Today’s report on initial jobless claims was just plain bad. While economists were expecting claims to come in at a level of 270K, the actual reading was 24K higher at 294K. This was the highest weekly reading in claims in over a year (February 2015) and also the biggest miss relative to expectations since March 2015. Not only was this a big increase relative to last week, but it also comes on the heels of a relatively big increase in the prior week (+17K). On a combined basis, jobless claims have now risen by 37K over the last two weeks, which is the largest two-week increase in more than two years (April 2014). The next milestone to watch is the 300K level, which hasn’t been breached to the upside since March 2015 (62 weeks).
Given the increases in the last two weeks combined with the relatively low readings that dropped off of the count, the four-week moving average for claims rose by over 10K this week to 268.25K. Two weeks ago, the four-week moving average made a multi-decade low, and after the last two weeks, that level may be in place for a long time.
On a non-seasonally adjusted (NSA) basis, claims rose by 18.3K to 261.8K. While we have often cited the fact that NSA claims were at their lowest levels in decades, for the current week of the year they were actually lower last year. That being said, this week’s reading was still 74K below the average reading of 336K for the current week of the year dating back to 2000.
Finally, from their cycle low of 248K three weeks ago, jobless claims have now risen by 45K. While this is a very big increase relative to the last several weeks, we would note that it is hardly a rare occurrence. As shown in the chart below, since 1967 initial claims have risen by 45K or more over a three-week period 77 times, including five occurrences in the current economic expansion. Also, as far as this kind of increase being an indicator of recession or not, 38 have occurred during recessions, while 39 have been during periods of growth. That being said, the amount of time that the economy has been in a recession is considerably less than the amount of time it has been in an expansion, so as a share of total weeks, these kind of spikes are much more common in recessions.