Unemployment Initial Claims Funny Numbers

After assessing the unemployment numbers I have made a few observations.  First we must thank ZeroHedge for the information that 9 states did not report UI claims due to the holiday, so this number is really understated (look forward to a great revision).  We seem to be widening the gap on the number we announce and what the revised number becomes.  By spinning the number lower when the UI number is announced you make the number look like it improved or make it beat forecasts.  The problem is that the market doesn't look at the revision for last month, because this latest month shows better numbers (and lets face it, it's in the past).  To show you the delta, please look at this chart:

We can see that we have overstated the actual numbers only a few times and not so much in recent weeks.  A majority of the time we understate the numbers and then revise up (over the past year).  Lets look at the raw numbers and observe the effect this has:

As you can see, in prior weeks having the revisions, almost every case the numbers looked worse and it has effected whether we beat forecasts (falsely elevating the market as it looks like a beat).   Market pundits say the market prices in the forecast, therefore a beat makes the market go up, but revisions are not highlighted in an announcement (like the initial number) so it tends to get glossed over and not figured into the market price.  You have to take this information into consideration when looking at the health of the economy.  Don't celebrate too soon.