The BoJ's research agenda telegraphs the exit problem

A central bank's discussion paper series is a leading indicator of what its senior staff are worried about. Read that way, the Institute for Monetary and Economic Studies

' 2026 output is a coherent document. Across eight papers the IMES is triangulating a single question: how does an economy that spent a generation at the effective lower bound behave when nominal rates start to matter again — for housing, for banks, for the JGB market, and for labour. The dossier is thematic, not tactical, but the direction of travel is unambiguous.

The organising theme is the transmission of nominal rates through channels that atrophied under yield curve control. One paper revisits the real effects of nominal interest rates on housing and activity; another tackles the challenges of exiting prolonged low nominal rates directly. A third extends shadow short-term rate models — the family of term-structure tools that back out an implicit policy stance when the observed rate is stuck near zero — to Japan's ultra-low environment. Taken together, these are the papers a research department writes when its principals need a defensible framework for what normalisation actually does to the balance sheet of the household sector.

Plumbing, not policy

The market-microstructure cluster is the more revealing half. One paper studies dispersion

in interbank loan rates — the gap between what different banks pay for overnight funding, which widens when the policy rate stops being a hard anchor. Another applies interpretable machine learning to the determinants of JGB (Japanese government bond) liquidity. A third, 'Distance and Loan Pricing Revisited', uses structural estimation to decompose how lender–borrower distance feeds into pricing through monitoring costs, information frictions, and market power. These are the papers you commission when you suspect the plumbing has changed shape while the rate was pinned, and you want to know what breaks first when it moves.

The standard disclaimer matters here more than usual. IMES output is not policy signalling in the way a Governor's speech is, and the dossier offers no quantified forecasts on the timing of the next BoJ move — readers looking for a probability on the next hike will not find one in this material. What it offers instead is the intellectual scaffolding the Bank will lean on when it does move: shadow-rate diagnostics for measuring the current stance, structural loan-pricing models for anticipating credit-channel effects, and machine-learning liquidity indicators for monitoring the JGB market through the transition.

You do not commission eight papers on the mechanics of exit unless exit is on the agenda.

The labour papers — one on computerisation's effect on occupational schooling, one on employment protection and job tenure — sit slightly apart, but fit the same frame. A normalising rate environment revalues human capital and reallocates workers; a research shop preparing its principals for that shift would want the labour-market fluidity question answered before, not after, the policy pivot. The 2026 slate is not a forecast. It is a reading list for the exit.