can provide. A residence right is valuable to a customer precisely because places are scarce. An insurer that forgets this will be reminded when its letter holders arrive.