We take the data on its merits as it comes in. Initial claims fell 7k last week to 444k from 451k (revised up 3k). This brought the rolling four-week average down by 5k to 458.5k. While this marks the third consecutive week of improvement the fact remains that at 444k, claims are still where they were in late 2009, 4-5 months ago. We've been highlighting for the last few weeks the fact that a divergence has emerged between claims and XLF performance. For now that remains the case, although XLF has given back 6.6% in the last 15 trading days, so the divergence appears to have narrowed (albeit for unrelated reasons, i.e. Greece/EU concerns).
We remain concerned that without significant improvement in claims, a leading indicator, there can be no meaningful improvement in unemployment, a lagging indicator. By extension, without improvement in unemployment it will be difficult for credit costs to return to what are considered "normalized" levels. At a minimum, a return to those normalized levels will be delayed. Remember, for unemployment to fall meaningfully, initial claims need to fall to a sustained level of 375-400k. We remain 45-70k above that level - roughly where we've been for five months now.
As a reminder around the census, we had been bullish on the lift the census would add going into its peak employment months. However, now that we're into May, it's time to start focusing on the drag it will create on the backside.
The following chart shows the raw claims data.
The following chart shows the census hiring timeline. If the past two cycles are an appropriate model for this year's census, we should start to see Census employment draw down as we move into June, creating a headwind for employment.
Joshua Steiner, CFA