Last Wednesday's ADP employment report caused quite a stir, reporting a measly estimated 38,000 jobs gained for the month of May. This in part caused the S&P 500 (SPY) to drop almost 2.5% on the day, as the strength of the US economic recovery is called into question. However, before you jump to conclusions or go buying 10 year Treasury Bonds at <3% yield, please take a second look at the usefulness of this report.
To call the initial, unrevised data from this report "inaccurate" is an understatement, and may be akin to calling a root canal "uncomfortable." Wildly inaccurate might do it justice.
I ran some data up to last May 2010 on ADP. Please take a look.
These numbers are in thousands, and since I don't have the revised BLS numbers for April, I just went ahead and re-used the unrevised number. Generally, the BLS revisions are not very large.
| Date | ADP Unrevised | ADP Revised | BLS Private | BLS Private Revised | High/Low | Difference Between ADP Unrevised Versus BLS Revised |
| May 2011 | 38 | |||||
| April 2011 | 179 | 177 | 268 | 268 | Low | 89 |
| March 2011 | 201 | 207 | 230 | 231 | Low | 30 |
| Feb 2011 | 217 | 208 | 222 | 240 | Low | 23 |
| Jan 2011 | 187 | 189 | 50 | 68 | High | 119 |
| Dec 2010 | 297 | 247 | 113 | 139 | High | 158 |
| Nov 2010 | 93 | 92 | 50 | 79 | High | 14 |
| Oct 2010 | 43 | 82 | 159 | 160 | Low | 117 |
| Sep 2010 | -39 | -2 | 64 | 107 | Low | 146 |
| Aug 2010 | -10 | 10 | 67 | 93 | Low | 103 |
| Jul 2010 | 42 | 37 | 71 | 107 | Low | 65 |
| Jun 2010 | 13 | 19 | 83 | 31 | Low | 18 |
| May 2010 | 55 | 57 | 41 | 33 | High | 22 |
You'll see that there are huge errors in the ADP Unrevised Data when compared to the actual private jobs numbers as reported by the Bureau of Labor Statistics. This is often 100s of thousands of jobs, which is very significant.
For those who believe that perhaps the data is rigged by the BLS to present a rosier outlook, you'll also notice that that ADP is frequently high as well as low.
Also, you'll see that, unlike the BLS, there are some very large revisions on ADP's part, occasionally 40-50,000 jobs.
While occasionally close, this report is highly inaccurate statistically. I would recommend waiting to see the actual statistics Friday before making any moves in your investments. Additionally, I'd recommend in the future that you remain skeptical of this report.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
