Macro · Research note
Central Bank Dilemmas: Navigating Interest Rates in a High-Inflation Decade
The ECB raised rates in June while forecasting inflation back at target only in 2028 and growth of 0.8%. That single decision contains the whole argument about what central banks are now optimising for.
On 11 June 2026 the European Central Bank raised its three key rates by 25 basis points, taking the deposit facility to 2.25% effective 17 June. It did so while publishing projections showing euro area growth of 0.8% for the year and inflation returning to 2% only in 2028. A central bank that raises rates into a slowdown it has itself forecast is telling you what it values: it is not optimising for output, and it is not optimising for the fastest possible return to target. It is defending the belief that it will eventually get there.
This piece argues that credibility, not inflation, is the scarce resource in 2026, and that the binding constraint on how much credibility a central bank can buy is fiscal rather than economic.
Why this shock cannot be looked through
Textbook guidance on supply shocks is to look through them, and the guidance is usually right. A tariff, a harvest failure or an oil spike raises the price level once and then drops out of the annual comparison; tightening into that is a policy error, since it destroys demand to fight an inflation that was leaving anyway.
The 2026 shock resists this treatment for one reason. Part of the supply loss is permanent. Iranian strikes on Qatar’s Ras Laffan complex on 18 March wrote off two liquefaction trains carrying 12.8 million tonnes per annum, about 17% of Qatari capacity, with a recovery window of three to five years. That is not a disruption which resolves itself over four quarters, and the damage is not confined to gas. Furthermore the Gulf supplies roughly half of globally traded urea and a large share of sulfur, and fertiliser prices reach food prices with a lag of six to nine months. A shock with a permanent component and a long lag produces exactly the inflation profile that central banks are told to ignore and cannot afford to.
Figure 1 shows why the two major central banks reached different conclusions. US headline CPI ran at 4.2% in May against core of 2.9%, a gap of 1.3 points that is mostly energy passing through; euro area headline was 3.2%. The gap between the two lines is the part of the problem that rate policy cannot reach directly, and it was wider in the United States than in Europe. The ECB nonetheless hiked and the Federal Reserve did not, which means the difference in behaviour was not driven by the data.
The ECB: buying credibility at a known price
The Governing Council was explicit about the trade: President Christine Lagarde noted that a look-through strategy had been considered and rejected, on the grounds that indirect and second-round effects were large enough to require action.
Figure 2 is the honest version of the decision. The ECB does not expect headline inflation at target until 2028; core inflation stays at 2.5% through 2027, above target for two more years, and growth is 0.8% in 2026. The bank raised rates anyway, and it downgraded growth in the same publication. Therefore the hike cannot be read as an attempt to hit the target sooner, because the bank’s own forecast says it will not.
The hike is better read as a purchase: the ECB has missed its target repeatedly since 2021, and each miss costs something. Raising into weakness is expensive in output and cheap in credibility, and the Council judged that it had more output to spare than credibility. Whether that judgement is correct depends entirely on whether expectations were actually at risk, which is unobservable at the time of the decision.
The Federal Reserve: a different instrument, not a different view
Coverage of the Fed in 2026 has sorted policymakers into hawks and doves. That framing obscures what Kevin Warsh, sworn in as the seventeenth Chair on 22 May, has actually proposed.
Warsh’s stated approach is a barbell. It pairs relative patience on the policy rate with aggressive reduction of the balance sheet, which stood near $6.7 trillion and about 21% of GDP by mid-2026. The reasoning is that quantitative tightening removes accommodation through the term premium rather than the short rate, and that supply-side improvements from artificial intelligence and deregulation are disinflationary. Hence a Fed that leaves the funds rate alone is not necessarily easing, and describing the position as simply hawkish or dovish misses the mechanism entirely.
| Central bank | Action, June to July 2026 | Instrument | What it is protecting |
|---|---|---|---|
| European Central Bank | +25bp on 11 June, deposit facility to 2.25% | Policy rate | Credibility, at a cost of 0.8% growth |
| Federal Reserve | Held at 3.50 to 3.75% on 29 July, committee divided | Balance sheet, ~$6.7tn | Optionality, while QT does the tightening |
| Bank of Canada | Held at 2.25%, fifth consecutive meeting | Neither | Room to move once the shock resolves |
Sources: ECB monetary policy decision, 11 June 2026; FOMC statement, 29 July 2026; Bank of Canada, June 2026.
The practical result is the divergence in Figure 3. Three central banks facing overlapping shocks chose three different instruments: the ECB moved the price of money, and the Fed is moving its quantity. The Bank of Canada held at 2.25% for a fifth consecutive meeting, with Governor Tiff Macklem describing the choice as one between crushing a slowing economy and tolerating persistent inflation.
There is a risk in the barbell that deserves stating: draining reserves raises borrowing costs through a channel the Fed controls less precisely than the funds rate, and it does so at the same time that Gulf sovereign wealth funds are withdrawing capital from US markets. The Public Investment Fund cut its international allocation target from a peak of 30% to 20% under its 2026 to 2030 strategy, redirecting a portfolio of about $925 billion toward domestic use. Two large sources of marginal demand for US assets are therefore contracting at once, and that combination is not something quantitative tightening was designed around.
The ceiling: fiscal dominance
Everything above assumes a central bank can raise rates as far as its mandate requires; sovereign debt levels increasingly determine how far that is.
US general government debt is projected to pass 120% of GDP by 2027 with a 2026 deficit near 7.9% of GDP. In a balance sheet of that size, the interest bill responds to policy rates with a lag but with force.
One caveat on Figure 4 is important, because the underlying figure is routinely misused. The often-quoted estimate that a one percentage point rise in rates costs the Treasury close to $400 billion a year is a long-run number, not a first-year one. Only maturing debt reprices, so the cost accumulates as the existing stock rolls over across the projection window. Quoting it as an immediate expense overstates the near-term squeeze and, more importantly, hides the reason fiscal dominance is dangerous: the pressure builds slowly enough to be ignored until it is structural.
The mechanism is a conflict between two arms of the state: fighting inflation requires high rates, and high rates raise the sovereign’s financing cost, which raises pressure on the central bank to cut or to buy debt. If markets conclude that the central bank has begun serving the financing need rather than the inflation target, the institution’s value collapses.
That value can be measured. IMF work finds that a 0.1 point increase on a zero-to-one independence index is associated with a reduction of roughly 0.6 to 0.7 percentage points in five year local currency sovereign yields. Independence is therefore not only a governance principle. It is a line item in the budget, and a government that pressures its central bank is charging itself for the privilege.
Expectation management, and what the evidence supports
When rate increases are constrained by debt, communication carries more of the load; the evidence here is real but smaller than the attention it receives.
Event studies of FOMC announcements suggest that a 10 basis point hawkish surprise in the expected policy path lowers market-based inflation expectations by roughly 4.5 basis points, and that unscripted remarks in the press conference can move them further than the prepared statement. Research on firm-level expectations finds that unanchoring is driven by disagreement between economic actors and by a belief that the central bank has temporarily set its target aside.
Note the size of these effects: a 10 basis point surprise buying 4.5 basis points of expectations is a real but modest transmission ratio. Communication supplements a rate policy that is working; it does not substitute for one that is blocked. Any argument that a constrained central bank can talk its way to price stability is asking more of the evidence than the evidence supports.
Where this framework falls short
The credibility argument is difficult to falsify, which is its main weakness. If inflation returns to target, the hike is credited; if it does not, the hike can be defended as having prevented something worse. A framework that survives both outcomes is not being tested by either. That weakness applies to the argument made here as much as to the ECB’s own defence of the decision.
Furthermore, the fiscal dominance argument treats a ceiling as though it were already binding, and it is not yet: the Fed held rates in July at 3.50 to 3.75%, which is a level it chose rather than one it was forced into, and no major central bank has been visibly overruled by its treasury. The honest statement is that the constraint exists and is tightening, not that it has already been reached.
Conclusion
Central banks in 2026 are not managing a standard cycle: they face a supply shock with a permanent component, an inflation profile with long lags in food and gas, and a debt stock large enough to price their own independence.
The ECB’s June decision is the clearest available statement of the resulting priority. Faced with 0.8% growth and a target it does not expect to reach until 2028, it raised rates. That is a bank spending output to protect the belief that it will eventually do its job; the Federal Reserve is buying the same thing through the balance sheet instead. Both are working within a fiscal ceiling that is closer than it was five years ago, and neither has yet had to find out exactly where it sits.
Update: 10 August 2026
Two things have happened since this was written on 12 June, and one of them undermines a forecast the piece leans on.
The Federal Reserve held its target range at 3.50 to 3.75% on 29 July, over a divided committee. Warsh declined to move on rates while stating that the Committee would act if inflation required it, and the statement attributed part of the overshoot to sector-specific supply shocks including energy. This is consistent with the barbell described above rather than with a conventional hawkish hold, and the division within the Committee suggests the approach is not settled internally.
The energy assumption underlying every June forecast has broken. Iranian attacks on shipping resumed in July, the United States reimposed its naval blockade on 12 July with a 20% cargo charge, and Brent gained more than 9% in a single session before trading in a wide band through early August. The ECB’s projection of 3.0% average inflation in 2026 was built on a path for energy prices that assumed the June memorandum would hold. It did not. Hence the September projection round is the one worth reading, since it will be the first to price a blockade that has been imposed, lifted and imposed again within six months.