Daily Growth Index vs GDP
On February 18th the markets rallied in part on the news that a widely followed leading economic index rose again in January (for the 10th consecutive month), this time by .3%. In contrast, the key indexes provided by the Consumer Metrics Institute show that the demand side of the economy is already shrinking at a nearly 1.5% annualized rate.
Much of the key data for the widely followed leading economic index is on page 7 of the official release, and that data provides a cautionary context that is lost in the summary observation of a .3% gain:
1) The movements of the sub-components of the index were very mixed, with five of the components rising, four dropping and one flat. The rising components during January (in order of greatest net contribution to the .3% gain) were the yield curve, vendor performance index, average work week, consumer expectations and stock prices.
2) Several of those five rising components beg disbelief. The most glaring is the rising stock prices for January (reportedly rising 1.2%), which does not remotely resemble the performance seen in month end statements (with the S&P 500 actually dropping about 3.5%). The statistical method applied in the LEI (an average of the daily closing prices) simply doesn't match the way real world portfolio performances are captured (comparing month end balances).
3) Meanwhile another component, the rising consumer expectations number, was sharply contradicted less than a week later by the same organization. Such massive revisions on such short notice should raise at least some doubt about the reliability or timeliness of the earlier data.
4) Simply replacing the consumer expectations and stock price components with more current information would cause the index to signal net contraction.
5) Another rising component (and the most positive of the lot) is arguably irrelevant to a recovery at this time. The yield curve has been sending a strong growth signal continuously since late 2007, but borrowing is obviously being constrained by factors other than interest rates. A favorable yield curve is simply not providing the kind of stimulus that has occurred historically with similar interest rates. Furthermore until something fundamental changes within this economy, no amount of low interest is going to make a difference.
6) Disregarding the yield curve component would also, by itself, have caused the index to signal net contraction.
7) Seven of the ten components were actually measured, but three were statistical "imputations"
"To address the problem of lags in available data, those leading ... indicators that are not available at the time of publication are estimated using statistical imputation. An autoregressive model is used to estimate each unavailable component. The resulting indexes are therefore constructed using real and estimated data, and will be revised as the unavailable data during the time of publication become available."
What this means is that 30% of the index is merely an estimate subject to later revision.
8) For investors all of the above arguably makes the index of marginal value, especially since the index includes old market data as a key component. Its not even clear that stock price movement data is a good predictor of the economy, since the relationship between the health of the equity markets and the health of the economy is both highly complex and subject to mutual feedback (just consult the 2008 sub-prime market meltdown for examples).
Recently the Consumer Metrics Institute has developed a radically different source of leading indicators that provide more timely and less frequently revised data. On a daily basis we mine on-line consumer tracking databases to extract macroeconomic intelligence about the demand side of the consumer economy. This places our indexes about as far "upstream" in the flow of the economy as possible. The combination of daily updates and a consumer focus result in a high resolution index that reacts quickly to significant consumer spending changes in a number of different segments of economy.
Currently our indexes disagree strongly with the upbeat story portrayed by the widely followed index. Our indexes generally peaked in August and have declined substantially since. Our daily "real-time" reading of the trailing 91-day "quarter" slipped into contraction on January 15th, 2010 and has lately been recording year-over-year contraction of about 1.5% -- placing that sliding "quarter" among the lowest one-eighth of all calendar quarters of GDP growth recorded since 1947 by the Department of Commerce.
Photo:
http://www.prlog.org/
