Inflation is one of the mortal enemies of interest rates. If dollars in the future buy less stuff than they do today, investors need to set higher and higher rates on the money they lend in order to realize the same returns. With that in mind, we'd be well within our rights to assume a surprisingly high reading on a key inflation report would push rates higher. In many past instances, that's exactly how things play out, but that's not what happened today.
To be fair, markets traded according to conventional wisdom in the first 20 minutes after the inflation report was released. But from that point on, the bond market rallied (i.e. bond prices rose and yields fell, thus implying lower rates for the mortgage market).
What's up with the paradoxical reaction? There are a few moving parts. The simplistic answer making the rounds is that this just "wasn't enough" inflation to derail the Fed's commitment to maintaining policies that support lower rates. That could be a factor, sure, but the unequivocal motivation is far more esoteric. It has to do with an imbalance of trading positions in the bond market and the subsequent exploitation or punishment of that imbalance.
In market jargon, it's known as a short squeeze and it basically means that so many traders were betting on higher rates that they were vulnerable to anything that pushed rates back in the other direction. Such a push would force those traders to buy bonds in order to prevent further losses. That buying pushes rates even lower, potentially hitting the stop-loss level for the next group of traders in line.
This is primarily a phenomenon that plays out in US Treasuries, but mortgage-backed bonds also improved steadily throughout the day. As a result, lenders increasingly recalled the day's initial mortgage rate sheets and offered improved terms (aka a "positive reprice"). Some of those lenders did that more than once, in fact! But even then, it's not in their nature to pass along all of the gains from the market to their rate sheets in one day. As such, bonds would only need to hold relatively steady for additional improvements tomorrow. To be clear, that's just an if/then statement and not something you'd want to take for granted as a prediction.