Making Some Sense of The Paradoxical Reaction

Retail sales came in at -0.6 vs +0.1 forecast. Even the "control group" (excludes autos/gas/building materials) was -0.4% versus +0.3%. On most any other occasion, this would result in an obvious rally for bonds. To be fair, it did that at first, but then we spent the rest of the day selling off. Here's why: online store sales led the drop, both in the headline and control group. Forecasters did a crappy job of accounting for Amazon moving Prime day to July--possibly because they were preoccupied with oil-related volatility last month. The rest of the weakness was directly or indirectly related to fuel price deflation. Data aside, it was a rally week for bonds, and we often see Friday position squaring on rally weeks. Last but not least, inflation expectations ticked higher again in the Consumer Sentiment data.