What We Got Wrong: What We Got Wrong, Week of August 21, 2026

What We Got Wrong, Week of August 21, 2026 The Hormuz covert flows piece deserves a hard look in the rearview. We ran "Oil market is pricing covert Hormuz flows it cannot verify" and the premise was right there in the headline, the thing we were reporting on is, by definition, unverifiable. That's a shaky foundation for a price-explanation story. We were essentially saying: Brent is at $94 because of flows nobody can audit. Readers who noticed that we couldn't actually support the claim would be right to raise an eyebrow. We should have leaned harder on the Kpler data showing suppressed visible traffic and let that contradiction do the work, rather than anchoring the piece on an analyst assertion about secret shipments. The structural tension between what Hormuz traffic data shows and where prices are trading is a real story. The covert-flows framing turned a verifiable puzzle into an unverifiable explanation. The China demand coverage had a similar lean-on-one-source problem. Ole Hansen explicitly flagged a "massive divergence" between IEA, EIA and OPEC forecasts, and we led the week with the IEA's plateau assessment while giving the other two agencies a brief mention. We know OPEC and EIA are more bullish on Chinese demand, that's why the divergence matters. The piece read like we'd picked a side rather than mapped the disagreement. The IEA is arguably the most credible here, but when three major agencies are reading the same data differently, our job is to explain why, not just feature the most bearish outlook. The Russia refinery numbers didn't reconcile between articles and we didn't flag that. One piece cited the Oxford Institute for Energy Studies putting current capacity at 3.8 million barrels per day against a pre-war figure around 5.2 million, roughly a 27% drop. Another piece cited Reuters reporting roughly 17% of capacity had been taken out. Those figures aren't the same story. We ran them in separate articles without acknowledging the gap. Data center coverage this week was our heaviest cluster, four pieces across PJM, ERCOT, New York's moratorium, public opposition trends, and the 100 GW supply gap. Each piece was fine on its own but they shared enough ground that readers following the week would have encountered the same BofA 100 GW figure twice and the same theme of demand outpacing permitted supply without a unifying thread connecting the policy fragmentation to the investment numbers. A single synthesized piece would have served readers better than four partially overlapping ones. The BP governance story ran as a fragment. The headline promised a chairman departure and bearish North Sea sentiment; the article delivered neither a full account of what happened nor enough context on why it moves sentiment. If we couldn't complete that story, we should have held it. Where was coverage thin overall? The Alpine hydro drought and the EU's €6 billion extra import bill ran as separate stories that were obviously connected, one of the main reasons Europe's gas prices are elevated is that hydro shortfalls are pulling harder on the gas system. We didn't make that link explicit. Readers connecting those dots themselves shouldn't have to.