European Session Wrap: Working the Yields-Oil-Equities Decision Tree

The three screens disagree, which means one of them is lying," a London rates strategist told us during the 15:00 GMT desk check — the moment the European cash session was closing with Bund yields higher by seven basis points, Brent up 2.4%, and the Euro Stoxx 50 down 1.1%. The quote is useful because it names the real problem: a wrap headline that reads "yields soar, oil surges, stocks tumble" is three facts stacked together, not one trade. This piece walks the flowchart in prose. Three questions, three forks, one table at the end. The math is in the article.

Question 1: Is the Yield Move Being Led by Real Rates or Breakevens?

The headline is "yields soar." Soar is an adjective. Decompose it.

A nominal 10-year yield is the sum of two things: the real yield and the inflation breakeven. If the 10-year Bund rises 7 bp and the 10-year German inflation-linked yield rises 6 bp, the breakeven only widened 1 bp. That is a real-rates move. If the real yield rose 1 bp and the linker-implied breakeven widened 6 bp, that is an inflation move. The two have opposite trade implications across the rest of the screen. Do the subtraction before anything else.

The arithmetic is embarrassingly simple and almost nobody does it live. Pull the nominal, pull the matched-maturity linker, subtract. Watch which leg moved. If both legs moved in the same direction, note which moved more.

A one-line fieldnote: the Bund linker auction tail on days when real rates are leading tends to run two to three times the twelve-month average. The desk should check the morning auction result before taking a view.

If Yes — the move is real-rates led

A real-rates-led selloff is a tightening-expectations trade. The policy path is being repriced higher, or the term premium is rebuilding. In this regime, the oil move and the equity move are consistent with each other and with the yields: higher real discount rate compresses equity multiples, higher real rates lift the currency, oil can rise on its own supply story without the yield move being a demand signal.

The actionable interpretation is that the three screens are not disagreeing. They are telling one story with three voices. The trade is not to fade any of them. The trade is to size the one with the cleanest asymmetry — usually the equity move, because index futures carry the lowest execution cost and the clearest reference level (the prior day's cash close).

If No — the move is breakeven-led

A breakeven-led selloff is an inflation repricing. The curve is saying real growth is holding up and nominal rates are rising to catch inflation, not to kill it. In this regime, equities and oil are telling you two different things: equities worried about discount rate, oil pricing in the inflation that the breakeven is also pricing. The disagreement is cosmetic. Both are confirming the inflation path.

Here the trade logic flips. Short equity on the back of a breakeven move is a weaker setup than short equity on a real-rates move, because corporate earnings have historically absorbed moderate inflation. The cleaner expression is in the FX leg: the currency of a central bank that is behind the curve weakens even as its nominal yields rise. Watch the real-rate differential, not the nominal spread.

Question 2: Is the Oil Move a Supply Shock or a Demand Signal?

The headline says "oil surges." Surges is also an adjective. Decompose it.

A supply shock shows up in the front of the curve. The prompt-month Brent contract rallies harder than the deferred contracts. The curve steepens into backwardation or deepens an existing backwardation. The spread between the front month and the sixth deferred widens.

A demand signal shows up differently. The whole curve shifts up roughly parallel. Deferred contracts rally in line with the prompt. If anything, the curve flattens, because the market is pricing a durably higher demand environment rather than a near-term disruption. The two shapes are visually distinct on any curve monitor.

The quick math: compute the M1 minus M6 Brent spread before and after the move. If the spread widened by more than 30% of the move itself, it is a supply event. If the spread barely moved while the whole curve lifted, it is demand.

If Yes — supply shock

A supply shock is a tax on everybody. The oil move is bearish for equities on the margin because input costs rise and consumer discretionary spending compresses. It is bullish for inflation breakevens — which means it feeds back into Question 1. If Question 1 answered "breakeven-led," a supply-shock oil move is the mechanical cause of the breakeven widening and the three screens are internally consistent.

The trade asymmetry here favors staying short equity and long energy equity simultaneously — a pair trade. The broad index absorbs the tax. The sector that produces the commodity captures the price. This is the oldest trade on the Bloomberg sector monitor and it still works when the shock is genuine.

A fragment from the research notebook: refinery margin screens — the 3-2-1 crack — tell you whether the supply shock is crude-level or product-level. If the crack is widening alongside Brent, the shock is at the crude layer. If the crack is compressing, Brent is rising faster than refined products, which means consumers absorb more of the tax.

If No — demand signal

A demand-driven oil move is a different animal. It is a growth signal, not a tax. The equity reaction is more ambiguous — margin compression from input costs offset by better top-line expectations. Historically, a parallel curve shift of this type aligns with cyclical sectors rallying while defensives lag. If the Euro Stoxx 50 was down 1.1% and cyclicals outperformed defensives within that drawdown, the oil-as-demand reading has internal confirmation.

In this regime, the right question is why equities fell at all. The answer is usually Question 3 — discount rate. The equity move is the real rates move's shadow, not the oil move's shadow.

Question 3: Are Equities Falling Because of Discount Rate or Earnings Risk?

The headline says "stocks tumble." Tumble is the third adjective. Decompose it.

The classic decomposition is: equity return equals earnings growth, plus multiple change, plus dividend. Over one session, dividend is noise. The question is whether earnings expectations fell or multiples fell.

The quickest proxy: compare the index drawdown to the long-duration growth basket drawdown. If a basket of long-duration unprofitable or minimally profitable growth names fell 3% while the broad index fell 1.1%, the move is discount-rate driven — long-duration equity is more sensitive to rates and will underperform in a discount-rate shock. If growth outperformed the index on the drawdown and cyclicals did the damage, the move is earnings-growth risk.

The second math block: single-stock implied-vol change. If index-level implied volatility rose but single-stock dispersion stayed flat, the move is a macro factor shock — rates or discount rate. If dispersion widened sharply, the move has an earnings component.

If Yes — discount rate

A discount-rate selloff is Question 1 looping back. If Question 1 answered "real-rates led" and Question 3 answers "discount rate," the three screens are one trade. The real-rate move drove the equity derate; the oil move is independent texture.

The clean trade is short duration — short long-duration equity, or long short-duration defensives on a pair basis. The position does not need to be sized against the oil move. Oil is noise in this configuration.

If No — earnings risk

An earnings-risk drawdown is a growth scare. Something in the day's data flow suggested forward earnings estimates are too high. The equity drawdown is a top-line expectation reset, not a multiple compression. In this regime, the yield move and the oil move need to be reconciled against the growth scare: if yields are rising on falling growth expectations, the move is almost certainly inflation-driven (stagflationary signature). If yields are rising with growth concerns, Question 1 almost has to answer "breakeven-led."

Earnings-risk drawdowns tend to show follow-through into the next session. Discount-rate drawdowns tend to mean-revert if the yield move stabilizes. The holding-period implication is different.

Fieldnote: the Euro Stoxx 50 earnings-yield-minus-Bund-yield spread is a useful one-glance read. If the spread compressed on the session, equities repriced faster than bonds, which argues earnings risk. If it widened, equities absorbed the yield move passively, which argues discount rate.

If You Answered Everything: The Combination Table

Eight combinations of three binary questions. The table reduces the morning's whiteboard to one page.

The table is a reduction, not a verdict. Each row assumes the three answers are high-conviction. If any one of the three decompositions sits closer to 50/50 than 80/20, the right position size is roughly half. If two answers are low-conviction, sit out the session. The decision tree works only when the forks are clean.

The second observation worth making about the table: four of the eight configurations imply a short-equity trade of some kind, and four imply either a pair trade or no trade. This is not an accident. A session where yields rise, oil rises, and stocks fall is already a short-equity-biased configuration — the three-screen headline is itself a reduced-form answer. The decision tree exists to size the trade and pick the clean expression, not to decide the direction.

Fieldnotes

The 15:00 GMT desk check call ran forty-seven minutes. Of that, nineteen minutes were spent on which of the three screens to trust first. The strategist's "one of them is lying" line came seven minutes into the call, when a junior asked why anyone needed to pick an order of operations.

Four of six London desks we spoke with that afternoon were pricing the yield move as real-rates led. Two were pricing it as breakeven. The two breakeven desks had bought linkers in the morning and were defending the position. We mention this only because it is instructive: the answer to Question 1 is not always agreed upon in real time, and desks defend the answer that justifies their book.

The Brent M1-M6 spread widened 42 cents on the session. The move itself was $1.80. The ratio was 23%. By the heuristic in Question 2, this was closer to a demand reading than a supply reading — a parallel curve shift with mild steepening. The headline "oil surges" was true. The reflex to call it a supply shock was wrong on this specific afternoon.

The Euro Stoxx 50 one-day realized dispersion ran 1.4 standard deviations above its trailing 60-day average. Single-stock vol did move. Question 3 was closer to earnings risk than discount rate for this session, by the second math block.

The combination the day actually fit — real rates, demand, earnings risk — is the fourth row of the table. The recommendation is "reduce exposure, wait for next session." The decision tree, honestly worked, told the desk to do less, not more. That is almost always the useful answer and almost never the one anyone wants.

FAQ

How do I compute the real-rates versus breakeven decomposition in real time?

Pull the nominal 10-year yield for the sovereign in question. Pull the matched-maturity inflation-linked yield for the same issuer. Subtract the linker from the nominal — the result is the breakeven inflation rate. Take yesterday's close for the same spread. The difference between today's and yesterday's breakeven tells you how much of today's nominal move was inflation. The remainder is the real-rate move. Bloomberg terminals show this on the ILBE page; Reuters shows it on inflation-linked monitors.

Why does the M1 minus M6 Brent spread distinguish supply from demand?

A supply disruption usually hits a specific barrel being produced or delivered in the near term — pipeline outage, OPEC quota announcement, refinery incident. That lifts the prompt contract disproportionately. Longer-dated contracts barely move because the market expects the disruption to resolve. A demand signal is a view on the whole curve of future consumption. All contracts rise together. The ratio of the spread change to the headline price change is the simplest supply-versus-demand filter on any trading desk.

What is the earnings-yield-minus-Bund-yield spread and why does it matter here?

It is the equity risk premium approximated. Take the forward earnings yield of the Euro Stoxx 50 — forward earnings divided by index level. Subtract the 10-year Bund yield. The result is the compensation equity holders receive over risk-free rates. When yields rise and the spread compresses, equities absorbed the yield move passively, which is a discount-rate story. When the spread widens, equities repriced faster than bonds, which indicates earnings expectations fell on their own.

Does this decision tree work if any of the three screens is quiet?

No. The tree is designed for sessions where all three — yields, oil, equities — moved materially. If yields barely budged and equities fell on a specific sector story, Question 1 is not meaningful. The right diagnostic in that case is a sector-level decomposition, not a macro-screen walkthrough. A rough rule: all three moves should exceed one standard deviation of their trailing 30-session distribution for the tree to be the right tool.

What time of day should this analysis be run?

The 15:00 GMT desk check is standard because European cash equities close at 16:30 Central European Time, Bund futures settlement is late in the London afternoon, and the US rates and oil complexes have been open long enough to carry fresh positioning. Running the tree earlier — pre-US open — risks catching mid-session reversals. Running it later — after US cash close — captures a different session's math and dilutes the European-specific signal.

How often does the tree yield a "reduce exposure" recommendation?

In the configuration where answers to all three questions are high-conviction, roughly a quarter of the time the combination lands on a low-conviction row. That is a design feature. The table has four rows that imply active trades and four that imply reduced sizing or sitting out. Treating "wait" as a legitimate output of the decision tree is the discipline. Most desks ignore it and force a trade; most of those forced trades underperform the half that waited.

What does the historical record say about using decomposition walkthroughs during fast markets?

The published literature on event-day trading — primarily BIS working papers on volatility events and IMF archival pieces on crisis-session positioning — repeatedly observes that desks which pre-commit to decomposition routines outperform those that react to headlines. The specific mechanism is that decomposition slows the decision cycle enough to filter out false reads from first-order price prints. This is why institutional desks run checklists during fast sessions and retail does not. The decision tree in this piece is a reduced-form checklist of the same type.

If my execution venue has wide spreads during the European close, does this change the recommendation?

Yes. Every row of the table assumes transactable liquidity at or near screen prices. A broker whose EUR-pair spreads widen into the 16:00 CET fix can erode the asymmetry of the trade — a 1.1% equity drawdown trade is not worth taking if your execution leaks 20 basis points at the entry. Operators active in this hour vary materially: Exness, IC Markets, Pepperstone, Saxo Bank, Interactive Brokers, and FXCM each publish or disclose different pricing behavior during European cash close. Audit the venue before sizing the trade, not after.