The graph says Alan Taylor and monetary policy are two hops apart; the more useful question is what a historian of credit cycles would make of a 2.24% breakeven sitting under 4.25% realized inflation.
Which of these is the anomaly: a Bank of England policymaker linked in a knowledge graph to the policy he helps set, or a 10-year breakeven of 2.24% sitting beneath realized inflation of 4.25%?
Only one is a discovery. The graph says Alan Taylor is MEMBER_OF the Bank of England and MEMBER_OF the Monetary Policy Committee, both at 0.95 confidence, and that Monetary Policy is itself MEMBER_OF the Bank of England. That is a rate-setter joined to rate-setting. The mispricing, if one exists, sits somewhere else.
The short version
- The graph's "finding" is a definition: a policymaker connected by MEMBER_OF to the policy he helps set. The Monetary Policy node's four recent-claim records carry null text and null claim type, and Taylor carries no claims at all in thirty days.
- A 2.24% close on the US 10-year breakeven on 10 July sits against realized CPI of 4.25% YoY (May vintage). A breakeven is long-horizon inflation compensation, expectations plus risk and liquidity premia, not a forecast that today's print falls two points.
- Oil is one proposed contributor to that gap, and its contribution is decaying by arithmetic rather than by price: the 30-day WTI window slipped from -24.0% to -21.5% while live WTI at 71.41 sat ten cents below the prior print.
- The 1998 parallel is about a rolling energy base, not about a policy error; that year's cuts also answered a financial-stability shock.
- Every figure here is American while Taylor's seat is British, so this is an argument about a mechanism, not about a vote, and June CPI on Tuesday 14 July begins to settle it.
Who is Alan Taylor, and why does he sit on the Bank of England's Monetary Policy Committee?
The graph records him MEMBER_OF the Bank of England and MEMBER_OF the Monetary Policy Committee, both at 0.95 confidence, with further links to the Banque de France (ALLIED_WITH, TRADES) and the Paris School of Economics (TRADES). Those labels establish edges, not affiliations, and this piece treats them as nothing more.
Get analysis like this delivered daily. No account required.
This analysis was produced by the Convex Research Desk from live economic data and is for informational purposes only. It does not constitute financial, investment, or legal advice. See our editorial standards and terms of service.
Convex combines institutional macro research with AI-powered signal generation. Access live trading signals, portfolio analytics, and strategy backtests.
Zero claims for him appear in the feed across thirty days, a gap in the feed rather than a silence on his part. The public record is loud. At the meeting ending 17 June 2026 the MPC voted 7-2 to hold Bank Rate at 3.75%, with Megan Greene and Huw Pill dissenting in favour of a quarter-point rise to 4.00%. Note the direction. The dissent was hawkish, which is worth registering in a piece about whether disinflation is durable. Taylor voted with the majority to hold and spoke at Barclays and CEPR that same month, having told an audience in March that he had by then cast twelve votes.
The seat is real and the man is on the record. Neither fact is the story.
What the Bank of England link between Taylor and monetary policy actually shows
A tautology dressed as a signal. The detector traced MEMBER_OF from Taylor to the Bank of England, MEMBER_OF from Monetary Policy to the same node, and reported a two-hop link between a policymaker and the policy he sets. That is what a rate-setter is.
The nodes are empty as well. Four recent-claim records attach to Monetary Policy between 30 June and 3 July, every one carrying null text and null claim type, and Taylor's claim list for the last thirty days is blank. Whatever that edge weighs, it carries no information about prices, and a market read stacked on top of it would be an author's macro view wearing a graph as a costume.
What survives is a question the graph cannot answer but the tape can: how much of the disinflation compensated in US breakevens depends on a thirty-day window that is about to change shape.
A 2.24% breakeven against 4.25% realized inflation: the numbers behind the disinflation trade
Start with the concession that matters. Every figure below is American: the 10-year breakeven, DFII10, DGS10, realized CPI, GDPNow, WTI. Taylor's seat is British, and a US breakeven binds nothing about what a BoE external member confronts. Read what follows as a mechanism, never as a vote.
On 10 July the 10-year breakeven closed at 2.24%, unchanged from the prior state, against realized CPI of 4.25% YoY on the May vintage. That is not a forecast that CPI falls two points. A ten-year breakeven is inflation compensation over a decade, and it embeds a term premium and whatever liquidity discountTIPS carry. The gap is still wide enough to ask what sits inside it.
Oil is one candidate, and nothing here decomposes the breakeven to prove it is the only one. What can be shown is that oil's statistical contribution is shrinking without oil moving: the 30-day WTI window now reads -21.5% (88.62 on 12 June to 69.60 on 6 July) against -24.0% measured a day earlier. No barrel rose to produce that. Live WTI at 71.41 sits ten cents below the 71.51 the prior state recorded, so mid-June's higher prints aged out of the base and the disinflation statistic shrank by division.
The rates side is more legible. Across thirty days the nominal 10-year rose 6bp (4.48 to 4.54), the real yield rose 14bp (2.17 to 2.31), and the breakeven fell 7bp. Real yields more than account for the entire nominal selloff. The move is being made in the real leg, and DFII10 at 2.31% is the closest structural trip-wire in the book.
The other side of the argument holds together. Claims at 215,000 and falling 6.5% over thirty days, GDPNow at 1.3%, NFCI easing to -0.515, HY OAS at 2.70, and live de-escalation optionality in Iran's succession and Hamas ceding Gaza governance: assemble those and 2.24% looks like compensation priced off tomorrow's demand rather than yesterday's oil. That configuration carries 15% of the mapped scenario weight, and it is the path on which this argument loses.
What happened the last time an oil crash drove disinflation
The 1998 comparison is worth making carefully. Energy prices collapsed, headline inflation ran low, and the Federal Reserve eased. It also eased into a financial-stability shock: Russia's default and the rescue of a large hedge fund had frozen credit markets, and those cuts answered that emergency as much as they answered the inflation prints. Calling the episode a misreading of energy-led disinflation would be an assertion rather than a finding, and this piece does not make it.
The part that transfers is arithmetic. The following year the energy base rolled. The prints that had looked disinflationary were accurate, but they sat on a moving base, and that base reversed on its own schedule regardless of what anyone traded or voted. Inflation reaccelerated and policy turned back the other way. The lesson is narrow: a disinflationary statistic built out of a recent window changes when the window moves, whether or not the world changes with it.
Whether the same arithmetic is running beneath a 2.24% breakeven has a date attached to it.
What would have to be true for the market to be right, and when we find out
Three things move the argument. June CPI prints Tuesday 14 July against May's 4.25% base; if it lands near that level, the case for durable disinflation weakens, and the nominal 10-year, live at 4.57%, is the first place a response would show. The next GDPNow revision is the growth arbiter, its 1.3% reading unchanged while claims at 215,000 argue the other way. Real yields are the third test and the one that would bite fastest, since a further leg up in DFII10 would pressure equity valuations without waiting for a data release. None of that is a recommendation to act on it.
What would prove this piece wrong: a soft June CPI arriving alongside genuine de-escalation in the Gulf, bleeding the risk premium out of crude and handing the breakeven a demand-side reason to be right. The gaps in the evidence are real too: no CFTC refresh this cycle (the 7 July vintage is two cycles old), no TGA or M2 read, no USD/JPY for the BOJ tail.
What is different from 1998 is where the exposure sits: then in the policy rate, now in a market price set against a realized 4.25% that anyone can see. What is not different is the base, and the base is doing the arithmetic.