Back on Track
This morning's initial jobless claims number is meaningful in that it reflects two weeks of data. The previous week's numbers were estimated due to the omission of 19 states, including California and Texas, due to the holidays.
Importantly, the data is reasonably strong, continuing a trend we've seen since September, adjusting for the one-time effects of Hurricane Sandy. As a reminder, we expect claims will continue to improve through February and should decline to a level in the 340-350k range on a 4-week rolling basis. Then, we would expect tailwinds to reverse to headwinds in March, as has been the case for the last three years. See the first chart below for an illustration of the trend.
Recall that to ferret out the noise of seasonality distortions, we look at the annual change in rolling NSA claims. Last week the NSA YOY change was -6%, which compared with -5.9% in the previous week. A larger negative number represents accelerating improvement.
Initial jobless claims rose 12k to 372k from 360k. The prior week's number was revised up by 12k to 362k from 350k. Normally we look at the unrevised vs. unrevised for apples to apples comparisons, but this week, due to the missing states, we're comparing unrevised to revised. Rolling claims rose 0.25k WoW to 360k.
Joshua Steiner, CFA