The lazy, hazy days of Summer: There's not a whole lot going on during these low volume, final weeks of August. Traders are on vacation, earnings season is mostly over. We’re in a news vacuum, a "drift and bounce" mode.
The biggest stories this morning are Anna Kournikova's new sports bra, and the new flavor slurpee at 7-11. Talk about your slow news day.
My takeaway from much of the recent economic releases is how every number comes with an *asterisk, every release has some hair on it. Whether this is a coincidence or not I am unprepared to say at this time.
Example 1: Take this week’s and next week’s New jobless claims data with a grain of salt: Claims fell 17,000 to 386,000 for the week ending August 16th; Sounds good, until you realize that this number under reported new claims -- the Northeast was closed late Thursday and all of Friday due to the blackout; By the same token, the report next week will overreported new claims.
What was lost in the jobless discussions was the prior week's jobless claims were revised upwards to (ping!) 403,000. I hasten to point this out only because so many people have proclaimed 400,000 as some kind of a magic number.
Another example, the Philly Fed index of current activity "increased notably this month, from 8.3 in July to 22.1 in August." That's a tripling in a month. What makes this survey number so suspect is:
Despite the pickup in overall activity, shipments, and new orders, more firms than normal scheduled plant shutdowns and production slowdowns over the past several months. Thirty-six percent of the firms indicated that they regularly schedule plant shutdowns or production slowdowns during summer months, but 45 percent indicated they did so this year. For those reporting shutdowns this year, around 40 percent indicated that the production decreases were greater than usual. The most common reasons given for scheduled reductions were demand conditions, plant maintenance, and summer vacations.
Apparently, they are 3 times as active this summer in the Philly are, yet plant idlings are up nearly a third, from 36% to 45%. Fascinating. I wonder how employment is doing in the area:
The purported improvement in manufacturing conditions this month is not yet resulting in increased employment at the region's manufacturing plants. More firms reported declines in employment (20 percent) than reported increases (11 percent), and the current employment index fell from 0.8 in July to -8.7 this month. The diffusion index for the average workweek, however, remained positive for the second consecutive month.
The Philadelphia Story: More activity with 1/3 less plants open and the employment index dropping 8.7%. Now if only we could spread the Philly miracle to the rest of the country . . .
Our 3rd and final *asterisked release is the leading economic indicators (LEI). This composite of ten economic indicators “predicts” economic activity six to nine months in future.
In the hall of mirrors that is Wall Street, we see some of the same datapoints bootstrapping each other. The S&P500, money supply, and the interest-rate spread (10-year Treasury and fed funds rate) are 3 key components of LEI. Money supply and the rate spead (or yield curve) are within the control of the Fed, and Alan & Co. Have been obviously working overtime to get the economy going again. So the LEI is a bit of a reflection of what the Fed hopes will happen, rather than what is actually happening.
As to the SPX, and its predictive value, lets do a quick Rally Comparison.
I direct your attention, dear readers, to the 4 prior rallies since March 2000. These SPX rallies ranged from 29 days to 108 days, with an average run of 70 days. The present move peaked on June 17th, lasting 97 days. The range of these gains were from 21.7% to 28.7%, with the bounces averaging 24.7%.
So I must ask the unpleasant question: What is the predictive value of any rally, given the prior four failures to predict anything?
Post script: I do not want to leave you with the impression that I’m some Uber-Bear.
My analytical model is based on five factors: 1) MacroEconomic data; 2) Technical Trends; 3) Monetary Conditions; 4) Market Sentiment; 5) SPX Valuation.
Each component is over/under weighted depending upon different factors.
Presently, the model shows a mixed picture: 2 positives, 1 negative and 2 mixed data series. Both the technical trend and monetary conditions are positive; Sentiment is now mildly negative (a little too bullish); In March, Sentiment was wildly Bearish, and hence positive.
Mixed components include MacroEconomics (Improving GDP and ISM, terrific Productivity, versus bad employment, low capacity utilization, and very bad debt data) and Valuation (the so-called Fed model is good, while P/E, P/B, P/S are all at historically pricey levels).
We are somewhat overbought at present, so I’m looking for a pullback into Sept/Oct, which we view as a buying opportunity.
I have no idea of where the indices will be year end, but I look at the market as ~12% undervalued based upon forward S&P500 earnings estimates, which are of course, frequently incorrect nonsensical analyst guesstimates.
Of course, this is subject to change -- mostly based on interest rates, employment data and end user demand.
TrackBack URL for this entry:
Listed below are links to weblogs that reference Dog Days:
The comments to this entry are closed.