Menu Close

Time To Play Wall Street’s Favorite Game

It’s time to play “Guess The Number!” Every month  we get together with Wall Street pundits and media types to play this fun for all ages game. It’s based on guessing a fictitious number that the majority never guesses correctly, but a few lucky winners  randomly do. Their prize for guessing right is… Wait for it… Nothing. That’s right, they win nothing unless they also guess how the market will react to the news. That depends entirely on whether bad is bad, bad is good, good is good, or good is bad. Your chance to win is as good as the experts!

Results will be announced Friday morning at 8:30 AM ET but many winners won’t be able to cash in until 9:30 when NY opens. If you are a futures player, congratulations! You can cash in right away when the news comes out.

The number we are all trying to guess is the seasonally adjusted abstraction of the BLS’s (Bureau of Liar Statistics) monthly first guess at the previous month’s non farm payrolls. Bear in mind that the BLS will revise the number 7 times before it’s finalized… in 2016. The BLS even warns that the statistical margin of error on the first guess is 90,000 +/-.  It used to be 100,000, but the BLS changed that to 90,000 a few months back. They’re getting better at guessing. Unfortunately, the conomic forecasting community isn’t.

Each month organizations like Bloomberg, Dow Jones, Econoday, and survey leading Wall Street conomists for their wild guess as to what “the number” is. If the consensus is off by approximately 20,000 or more, the market goes wild for 20 minutes in reaction to the misunder or misoverestimate. In terms of the total payrolls of around 135.5 million, a consensus guess that misses the mark by around 15 thousandths of one percent will start people running around with their hair on fire, buying or selling if they’re traders or, if they’re pundits, trying to explain what went wrong, when in fact nothing did. The actual, not seasonally adjusted data has actually been incredibly consistent over the past year. And getting to within 1 or 2 hundredths of a percent actually is pretty good. But the market and the media, which profits from the excitement of the game, demand even greater accuracy.

The trend of the actual, not seasonally adjusted (NSA) survey data has been boringly consistent, rising at the annual rate of 2.1 million, give or take 50,000, each month this year, which isn’t materially slower than the annual rate of 2.1 to 2.3 million from April to December 2012. The recent rate equates with a gain of 1.57%. The market does settle down to recognizing that the trend is your friend after a half hour or so. It’s just the first reactions in the futures and at the NY opening bell that are so much fun.

So you’re possibly wondering what my guess ,as a non-conomist technical analyst, of this month’s Number is. The data I look at suggests that the conomic consensus is way too low. The guess as of Wednesday morning in’s leading published survey is that payrolls will rise by 159,000. Applying some very simple ‘rithmetic to the just released ADP survey of private payrolls, and also the real time daily withholding tax data for May from the US Department of the Treasury, suggests that ultimately the number could, and should, be much larger than that. Whether it will be or not is a whole nother story.

With the vagaries of the seasonal adjustment process and other guesswork applied in the BLS Establishment Survey, the headline number often does not accurately represent reality. It usually heads in the right direction, although even that can go haywire at times.

Even if the indications of the ADP survey and withholding tax data align, the payrolls number initially may not. Sooner or later it invariably will, but nobody cares about later. They’re only interested in the advance flash estimate posted on the first Friday of the month following the month in question. It does not matter that this number almost always suffers a large revision the following month, the month after that, and every year on the benchmarking of the data. The first number is the only one the market cares about even though it’s always wrong.

ADP says that seasonally adjusted private payrolls rose by 135,000 in May. The consensus of conomists was for 157,000. They were too high. It suggests that they’re too high on their non-farm payrolls (NFP) guess, or in other words that the report will “miss.”

Another way to look at this data is by the usual difference between the total monthly ADP number and the BLS’s NFP number. Over the past 12 months, the BLS number has been between 21.9 million and 22.2 million higher than the ADP number. The differntial has been trending higher during that time. In April the difference was 22.2 million. That would have included any secastration furloughs. For the last 3 years the average difference in the May number has been 22.342 million.  Assuming that the difference in the May number this year is 22.2 million, the monthly gain in headline NFP would be just 125,000.  However, if the number is closer to the 3 year average May differential, staying on the conservative side by rounding down to 22.3 million, the month to month gain in jobs would be 225,000. You can see just how sensitive the number is to what are really inconsequential differences.

The tax withholding data suggests that the numbers are even stronger than even the best case from the ADP numbers.  I track this data every week in the Professional Edition Treasury Report and it has been a good indicator of whether the number will miss or beat, although it doesn’t forecast the actual number well.

The average monthly tax take was up approximately 11% year over year in May. The difference was virtually the same using the effective date of the BLS survey of the 12th of the month as a reference point.  The change in tax rates in January appeared to amount to a gain of 6.5%, leaving a net gain of 4.5%.  The big variable that’s difficult to estimate correctly is the impact of inflation. Compensation inflation ran at zero on a  year to year basis in April, but has averaged closer to 2% over the past year. If the gain due to inflation in May was 2%, then the implied gain due to an increase in the number of persons employed would  be around 2.5%.

Real Federal Withholding Taxes - Click to enlarge
Real Federal Withholding Taxes – Click to enlarge


Unfortunately there hasn’t been a clear correlation between gains in withholding taxes and gains in jobs. The withholding tax gains are skewed by bonuses, commissions, and non-employment withholding.  The difference between the year to year gains in withholding and the year to year gain in jobs has varied, but one thing has been consistent. When withholding gains have been this strong, the headline jobs number has consistently handily beaten the consensus guesstimate. Based on that relationship, my guess would be that the the gain in jobs will be greater than the recent year to year trend gain of 1.57%. That would equate with a year to year gain in May of 2.1 million or a monthly gain of approximately 175,000. That’s simply if the data stays on trend.

A year to year gain of 2.5% would be off the charts. That’s just not going to happen this month or probably ever. But it does suggest that the May number will be on trend at a monthly gain of at least 175,000 or better.

That’s good news that is bad news. A number that strong would ignite market fears of the Fed Taper Worm (First ZIRP Bernankecide kills all the elderly savers, then The Taper comes for the rest of us). Pundits and traders would immediately conclude that the strong number means that the Fed will begin to cut QE sooner rather than later, and everybody’s knee jerk reaction to that would be to SELL!

Now that we know that, we can all panic now and avoid the rush.

That’s my handicapping of this month’s Guess The Number, Wall Street’s favorite game that everyone can play. This report is for entertainment purposes only of course.  If you suspect you have a gambling problem call 1-800-GuessIt now!

Stay up to date with the machinations of the Fed, Treasury, Primary Dealers and foreign central banks in the US market, in the Fed Report in the Professional Edition, Money Liquidity, and Real Estate Package. Try it risk free for 30 days. Get the research and analysis you need to understand these critical forces and stay ahead of the herd. Click this link and begin your risk free trial NOW!

Read Payrolls Gain But QE Has No Impact On Growth Rate

Read Claims Data Gets No Help From QE, While Stock Prices Were Extended

Follow my comments on the markets and economy in real time @Lee_Adler on Twitter!


Join the conversation and have a little fun at If you are a new visitor to the Stool, please register and join in! To post your observations and charts, and snide, but good-natured, comments, click here to register. Be sure to respond to the confirmation email which is sent instantly. If not in your inbox, check your spam filter.

This site uses Akismet to reduce spam. Learn how your comment data is processed.