When policy shocks meet market relationships

Governments can change the information markets use in an instant. Prices then adjust on different clocks: Treasury yields can reprice within hours, while fuel prices pass through to retail surveys over days or weeks. To make these shifts visible, we compare a treated price with a related comparison price before and after each event.

Government bond yields: different paths after the announcements

First, we subtract the euro-area AAA 10-year par yield from the U.S. 10-year Treasury yield. Both series are in percentage points. The comparison controls for some movements shared by bond markets, but the euro area was itself exposed to both trade and geopolitical news.

The pre-event line falls sharply. After Liberation Day, the gap initially falls again, then rises over the following weeks and moves through several further phases. On April 3 itself the observed gap was roughly 11 basis points lower than on April 2; the higher intercept of the long post-event fitted line is an extrapolation from later data, not an immediate upward price jump. The −28 basis point DiD summarizes nearly eighteen months of subsequent developments and cannot be read as the tariff announcement’s isolated effect.

In the shorter June window, the pre-event yield gap drifts down, while the fitted post-event line slopes upward from a lower level. That change in slope is more informative visually than the single −14 basis point average. It is also a crowded event window: Israel began striking Iran on June 13, before the June 21 U.S. strikes. Market expectations were already changing.

At the pump: diesel versus gasoline

The second comparison subtracts U.S. regular gasoline from U.S. retail diesel. Both are weekly national prices in dollars per gallon. Their difference is the retail diesel premium, and the DiD is its change in dollars per gallon. Weekly surveys limit how precisely we can date the first response.

After Liberation Day, gasoline’s average rises while diesel’s average edges down. The premium was already narrowing beforehand, however. The fitted lines suggest a change in the pace of that decline, not a clear step downward on April 7. This chart is evidence of a relative price change in this selected window, rather than evidence that tariffs caused it.

Here the post-event dots visibly separate from the earlier diesel–gasoline relationship. The diesel premium jumps in the first post-week and remains elevated. Yet the June 13 Israeli strikes fall inside the pre-period and may already have altered oil and diesel expectations. The +25 cent contrast describes the price path around the wider conflict, not the marginal impact of the U.S. strike alone.

After U.S. forces began Operation Epic Fury on February 28, both fuels rose. The first March 2 diesel premium was barely above the preceding week’s; the divergence built over subsequent weeks. The scatter therefore shows a pronounced change in slope and cumulative repricing rather than a one-week 51 cent jump.

What the five charts establish

Markets did not react in one uniform way. The Treasury–euro gap changed direction across event windows; the diesel–gasoline gap narrowed around Liberation Day and widened during the Iran conflicts. The same DiD machinery makes these different responses directly legible.

These are descriptive contrasts, not identified treatment effects. The euro-area yield is affected by global policy news; gasoline is affected by the same oil shock as diesel. Monetary policy, refinery margins, seasonal fuel demand, advance expectations and other decisions also move the gaps. The fuel estimates rely on only four to six weekly observations on each side, while the long yield estimate spans many unrelated developments. DiD would need a credible unaffected comparison group and a plausible parallel-trends argument to support a causal claim.

The useful lesson is narrower and more durable: a political decision can arrive at a known date, but its market footprint can be an immediate price move, a new slope, or a slow pass-through. The dots and separate fitted lines force us to distinguish those stories before attributing a number to an event.