Reporting/Market Focus  from August 2006 Edition of the SGS Newsletter

In response to subscriber demand, SGS has introduced and published in downloadable form on the www.shadowstats.com Alternate Data page, three alternate economic series. The SGS Continuing M3 is an estimate of growth in an ongoing version of the now defunct broad money supply series. The SGS Alternate Consumer Inflation and Alternate Gross Domestic Product estimates attempt to provide historical series that more closely track economic reality and common experience than do current, official government reports.
Continuing Money Supply M3. The Federal Reserve discontinued publishing M3, its broadest measure of the money supply, on March 23, 2006. The SGS continuing series estimates year-to-year change in the monthly average of seasonally-adjusted M3. The results are shown in a graph in the Opening Comments.
M3 consists of M2 plus institutional money funds, large-denomination time deposits, repurchase agreement liabilities and eurodollar holdings at foreign branches of U.S. banks.
More than 70% of the non-M2 components of M3 are accounted for by institutional money funds and large time deposits. The Fed has continued reporting institutional money funds as a memorandum item in its H.6 report on Money Stock Measures.
Large time deposits at commercial banks is reported regularly in the Fed’s H.8 report on Assets and Liabilities of Commercial Banks. These numbers allow modeling of a good estimation of the large time deposit number used in M3 calculations.
Representing less than 30% of non-M2 components of M3 and less than 10% of total M3, the repos and eurodollars are being modeled by SGS econometric models. They will be updated, along with large time deposits, as benchmarking information becomes available in the quarterly flow of funds accounts. Accordingly, published SGS M3 growth rates are subject to revision, as are the still-published M1 and M2 series as published by the Fed. Nonetheless, the bulk of the M3 estimation is solid and should allow the production of annual M3 growth estimates that are representative of Fed activity, as well as representative of what the Fed still likely follows with its own unpublished estimates of M3.
SGS Alternate Consumer Inflation Measure. As discussed in the October 2005 SGS, the BLS publishes a "current methods" CPI that restates historical CPI reporting for methodological changes made to the series since the early 1980s. By reverse engineering the series, it is possible to calculate what the BLS estimates the different changes to CPI reporting have added or subtracted to reported annual inflation. With nearly all of the changes resulting in reduced CPI inflation, one would have to add 3.6% to current annual CPI reporting to approximate what inflation would have been, based on the older methodology. This BLS series, however, allocates only 0.2% to the annual effect of the changeover to geometric weighting in the 1990s.
SGS estimates show that net effect of the geometric weighting change is roughly 3.1% CPI reduction in terms of present impact. According 2.9% (3.1% - 0.2%) is added on top of the other adjustments. The SGS Alternate Consumer Inflation measure has been calculated over time, adjusting for the effects of all the methodological changes on annual not-seasonally-adjusted CPI-U inflation, as well as using a three-month moving average centered on the month in question. The latest month is published as is and then is revised with the publication of the next month’s reporting.
Given the timing issues of when the BLS recognizes such factors as rising energy prices, the process eliminates some the unnecessary short-term volatility in the reporting. The results are shown in a graph in the Opening Comments.
Since the adjustment process is largely additive to existing reporting, the alternate series tends to show the same general up and down patterns of the underlying official CPI-U. The July 2006 annual rate of the SGS Alternate Consumer Inflation is 11.0%, which is reasonably close to rough estimates of the fully independent SGS inflation series that still is under development.
Shadow Government Statistics Alternate Gross Domestic Product Measure. Over time, the GDP series has strayed from reality, based on methodological changes often tied to the deflation process. The SGS Alternate GDP is an estimate of the historical series net of those changes, and other changes, where quantifiable, that I have been able to identify. The results are shown in a graph in the Opening Comments. The series is far from perfect. The unofficial recession of 1986 is almost there. The shallow 1995 downturn is in place, while the 2000/2001 recession is much longer and deeper than officially reported and already has started into a double-dip. In official reporting, the 2000/2001 recession has been revised away.
This series is an attempt to approximate reality, and often is adjusted as an additive function on top of the current GDP series. Accordingly, most of the ups and downs tend to move with each other in both the SGS and official series.
Separately in the works is an independent GDP measure based on a regression analysis of underlying economic series of known good quality. The patterns in that series should be close to what is shown here. As with the SGS Alternate Consumer Inflation series, the three-quarter moving average on annual growth rates is used to smooth insignificant variations in growth that often result from reporting games playing.