The lobster is sometimes said to be immortal. It shows almost no sign of senescence: it does not weaken or lose fertility with age, and its cells keep dividing as though the years did not count (Smithsonian). But not ageing is not the same as not dying. The lobster dies. It dies not of old age but of the very mechanism of its growth. A lobster can grow only by shedding its whole shell in a moult, and each moult costs exponentially more energy as the animal gets larger. Between 10% and 15% of older lobsters die each year from a failed moult, and in time they stop moulting at all to conserve energy (Smithsonian, citing Maine’s Department of Marine Resources). The old shell then wears, bacteria get in, and scar tissue fixes the body to the shell. Trapped in the shell it will not shed, the animal rots. It dies because it stopped shedding, not because its time was up.
Markets price Japan much as they price the immortal lobster: large, old, unkillable. So foreign money buys the reform of Japanese firms, the end of deflation, the turn to active fiscal policy. The government of Sanae Takaichi, formed in 2025, is sold as exactly that moult: a break with austerity, a new fiscal shape. Yet look for a hand on the fiscal and household shell, and there is none.
What is new is not the behaviour. It is the price tag. Taking quietly from households through fiscal drag and social-insurance contributions while avoiding permanent tax cuts is not Ms Takaichi’s invention; it is the base metal of Liberal Democratic Party fiscal policy going back decades. What the market buys as a brand of active, anti-austerity policy is, in behaviour, the same regime as before. What changed is the constraint around that behaviour. In the zero-rate years the take-and-keep game had no visible cost: the real rate was negative, the Bank of Japan absorbed the bonds, and debt service could be ignored. No longer. The policy rate is 0.75%, the ten-year yield touched 2.125% in January, its highest since February 1999 and so for almost 27 years (Nikkei), and debt service reaches ¥31.3tn in the 2026 budget and, on the finance ministry’s own projection, ¥41.3tn by 2029, overtaking the social-security line of the national budget for the first time (Nikkei). The same refusal to moult now carries a price it did not before.
The clearest place that price shows is the yen. Late on 30 April 2026 the yen jumped to around ¥155, snapping back from the high ¥160s on what looked like intervention to buy yen, signalled by the finance minister, Satsuki Katayama, and the vice-minister for international affairs, Atsushi Mimura (Nomura). The intervention worked. Yet by 1 July the yen was back at ¥162.80, its weakest since December 1986 and so in nearly 40 years (Nikkei).
The level itself is largely an American story. A re-acceleration of American inflation and expectations that the Fed will tighten are pulling the dollar up, as Nikkei itself reports (Nikkei). What sets the level is the rate gap, not Japanese fiscal policy.
The reason intervention will not hold, though, sits on the Japanese side. Household net buying of foreign assets through the new NISA scheme is put at ¥0.7tn to ¥3.9tn a year, enough to push the dollar-yen ¥1 to ¥6 higher through 2027 (Japan Research Institute). It is not speculation reacting to the rate gap but real-money yen selling built into the scheme, and it continues even as the gap narrows. Households that cannot earn a real return at home are going abroad. The structure of the balance of payments itself tilts towards a weaker yen, as Sumitomo Mitsui DS Asset Management also notes (SMDAM). So April’s intervention returned the yen to ¥155, and two months later it had been pushed back to ¥162. Speculative net yen shorts stood at around 124,000 contracts as of 7 July, above their level before April’s intervention (CFTC). Reports of an online call between Ms Katayama and Scott Bessent, the American treasury secretary, sharpened the market’s wariness of further intervention (SMDAM). What that exchange signifies I have set out separately (The Silent Lever).
I wrote before about a squeeze that caught yen shorts (an earlier piece). That was a cyclical story: crowded shorts, one trigger, a run higher. This is a structural one. Even when a cyclical squeeze fires, the regime of tax, social insurance and real rates, the machine that keeps taking from households and denies them a real return, pulls the yen back to soft. The level is set on the American side; the reluctance to recover and the downward stickiness are set on the Japanese side. Intervention works on the cycle. It does not work on the structure.
If the yen is soft by structure, what of the purchasing power of the households that live in it? The surface numbers look bright. Real wages rose 1.9% year on year in April 2026, a fourth straight monthly gain (Ministry of Health, Labour and Welfare); 2025 had run 12 months in the red, so this looks like a turn, and the May flash reading of 1.4% extended it to five (MHLW). But break the turn down and it is made of policy. April’s consumer price index, excluding imputed rent, rose 1.5%. Energy pulled it down: petrol fell 9.7% on government subsidy and the end of the old provisional tax rate, and education fell 6.1% as high-school tuition support widened (Nikkei, on the ministry’s CPI). Nominal pay did not so much outrun prices as the government held part of prices down. Dai-ichi Life Research flags a downside risk to real wages from the autumn if oil climbs again (Dai-ichi Life). Subsidies are a budget measure and the tax-rate cut is a one-off level shift; both roll off next year. A real-wage gain built by policy vanishes when the policy stops. This is not a moult. It is polishing the outside of the shell to make it shine.
The sharpest objection runs like this: if nominal pay rises feed on themselves, and real wages go on rising without policy support, that is the shell coming off. It is a fair point. But the evidence for self-propulsion is, for now, hidden beneath the price suppression that policy has manufactured. The day it shows, this piece is wrong.
The permanent tools are being avoided. The zero rate on food is for two years, food only, and designed to stay within a no-deficit-bond envelope. Ms Takaichi herself says the cost can be met without deficit bonds, from non-tax revenue and a review of tax expenditures (Jiji). The rise in the basic deduction to ¥1.78m leaves residential tax and the tax brackets untouched; Dai-ichi Life called it insufficient against bracket creep (Dai-ichi Life). The reversible support is thick; the permanent cut is thin, slow and hedged. The taking is structural; the giving back is a feint.
Against a giving-back that is a feint, the taking rises on its own. Social-security benefits swelled from ¥78.4tn in 2000 to ¥140.7tn in 2025, and the insured’s contributions from ¥29.9tn to ¥43.5tn, roughly half as much again (Dai-ichi Life). The aggregate is shaped by headcount and wages, but the rate has risen too: the long-term care contribution rate has trebled from 0.6% at the scheme’s creation to 1.82% in 2023, a record (Kyokai Kenpo). The burden of tax and social insurance as a share of workplace income has risen over the long run as well (Dai-ichi Life). No matter how high nominal pay climbs, if the slice taken from it keeps rising, real disposable income is pared.
Why never give it back for good? Because a transparent tax cut carries one of two costs: accept a permanent fall in revenue, or fund the cut with bonds and have the market question the public finances. Either forces a reckoning on those who gain now. What is protected is the older, asset-holding cohort: pensions track prices, price suppression holds down the cost of living, a negative real rate slows the erosion of savings and supports asset prices. The burden falls on the young and the working-age, with fiscal drag and imported inflation landing on them. Whether this allocation was chosen with intent I will not assert. But the result does not move: permanent tax cuts and a rising real rate both force the reckoning the median older voter dislikes, and reversible support with a repressed rate defers it. The generational bargain is the floor beneath the standstill.
That floor does not sit under the voters alone. It sits under the policymakers too. Tax expenditures, subsidies and funds, once created, do not die: each acquires a bureau, a budget line and a beneficiary, and abolishing any of them cuts into some bureau’s turf. No bureau shrinks itself.
The first pass of Japan’s DOGE, its copy of America’s government-efficiency drive, laid this bare. The office for reviewing tax expenditures and subsidies, set up within the Cabinet Secretariat in November 2025, is staffed by around 30 people drawn mainly from the finance ministry (JBpress). Where the American original was led by an outside businessman and cut into headcount and whole agencies, the Japanese version ran as a self-inspection, each ministry vetting its own tax breaks and subsidies. The result: of about 120 items published by 13 ministries and agencies by 8 July 2026, exactly one was marked for clear abolition (NRI). Ms Katayama, the minister in charge, allows that the outcome does not satisfy her (NRI). An organisation created on the admission that reform needs a push from outside, then run by insiders inspecting themselves, had already given its answer.
The generational bargain and an administrative machine that cannot shrink itself form a double floor under the standstill. The structure by which the same finance ministry designs both revenue and spending belongs in a separate piece.
There is a counter-argument. Withholding tax cuts is fiscal discipline, and discipline should support the currency. But this government secures revenue while letting spending swell. What it takes it does not return; it recycles into growth investment and crisis-management investment. The fiscal impulse, the net push or drag fiscal policy puts on activity, is not tightening. Were the discipline real, the Bank of Japan would let the real rate rise. It does not: even as it keeps raising the policy rate, the bank still describes the real rate as very low (Bank of Japan, January 2026 Outlook). The reluctance is not discipline; it is a transfer the market is not allowed to price. The discipline that would support the yen and this reluctance are two different things.
The opposite pole offers no exit either. The loudest criticism of the finance ministry comes from the reflationists. They call a weak yen an engine of growth and argue that a weaker yen lifts tax revenue and eases the public finances. Part of that is true: a weak yen does lift nominal revenue and corporate profits. But it does so on top of an erosion of household real income. That a weak yen eases the finances is precisely the mechanism this piece is about: the release valve is open, so the government need not touch the skeleton. And many of this school sit on the boards and committees that advocate deregulation and wider foreign-worker employment. A cheaper currency and a larger labour supply are both prescriptions that stand with capital, with the burden landing on households and the working-age. The critics of the finance ministry stand not outside the system but inside the side that protects firms and assets. The opposite pole was not an exit; it was an extension of the regime.
Seen from the outside, nobody chose decline. Each actor optimises to avoid conflict, and the sum of that is simply the path of least resistance: the government protects the brand, the Bank of Japan holds the rate down, the older majority protects its share. No one is at fault. That is why it does not stop. And the decline does not show in the index. It does not show in the Nikkei. The reform and governance gains in Japanese firms that foreign money is buying are real, and those who bet on “Japan is finished” and sold have long been burned. The decline shows where the system is willing to give in order to avoid conflict: in the yen, and in the real living standards of households. A reformed Japan and a Japan left soft; those two Japans are the subject of an earlier piece (Two Japans).
The implication for investors can be put as a direction. No tickers. The focus is domestic. As long as reversible support rolls off, social-insurance contributions pile up and the real rate stays negative, household real purchasing power will not rise as far as the headline pay-round reports suggest. The view that a tax cut brings consumption back misses how reversible the support is. Sectors that lean on domestic household spending, retail, restaurants and staples, need to price this structural headwind. The export and yen-beneficiary side is a well-worn trade, no need to rehearse it here. The specific sector calls belong in a separate piece. This is about macro structure, and it is not investment advice.
Back, finally, to the shell. The lobster does not die by fate. It dies only when it stops moulting. So this piece can be proven wrong. If Japan’s DOGE, on its next pass and unlike its first, actually cuts tax expenditures, if the Bank of Japan lets the real rate rise and if the brackets and social-insurance contributions are truly reformed, then the shell is coming off and what is written here is mistaken. Until one of those happens, the impressive old body sits in a shell it will not shed. And the yen carries the strain.