Issue Number 47 - Reporting Perspective
Section Three
(View all Sections of this Issue)
REPORTING PERSPECTIVE
The Big Three Market Movers
Most underlying economic fundamentals have deteriorated sharply in recent reporting, with a downside acceleration and catch-up seen in post-election payroll reporting. The election is over, so any short-term manipulations are likely to be financial-market oriented, aimed at pacifying the still highly unstable financial conditions. The apparent "concurrent seasonal factor bias" in headline reporting of payrolls continues and is assessed further in the Employment/Unemployment section.
The extraordinary drop in oil and gasoline prices will reduce the reported annual CPI inflation for October and probably November. Reporting patterns, again, are likely to reflect the rapid flow through of oil price swings to the downside, in contrast to the slow pick-up in reporting of energy-related costs when oil prices are rising. Those patterns tend to run counter to common experience.
Messrs. Bernanke and Paulson need a stable or relatively strong U.S. dollar in the evolving systemic solvency crisis, and such requires stronger economic data than commonly might be expected in the now widely accepted recession. With the financial crisis threatening national security, almost anything remains possible in the arena of data and market manipulations. Data manipulation is an extremely inexpensive and effective policy tool.
Absent manipulation, and against market expectations that have shifted sharply to the downside, most near-term economic reporting still should tend to surprise the markets on the downside. Market-moving reports, such as employment and GDP, however, still may show massaged results that come in on the upside of expectations. With inflation expectations tanking along with oil prices, beyond the October CPI report, most inflation reporting should surprise expectations on the upside.
Employment/Unemployment — As discussed and graphed in the Opening Comments, and as noted in the November 7th Flash Update, annual payroll contractions continued to show a pattern never seen outside of recession. Against a background of rapidly deepening recession and intensifying weakness in all other key employment/unemployment indicators, the Bureau of Labor Statistics (BLS) reported that the October seasonally-adjusted payroll levels plunged and that the unemployment rate soared, a picture much closer than usual to reality, but still something shy of it.
The pattern of revisions in the October report (released post-election) showed that pre-election reporting of job losses actually had been well shy of reality in initial estimates of August and September payrolls. Those happy headline numbers not only were relatively good news at the time for political incumbents, but also were taken positively by troubled financial markets that had been expecting something worse, each time.
The initial reporting of jobs loss in August of 84,000 was revised to a 73,000 decline with the September report. In the October report, August’s decline deepened in revision to 127,000. September’s employment report — the last before the election — also showed a less severe than expected 159,000 drop in employment and an unchanged unemployment rate. Now, with October’s reporting (the first since the election), September’s payroll data underwent extreme negative revision, to a 284,000 decline, while October showed a much larger than expected surge in unemployment. History would suggest this pattern is more a function of catch-up following political manipulation of the data, than it is of normal variation in monthly reporting. Also, as suggested by the ongoing upside concurrent seasonal bias, misreporting of the data continues.
The BLS also announced that its annual benchmark revision will adjust for an unbelievably low 29,000 jobs overstatement as of March 2008. The issues tied to the benchmarking process and the related use of the gimmicked birth-death model will be reviewed in a later newsletter.
Payroll Survey. The BLS reported a statistically-significant, seasonally-adjusted jobs loss of 240,000 (down 419,000 net of revisions) +/- 129,000 for October, following a revised 284,000 (previously 159,000) jobs loss in September. Annual contraction (unadjusted) in total nonfarm payrolls continued to deepen, down 0.85% in October, versus a revised 0.52% (previously 0.43%) in September. The seasonally-adjusted series also contracted year-to-year, down by 0.78% in October, versus a 0.51% (previously 0.38%) contraction in September.
Concurrent Seasonal Factor Bias. The pattern of impossible biases (see the Reporting/Market Focus in SGS Newsletter No. 43 of June 10, 2008) being built into the headline payroll employment changes intensified with October reporting. Instead of the headline jobs loss of 240,000, consistent application of seasonal-adjustment factors — net of what we are calling the concurrent seasonal adjustment bias — would have shown a more-severe monthly jobs loss of about 308,000. This upside reporting bias has been seen in 11 of the last 12 months, with a rolling 12-month total upside headline-number bias of 678,000.
Examining the seven months of revised data, in place since the last benchmark revision, confirms that an unusual upside reporting bias has been in play for the headline payroll numbers. In recent reporting, headline payrolls generally have come in slightly better than consensus expectations, helping the financial markets of the reporting day.
Before the standard two revisions that follow a headline release, the headline bias was positive in six out of the seven months, aggregating to 275,000. (In 11 out of the last 12 months, again the bias was positive, aggregating to 678,000.) After two revisions (non-benchmark), however, the seven months of data show now that three months ended up with negative biases, and four months had positive biases, aggregating to plus 44,000. The latter pattern is what should be typical for headline number reporting.
Birth-Death/Bias Factor Adjustment. A minor element in October that added upside pressure to the payroll number was the monthly bias factor (birth-death model). Never designed to handle the downside pressures from a recession, the model added a 71,000 upside jobs bias to October 2008 (same amount as in October 2007), and followed a net upside bias of 42,000 jobs in September 2008. The process boosted financial-activities and construction jobs by 13,000 and 7,000, respectively, for the month. Although the adjustments are made to the unadjusted series, they flow through at roughly the same magnitude in the seasonally-adjusted series.
Household Survey. The usually statistically-sounder household survey, which counts the number of people with jobs, as opposed to the payroll survey that counts the number of jobs (including multiple job holders), showed household employment fell by 297,000 in October, following a 222,000 loss in September.
The October 2008 seasonally-adjusted U.3 unemployment rate showed a statistically-significant increase to 6.50% +/- 0.23% from 6.12% in September. Unadjusted U.3 rose to 6.1% in October, versus 6.0% in September. The broader October U.6 unemployment rate jumped to an adjusted 11.8% (11.1% unadjusted) from 11.0% (10.6% unadjusted) in September. Refigured for the bulk of the "discouraged workers" defined away during the Clinton Administration, actual unemployment, as estimated by the SGS-Alternate Unemployment Measure (graphed in the Alternate Realities section), rose to 15.8% in October, up from 15.0% in September.
In 1994, the BLS completely redesigned and redefined the unemployment series and all its measures, broad and narrow, so that the new series going forward could not be compared with the old series. I still am struggling to take my alternate measure back before 1994, where finding consistent and good data is a major problem. That said, the U.6 broad measure of 11.8% unemployment is the highest since the same reading in January 1994, where such was the first reading under the new definitions, during a period of unofficial recession.
Employment Environment. The deterioration in October’s employment environment continued in line with, but still shy of reality, per trends indicated by the better-quality employment-environment indicators: September help-wanted advertising remained at its historic low; new claims for unemployment insurance continued to surge sharply in terms of annual growth; and deepening, recession-level employment readings were seen in both the October manufacturing and nonmanufacturing purchasing managers survey. Since the employment and unemployment indicators tend to be coincident markers of broad economic activity, weaknesses in these numbers are signaling an ongoing recession in place.
Next Release (December 5): Based on continuing deterioration in underlying economic activity, the November payroll survey should plunge by more than 300,000 jobs, along with a further spike in the unemployment rate. The Treasury/Fed’s needs to continue to salve the markets, however, suggest that reporting likely will come in a little better than consensus expectations in place in advance of the report.
Gross Domestic Product (GDP) — As discussed in the Opening Comments and noted in the October 30th Flash Update, the Bush Administration signaled a week before the data release that not only would the "advance" estimate of third-quarter 2008 GDP growth show a contraction, but also that the fourth quarter likely would, too. Such would place in the U.S. economy in a formal recession based on traditional definition (see the Reporting/Market Focus).
The Bureau of Economic Analysis (BEA) estimated a 0.25% contraction +/- 3% (95% confidence interval) in the "advance" estimate of annualized real (inflation-adjusted) growth in third-quarter GDP. The number was statistically indistinguishable from average economic growth or a deep contraction, and it was against estimated second quarter real growth of 2.83%.
With the sharp reversal in current quarterly growth against the unbelievable annualized 4.8% growth reported for third-quarter 2007, year-to-year change slowed sharply, with annual third-quarter GDP growth slowing to 0.81%, from 2.05% in the second quarter. The SGS-Alternate GDP estimate is for an annual contraction of roughly 3.3% versus an annual (not annualized) contraction of 2.9% in the second quarter. Against reporting of underlying economic series, and annualized quarterly contraction in excess of 2% would have been more realistic than the 0.25% estimate. The SGS-Alternate GDP measure is graphed in the Alternate Realities section.
The difference between the reported 0.3% annualized Gross Domestic Product (GDP) and the consensus expectation of a 0.5% contraction was no more than statistical noise, yet the reported result most certainly was manufactured so as to allow the hypesters on Wall Street and in Washington to spin their fairy tales of a "less-severe recession" in order to help draw the gullible back into stocks, at least for a day or two before the election. Such followed earlier economic scare tactics aimed at the public to help sell the "bailout" package.
The 0.3% contraction was a plug number, as its calculation included significant "guesstimates." For example, the "advance" estimate was based on only two out of three highly volatile and suspect monthly trade reports. The BEA could have brought in the reported growth at a small plus, just easily as a small minus, but the credibility of perpetual GDP growth may have reached its limits and was abandoned publicly by the White House. The reported "advance" growth estimate usually is massaged so as to come in close to consensus, in this case a little bit better. Keep in mind that the 0.25% contraction is an annualized rate, only 0.06% ($7.4 billion out of $11,720.0 billion) quarter-to-quarter, a magnitude well within the scope of regular monthly revisions and statistical noise.
Against the heavy upside biases built into GDP reporting, the BEA also had to allow an inflation surge in order to get real GDP into contraction. The GDP implicit price deflator (GDP inflation rate) exploded to an annualized 18-year high of 4.09%, versus the multiple-year low of 1.26% used to keep second-quarter GDP growth in positive territory (the lower the GDP inflation rate used, the higher the reported real growth).
The BEA’s GDP-like measures for third-quarter 2008, including Gross National Product (GNP), where GDP is GNP net of trade in factor income (interest and dividend payments), and Gross Domestic Income (GDI), which is the theoretical income-side equivalent to the GDP’s consumption-side measure have not been estimated, yet, given the particular lack of significance associated with the "advance" estimate. The GDI and GDP, which have shown various quarterly contractions since the fourth-quarter of 2006, likely will be updated with the next GDP release.
General background note: Although the GDP report is the government’s broadest estimate of U.S. economic activity, it is also the least meaningful and most heavily massaged of all major government economic series. Published by the BEA, it primarily has become a tool for economic propaganda.
Next Release (November 25): Underlying economic fundamentals suggest that the "preliminary" estimate revision of third-quarter GDP should revise to show a deeper contraction, but the initial contraction reported was shallow enough to revise away and become a shallow economic expansion. The September trade deficit was reported with enough improvement to allow for the possibility of the third-quarter GDP growth to revise into positive territory. The BEA will bring in the revision wherever it chooses to, likely helping to meet perceived financial-market needs of the moment.
Consumer Price Index (CPI) — As discussed in the October 16th Flash Update, September CPI inflation eased back again on both a monthly and annual basis, as the CPI continued to absorb the impact of an ongoing tumble in oil prices. Yet, monthly "core" inflation also eased (CPI-U core was up by 0.1% in September, 0.2% in August, and 0.3% in July) despite significant anecdotal evidence of surging non-energy and food inflation. As discussed in the Opening Comments, significant further hits on CPI inflation likely lie ahead in October and November reporting, thanks to the continued collapse in gasoline prices, but annual inflation will remain positive and likely rise in the months following.
CPI-U. The Bureau of Labor Statistics (BLS) reported seasonally-adjusted September CPI-U (I.6) eased by 0.03% (down by 0.14% unadjusted) +/- 0.12% for the month, versus a 0.14% (0.40% unadjusted) decline in August. Year-to-year or annual inflation in September backed off to 4.94% from 5.37% in August.
Annual inflation would increase in October 2008 reporting, dependent on the seasonally-adjusted monthly gain exceeding the 0.26% monthly increase seen in October 2007. The difference in growth would directly add to or subtract from September’s annual inflation rate of 4.94%.
CPI-W. Annual inflation for the narrower CPI-W — targeted at the wage-earners category where gasoline takes a bigger proportionate bite out of spending — eased to 5.4% in September from 5.9% in August and 6.2% in July. The CPI-W is used for making the annual cost of living adjustments to Social Security payments, and the 2009 adjustment — based on the July to September 2008 period — was 5.8%, two-and-a-half times the 2.3% adjustment for 2008. Such is not good news for federal budget deficit, discussed elsewhere.
C-CPI-U. Year-to-year or annual inflation for the Chain Weighted CPI-U (I.5) — the fully substitution-based series that increasingly gets touted by CPI opponents and inflation apologists as the replacement for the CPI-U — eased to 4.34% in September from 4.70% in August.
Alternate Consumer Inflation Measures. Adjusted to pre-Clinton (1990) methodology (I.7), annual CPI growth declined to roughly 8.3% in September from 8.7% in August, while the SGS-Alternate Consumer Inflation Measure (I.8), which reverses gimmicked changes to official CPI reporting methodologies back to 1980, eased back to roughly 12.9% in September from 13.2% in August. The alternate numbers are not adjusted for any near-term manipulations of the data.
The eight levels of annual inflation, I.1 to I.8, are detailed in the table in the Alternate Realities section, along with the graph of SGS-Alternate Consumer Inflation.
Next Release (November 19): Annual October CPI inflation should decline, based on a continued collapse in oil and gasoline prices. Average gasoline prices in October were down roughly 20% from September, which would tend to knock off roughly 1% from the not-seasonally-adjusted monthly CPI. The seasonally-adjusted effect should be less, though, given that gasoline prices tend to ease some at this time of year. Consensus expectations (briefing.com) of a monthly 0.8% drop in the adjusted CPI are not unreasonable. Longer-range impact from earlier dollar weakness and high oil prices, and high broad money growth rates will tend to lead to upside CPI surprises in later months.
As to annual CPI inflation, any seasonally-adjusted monthly increase above or below the 0.26% monthly gain seen in October 2007 would directly add to or subtract from September 2008’s annual inflation rate of 4.94%.
Other Troubled Key Series
Federal Deficit. Rapid deterioration continues in the U.S. government’s fiscal condition, partially as a result of the impact from the ongoing systemic solvency crisis. The 12-month rolling deficit through October 2008 — the first month of fiscal 2009 — rose to $635.1 billion, up from the September’s $454.8 billion and from October 2007’s $169.1 billion.
In terms of fiscal-year 2008 (year-ended September 30th) the official federal deficit was $454.8 billion, against official forecasts just three months ago of a $389 billion deficit, and against a $161.5 billion deficit in 2007. These are the officially-gimmicked numbers, not GAAP reporting, which will show a much larger deficit in statesments due for release in mid-December. The GAAP-based deficit for fiscal-year 2007 topped $4 trillion and likely will top that in the 2008 GAAP-based estimate.
Confirming that a recession was in place long before the solvency crisis came to a head, government revenues in 2008 fell by 15.3% versus 2007. The budget deficit outlook for the current 2009 fiscal year is for continued sharp deterioration, given the rapid decline in economic activity and prospects for further lost tax revenues, and given soaring government costs tied to the systemic bailout and promised new stimulus packages. Official reporting likely will show a 2009 federal budget deficit in excess of $1.5 trillion. The downturn in revenues tied to the weak economy also has started to impact state and local government fiscal operations, as well, with serious funding problems already surfacing in a number of jurisdictions.
Viewing the change in the level of gross federal debt bypasses several of the regular reporting manipulations of the government’s financial results and is a better indicator of actual net cash outlays by the federal government than is the official, gimmicked deficit reporting.
Gross federal debt stood at $10.574 trillion at October 31, 2008, up by an extraordinary $549 billion for the month and up $1.495 trillion from October 2007, which in turn was up $495 billion from October 2006. As of the end of September 2008, the close of the government’s fiscal year, gross federal debt stood at $10.025 trillion, up $379 billion for the month and up by $1.017 trillion from September 2007, which in turn was up $501 billion from September 2006.
Initial Claims for Unemployment Insurance — The continued, rapid rise in initial claims for unemployment insurance reflects the severe deterioration taking place in labor conditions. On a smoothed basis for the 17 weeks ended November 8th, annual growth hit 43.3%, up from 37.0% as of the 17 weeks ended October 11th, and up from 29.6% as of the 17 weeks ended September 6th. A rising growth trend in new claims is an economic negative.
General background note: More often than not, week-to-week volatility of the seasonally-adjusted weekly claims numbers is due to the Labor Department’s efforts to seasonally adjust these numbers around holiday periods (such as Veterans Day and Thanksgiving). The Labor Department has demonstrated an inability to do such adjusting successfully. When the new claims series is viewed in terms of the year-to-year change in the 17-week (four-month) moving average, however, such generally is a fair indicator of current economic activity.
Real Average Weekly Earnings — Reflecting impact from a further pullback in gasoline prices, September’s seasonally-adjusted monthly real earnings were unchanged in September, following a 0.6% increase in August and a 0.6% decline in July. Annual change in September remained in deep contraction, holding at August’s level of 2.5% year-to-year decline, following an annual contraction in July of 2.8%.
General background note: Gyrations in the poor quality of reported CPI growth account for most month-to-month volatility in this series. Adjusting for the major upside biases built into the CPI-W inflation measure used in deflating the average weekly earnings, annual change in this series shows the average worker to be under severe financial stress in an ongoing structuralrecession (see the Hyperinflation Special Report of April 8, 2008, and this month’s Reporting Focus).
Retail Sales — As discussed and graphed in the Opening Comments, the pattern of monthly quarterly and annual contractions in real (inflation-adjusted) retail sales continued to deteriorate in September and October, and likely will deepen further in November, as the holiday shopping season kicks in. These growth patterns show weakness never seen outside of major recessions.
The Census Bureau reported this morning (November 14th) that seasonally-adjusted sales for the month of October fell by 2.77% (down 3.14% net of revisions) +/- 0.6% (95% confidence interval), versus a revised decline of 1.29% (previously 1.16%) in September. The decline was exacerbated by the fall in gasoline prices, but as indicated in October’s "core" measure, it still was significant. On a year-to-year basis, October sales were negative for a second month, down by 4.11%, versus a revised 1.15% (previously 1.03%) annual decline in September.
Real Retail Sales. The reported monthly and annual declines in October retail sales also will show intensifying declines, net of inflation, despite a likely sharp monthly decline in the October CPI. Deflated by the September CPI-U, seasonally-adjusted real (inflation-adjusted) retail sales contracted by 1.26% for the month of September (down 1.29% before inflation adjustment), following a 0.31% decline in the month of August (down 0.45% before inflation adjustment). On a year-to-year basis, September retail sales fell by 5.81% (down 1.15% before inflation adjustment), versus a 3.66% decline in August (up by 1.50% before inflation adjustment).
The series now has suffered its fourth consecutive quarter-to-quarter real contraction (an annualized decline of 10.8% in the third quarter versus a 1.2% drop in the second quarter) and the third consecutive quarter of annual contraction (down 4.4% for the third quarter versus a 1.6% decline for the second quarter). The fourth quarter is off to an even worse start. Such ongoing negative growth patterns never have been seen outside of formal recessions.
Core Retail Sales. Consistent with the Federal Reserve’s predilection for ignoring food and energy prices, "core" retail sales — retail sales net of grocery store and gasoline station revenues — fell by 1.65% in October (down 2.13% net of revisions), versus a revised 1.64% (was 1.47%) decline in September. Those numbers contrasted with the official aggregate drop of 2.77% in October and the revised 1.29% decline in September. On an annual basis, October "core" retail sales fell by 5.95%, versus a revised 4.43% (was 4.26%) decline in September. The sharp declines shown in "core" retail sales are after a 12.7% decline in October gasoline station sales, tied to reduced gasoline prices.
Next Release (December 12): November retail sales should continue showing a pattern of deepening monthly and annual contractions, net of inflation. Odds favor a result somewhat weaker than likely consensus forecasts.
Industrial Production – As discussed and graphed in the Opening Comments and detailed in the October 16th Special Update, the Federal Reserve reported a 2.8% monthly contraction in September industrial production, following a revised 1.0% (previously 1.1%) drop in August. September’s weakness was attributable to a strike at Boeing and to oil and gas related shutdowns in the Gulf of Mexico region due to hurricanes. September year-to-year change plunged by 4.5%, after a 1.4% annual contraction in August. Adjusting for strike and hurricane impacts, however, third-quarter production still was down on both a quarterly and an annual basis.
Not adjusted for the special factors, following a 3.1% annualized quarter-to-quarter contraction reported in the second quarter, industrial production suffered its second consecutive quarterly downturn in the third quarter at an annualized 6.0% rate of decline. In conjunction with a 2.7% contraction in year-to-year growth for the third quarter, the series is showing growth patterns not seen outside of recessions.
Next Release (November 17): The October production numbers should not contract as sharply as in September, given the settlement of the Boeing strike and some resumption of curtailed oil and gas production in the Gulf of Mexico region. Nonetheless, production should be down both month-to-month and year-to-year, and likely will be weaker than already soft consensus estimates.
New Orders for Durable Goods — As discussed in the October 30th Flash Update, although the regularly-volatile new orders for durable goods reportedly rose by 0.8% in September, versus a 5.5% contraction in August, year-to-year change continued to decline in a recessionary pattern, down 2.4% from September 2007, versus August’s 8.9% annual decline. In the third quarter, the series showed its second consecutive quarter of year-to-year contraction (down 4.7%) and its third consecutive quarter-to-quarter contraction, down at an annualized 14.3% from the second quarter. The annual and quarterly contractions are before any adjustment for inflation.
New orders for nondefense capital goods also rose by 0.8% for the month, versus a 7.7% contraction in August, and fell 6.7% year-to-year, following a 5.2% annual drop in August.
General background note: Durable goods orders lost its status as a solid leading economic indicator when the semi-conductor industry stopped reporting new orders in 2002.
Trade Balance — The trade deficit reportedly narrowed in September, still reflecting likely oil import shortfalls and continued paperwork flow distortions. The Census Bureau and the Bureau of Economic Analysis reported the seasonally-adjusted September trade deficit narrowed to $56.5 billion from a largely unrevised $59.1 billion in August. With continued games-playing in reported oil imports, these data (net of inflation) could be used to help revise third-quarter GDP into positive territory.
The latest report showed continuing, unusual volatility in import and export carryover (imports or exports included in the current month that actually took place in earlier periods, when they should have been reported), along with an unusually sharp and unseasonable decline in daily crude oil imports to 8,443 thousand barrels per day, down 15.1% month-to-month, down 16.2% year-to-year. Separately, average imported oil prices were down 10.3% for the month to $107.58 per barrel.
Next Release (December 11): The trade deficit for October likely will be softened by falling oil prices. Whatever gimmicking has been taking place with the reporting shows no sign of abating. Significant catch-up in the deficit understatement is long overdue.
Consumer Confidence — As graphed and discussed in the Opening Comments, consumer confidence tumbled terribly in October, as the Messrs. Bernanke and Paulson’s efforts to terrify the public on the economy and individual financial prospects bore fruit.
The Conference Board’s Consumer Confidence measure suffered the largest percentage declines, ever, on both a monthly and annual basis, with confidence falling to its lowest level, ever. The Confidence measure fell by 38.1% for the month, following a 5.0% gain in September. Year-to-year October confidence fell by 60.1%, having been down by 38.3% in September.
The Reuters/University of Michigan’s Consumer Sentiment measure dropped by 18.1% in October, following a 12.6% rise in September. Year-to-year, Sentiment declined by 28.8% in October, versus a 15.7% decline in September.
These lagging, not leading indicators confirm that the economy has been in a deepening recession.
General background note: The Conference Board measure is seasonally adjusted, which can provide occasional, but significant distortion. The adjustment does not make much sense and is of suspect purpose, given that the Conference Board does not release the unadjusted number. The Reuters/Michigan survey is unadjusted. How does one seasonally-adjust peoples’ attitudes? Also, beware the mid-month Consumer Sentiment release from Reuters/University of Michigan. The sampling base is so small as to be virtually valueless in terms of statistical significance.
Short-Term Credit Measures — Annual growth in both consumer credit and commercial borrowing has continued to slow, intensifying recessionary pressures and highlighting difficulties the Federal Reserve still is having in stabilizing solvency issues in the U.S. banking system. With direct intervention as a lender in the commercial paper market, and with heavy jawboning of banks to lend to credit-worthy customers, the Fed is pursuing increasingly desperate approaches to stimulate both commercial and consumer lending.
For seasonally-adjusted consumer credit, which includes credit cards and auto loans, but not mortgages, annual growth was reported at 3.7% in September, down from 3.9% in August and 5.0% in July.
As reported by the Fed (Flow of Funds September 2008), home mortgage loan growth slowed from a seasonally-adjusted annualized growth rate of 6.1% in fourth-quarter 2007, to 3.3% and 1.4% respectively in the first and second quarters of 2008. The data, however, are of questionable quality.
In the current environment, where inflation-adjusted growth in income (see this month’s Reporting/Market Focus) is not adequate to support meaningful growth in the personal consumption component of GDP, GDP growth only can come from temporary debt expansion or savings liquidation. Accordingly, stagnating growth and eventual contraction in consumer debt remains an ongoing constraint on economic activity.
Annual growth in commercial borrowing continued to fall off. Commercial paper outstanding showed a 16.2% year-to-year contraction in October, versus a 13.2% annual decline in September, and an 8.0% drop in August. Fed intervention should begin to boost the annual growth rate.
Annual growth in September commercial and industrial loans slowed to 13.3%, down from 15.2% in August and from 17.5% in July. Slowing growth in commercial lending should place a damper on broad business activity.
Producer Price Index (PPI) — As discussed in the October 16th Special Update, the regularly volatile Producer Price Index (PPI) for finished goods contracted by a seasonally-adjusted 0.4% (0.1% unadjusted) in September, versus 0.9% (1.6% unadjusted) in August, as reported by the Bureau of Labor Statistics. The decline largely reflected the continued sharp decline in oil prices. Year-to-year PPI inflation in September eased to 8.7% from 9.6% in August.
On a monthly basis, seasonally-adjusted September intermediate goods fell by 1.2% (down 1.0% August), crude goods fell by 7.9% (down 11.9% August). Year-to-year inflation slowed but remained high, with September intermediate goods up by 15.4% (16.7% August) and September crude goods up by 26.0% (38.1% August).
Next Release (November 18): Allowing for the ongoing, regularly random volatility of the monthly price variations, PPI inflation reporting over the next six-to-nine months generally should favor upside surprises in official results, despite hits in October and November from the continued fall in oil prices.
Better-Quality Numbers
General background note: The following numbers are generally good-quality leading indicators of economic activity and inflation that offer an alternative to the politically-hyped numbers when the economy really is not so perfect. In some instances, using a three-month moving average improves the quality of the economic signal and is so noted in the text.
Economic Indicators
Purchasing Managers Survey: Manufacturing New Orders — The October purchasing managers manufacturing index (PMI) fell into official recession territory, even with the index reweightings that kept the aggregate index above 50.0 for a number of month’s this year. The Institute for Supply Management (ISM) reweighted its key index so that the PMI would better match GDP results. While the effort was well intentioned, altering the data to match the extremely overstated GDP growth rates damaged the reporting quality of the PMI. Fortunately, however, the more meaningful components of the index were not affected by the efforts to match the flawed government data.
The October PMI plunged to 38.9, down from 43.5 in September, to its lowest level since September 1982. While the ISM uses an index reading of 41.1 (now breached) as the break-point between recession in the broad economy and expansion, a reading below 50.0 means a contracting manufacturing sector. The 50.0 mark still has worked out as a solid, broad recession signal in my analyses that are unfettered by reliance on GDP data for a recession signal.
The various components of the ISM composite indices are diffusion indices, which are calculated as the percent of positive responses from the ISM survey plus one-half of the neutral or unchanged responses. Hence, a reading below 50.0 indicates a contracting series.
The October new orders index plummeted to 32.2, from 38.8 in September, and stood at its lowest reading since April 1980. The new orders have been in actual contraction since December 2007. Distortions from the seasonal factors calculated by the Department of Commerce can be minimized by viewing the series using year-to-year change on a three-month moving average basis. On that basis, the October new orders index fell by 25.9%, following a 19.9% decline in September and 15.1% drop in August.
The new orders component of the purchasing managers survey is a particularly valuable indicator of economic activity. The measure gradually has notched lower from its peak annual growth of 35.5% in April of 2004. As an SGS early-warning indicator of a major economic shift, new orders breached its fail-safe point in mid-2005, signaling pending recession.
Also a significant measure, the manufacturing employment component fell to 34.6 in October from 41.8 in September.
Service Sector Composite Index. This series does not have much meaning related to overall business activity, since new order activity at law firms, dentists, hospitals or fast-food restaurants has little obvious relationship to broad economic activity. With that as background, the October services composite index fell to 44.4% in October, from 50.2 in September.
Both the services employment and prices paid components, however, have some meaning. Covering the real estate and banking industries, among others, the October employment component remained below 50.0, falling to 41.5 from 44.2 in September. The heavily hit prices-paid component for both indices is covered in the Inflation Indicators.
Help-Wanted Advertising Index — (Newspapers and On-Line) — Please Note: The Conference Board has ceased issuing Web-based press releases on its help-wanted advertising in newspapers series, but the monthly data still are available for some undetermined period of time, upon request.
The seasonally-adjusted September help-wanted advertising index continued at its historic low level of 15, the same as in August, the lowest level seen since the index was first calculated at the end of President Harry Truman’s term in office.
The September reading was down by 37.5% year-to-year, versus a 34.8% decline in August. The annual change in the three-month moving average as of September was a 36.1% decline, versus August’s 33.8% contraction. Despite some of the historic weakness in the series being due to the loss of newspaper business to the Internet, and despite its looming abandonment by the Conference Board, the HWA remains a solid leading indicator to the broad economy and to the monthly employment report. It continues to signal severe deepening in the recession and ongoing deterioration in labor-market conditions.
Where the HWA series does not include a measure of on-line advertising, recent indices developed to measure Internet activity have serious definitional problems and still are too young to be meaningful indicators. That said, the Conference Board has reported that annual growth in its nascent on-line measure of help-wanted advertising has contracted on a year-to-year basis in each month from April through September 2008. Such cannot be a good sign for national employment or for broad economic activity.
Housing Starts — As discussed in the Opening Comments and as graphed there net of the New York City paperwork distortions in June’s reporting, housing starts continue to show a severe, deepening recession. The seasonally-adjusted September housing starts level was reported down by 6.3% for the month and down 31.1% year-to-year. August was down a revised 8.1% (previously down 6.2%) for the month, and down 34.8% (previously 33.1%) year-to-year. The latest numbers still show ongoing quarterly and annual contractions in housing activity never seen otherwise outside of recessions.
In home sales data, the seasonally-adjusted September new home sales rose by 2.7% +/- 14% (95% confidence interval), which was not statistically distinguishable from a loss, following a 12.6% (previously 11.5%) monthly decline in August. On a year-to-year basis, September new home sales dropped by 33.1%, following a 35.6% (previously 34.5%) decline in August. Reflecting the intensifying impact of foreclosure sales, existing home sales in September rose by 5.5% for the month, 1.4% year-to-year, while August fell by 2.2% for the moth and by 10.7% year-to-year.
Inflation Indicators
Money Supply — As graphed and discussed in the Opening Comments, significantly higher, broad money growth likely will result from recent Federal Reserve and U.S. Treasury actions, along with an increase in money velocity as bank lending and financial markets return to some semblance of normalcy (see the August 3rd Money Supply Special Report for a discussion of the practical measurement and analytical uses of money supply in assessing inflation prospects).
Annual growth in the seasonally-adjusted SGS-Ongoing M3 continued, but it is estimated to have slowed to 10.8% in October, down from 13.1% in September and off from its record-high 17.4% level in March. The slowing growth reflects serious difficulties the Fed has had in getting banks to resume normal commercial activity, although some positive impact has started to surface. With Fed actions working their way through the system, October’s annual growth should be close to the trough of the current cycle, with sharply higher M3 growth likely by the end of November and into December.
Outside of the last several months, the prior historic high of 16.4% was seen in June of 1971, two months before President Nixon closed the gold window and imposed wage and price controls. While October’s growth is well shy of 1971’s high, the current environment still promises heavy upside inflation pressure well into 2009.
For October 2008, annual growth for monthly M1 surged to 7.6% from 6.4% in September. Annual M2 growth increased to 7.4% from 6.2% in September, and the SGS-Ongoing M3 estimate showed annual growth of roughly 10.8%, down from 13.1% in September.
In terms of seasonally-adjusted month-to-month change, October M1 rose by 1.3%, following a 4.3% increase in September; October M2 rose by 1.4% following a 1.3% gain in September; and October SGS-Ongoing M3 fell by 0.4%, following a 0.7% increase in September. Monthly money numbers do rise and fall, with the latest monthly M3 contraction the ninth such occurrence in the last eight years.
One benefit of continuing to estimate the broader M3 measure is that it helps to explain what is happening in the narrower measures. For example, the recent strength reported in M1 and M2 did not reflect the Fed’s extreme attempts at systemic liquefaction, but rather a movement of funds out of M3 accounts into M1 and M2 accounts. As mentioned in the opening comments, a still-broader money measure than M3 likely would be showing rising annual growth.
General background note: Historical annual growth data and monthly levelsfor the money supply series, including the SGS-Ongoing M3 estimates, are available for download on the Alternate Data page of www.shadowstats.com. See the August 2006 SGS Newsletter for methodology. The indicated M3 levels are our best estimate and are provided at specific subscriber request. Keep in mind that regular revisions in the related Fed series affect historical M3. Usually, annual growth rates hold, although levels may shift a little. We have not attempted, nor do we plan to recreate a revised historical series for an M3 monthly-average level going back in time; the published series can be linked to earlier historical data available from the St. Louis Fed. The purpose of the SGS series was and is to provide monthly estimates of ongoing annual M3 growth. We are comfortable with those numbers and that our estimated monthly growth rates are reasonably close to what the Fed would be reporting, if it still reported M3.
Purchasing Managers Surveys: Prices Paid Indices — Both of the October purchasing managers composite surveys prices-paid indices took large hits, reflecting the severe plunge in oil prices.
On the manufacturing side, the October price index showed declining prices, with the index dropping to 37.0, from 53.5 in September, and hitting the lowest reading since December 2001. On a three-month moving average basis, October’s year-to-year change was a decline of 9.5%, versus a 17.1% gain in September and a 31.1% increase in August. The manufacturing price indicator is not seasonally adjusted and, therefore, is generally the better indicator of pricing activity.
On the non-manufacturing side, the seasonally-adjusted October prices diffusion index also dropped, but held in positive territory, falling to 53.4 from 70.0 in September. On a three-month moving-average basis, October’s annual gain was 2.7%, down from 18.9% in September and from 26.2% in August.
General background note: Published by the Institute for Supply Management (ISM), the prices paid components of the purchasing managers surveys are reliable leading indicators of inflationary pressure. The measures are diffusion indices, where a reading above 50.0 indicates rising prices.
Oil Prices — With the West Texas Intermediate (WTI) spot price closing at $58.24 on November 13th, oil prices have plunged by 60.0% since the record-high closing price of $145.66 on July 11, 2008, and have collapsed well below the $90 per barrel that level that promised ongoing severe near-term inflation problems in the U.S. economy. October’s monthly average spot price for WTI (St. Louis Fed) was $76.65 per barrel, down 26.2% from September’s $103.90, and down 42.8% from June’s $133.93 historic-high average. For October 2008, the year-to-year change in price level was a decline of 11.2%, versus a 30.0% annual gain in September.
As discussed in the Opening Comments, the lower oil prices likely will result in monthly declines in the CPI for October and November, but the long-term impact from recent high oil prices still is working its way through the broad economy, and still will add upside pressures to general inflation (exclusive of the near-term energy measures) into the first half of 2009. Those pressures reflect costs ranging from transportation of goods and services, to material costs in the plastics, pharmaceutical, fertilizer, chemical industries, etc. These broad inflationary pressures will remain intact for a while, despite any near-term oil price gyrations.
Oil prices remain highly volatile and sensitive to minor surprises. While slowing U.S. and global economies reduces oil demand, recent OPEC activities have been and likely will continue to be aimed at offsetting such, with production cuts. Adding some upside pressure to prices, the U.S. interregnum offers risk of intensified global military and political tensions, and other supply and demand risks/issues. Other than the possible impact of pre-election politics, the recent strength in the U.S. dollar has been key to the oil price decline, much as dollar weakness had been key to the earlier oil-price surge. At such time as heavy dollar selling resumes — and it will eventually — look for oil prices to spike anew, moving back above the $100 per barrel level and rekindling oil-price related inflation concerns.