Business cycle is still going down (as indicated by falling steel prices) and defensive sectors are supposed to outperform SPY in this environment. Lately XLU is underperforming SPY significantly, which happened twice in last several years. In 2015 it foretold a big market crash. In early 2019, while the divergence was relatively small, it predicted the May correction. This time there is a significant divergence. I expect a significant decline in stocks in coming weeks.