Fascinating set of charts via Martin Pring at Real Money, applying trend following and technical analysis to NFPs.
"Technicians assume that markets move in trends and that once a trend begins, it remains in force until enough indicators prove that it has reversed.
Trends, of course, can range from intraday trends to secular trends extending over several decades. Obviously it’s not possible to apply intraday analysis to economic data, because most of this information is published on a monthly basis.
To monitor these cycles in nonfarm payrolls, I decided to divide each monthly data point by a 12-month moving average. The great thing about a 12-month average is that it eliminates any seasonal biases. The result is the indicator plotted at the bottom of the first chart."
Here are 4 of Pring’s charts on the issue. They are rather instructive.
Larger chart:091107_pring01.gif
Larger chart: 091107_pring02.gif
Larger chart: 091107_pring03.gif
Larger chart: 091107_pring04.gif
Very interesting approach . . .
Source:
Trading Nonfarm Payroll Charts
Martin Pring
RealMoney) 9/11/07
http://www.thestreet.com/p/rmoney/technicalanalysis/10378989.html
The economy continues to struggle. The Retail Sales y/y % change declined.
I have posted charts about Industrial Production approaching a critical level. I plotted it against Real Durable Goods Consumption, which leads IP at turning points.
See
http://www.wrahal.blogspot.com
The thing that really stand out to me immediately is that under the booming Bush recovery, the employment indicator barely nudged above and for not very long above the top oscillator (.5%)line. Gee, like I have been saying, it was a jobless recovery.
Really interesting that the bottom line oscillator in each chart looks really enemic coming off the 2001/2002 bottom. If you draw a trend line above the highs of this oscillator from the 1978 peak, it shows a HUGE negative divergence to the various values shown in the upper lines in each chart. Hmmmm, are we on the cusp of an even larger turn down, similar say to a fifty or hundred-year rain? Flood?
test.
Weakest bounce since 1958?
Reading between the lines a bit further, the recent high in the NFP oscillator at the bottom of each chart reached a high of about 1.0 at the end of 2005 or early 2006. This means that the monthly NFP data point got up to, but did NOT get over the 12 month moving average.
All previous bounces in the NFP oscillator off its lows got well above the 12 month moving average (above 1.0 on the right scale) and stayed above it for various amounts of time. For instance in 1994, the oscillator went above the 12 month moving average by 50% (1.5 on the right scale) and then followed the 12 month moving average (which was presumably quite positive at the time) more or less until 2000 or so.
Currently, if the next few NFP monthly data points are also below the 12 month moving average (below 1.0 on the right scale), the oscillator’s recent bounce off the 2001/2002 lows will be the weakest SINCE 1958!
Hmmmmm.
Peter wrote: “If you draw a trend line above the highs of this oscillator from the 1978 peak, it shows a HUGE negative divergence to the various values shown in the upper lines in each chart.”
Interesting observation. The massive divergences display a classic example of the misallocation of resources of which Mises wrote.
Each chart emphasizes the boom/bust cycle inherent with central/fractional-reserve banking methodology, and the subsequent misallocation of resources to which those policies lead.
Interest rates in their pristine form are discounters of future consumption – a higher rate indicates present consumption is preferred instead of time-delayed consumption; however, central banks via fractional reserved banking time and again get this simple principle wrong, and attempt to manipulate supply and demand via interest rates when in fact interest rates are simply a measure of supply and demand. (Consider falling mortgage rates due to lack of present demand compared to rising CP rate due to increased present demand.)
Once again we are on the same path where falling real interest rates indicate a weakening of demand for present credit and present consumption. Without artificial stimulus, the misallocated resources must respond to current demand by reversion to supply/demand equalibrium.
It is not the reversion that causes the pain but the misallocation that preceded it. Bubbles cannot burst without first being created.
Not to dispute the general points, but this is not entirely accurate: “The great thing about a 12-month average is that it eliminates any seasonal biases.”
Only when assuming that the “seasons” and annual events fall at the exact same dates, which they are not. Some holiday schedules are off, week-of-the-month based events jitter, and the start and magnitude of weather seasons varies, all of which makes for overall jitter, which is the same problem the BLS has with its ARIMA model.
It only works when you are looking at a point “safely away” from varying events like Easter, spring break, etc. I’m inclined to believe any seasonal smoothing model will work then.