ShadowStats provides charts using “historical alternate data” for U.S. GDP, CPI, M3 money supply and employment. The site also supplies “as reported” historical data for those same indicators.
ShadowStats notes that The Federal Reserve Bank no longer published M3 after March 2006. However, the Fed has continued reporting M2, institutional money fund levels and large time deposits. These were the M3’s largest components. The SGS M3 Continuation estimate is based on this data plus a ShadowStats proprietary model to capture the remainder of M3. Changes in money supply have implications both for domestic economic activity and inflation.
ShadowStats calculates M3 data for its subscriber base. This is what caught my attention: ShadowStats generates econometric indicators that were used for market insight, yet are no longer released or accessible to market participants.
After the recent wild swings in asset values for exchanged-traded securities, there is a strong interest in longer-term views of economic cycles. I agree with that to the extent that more observations offers more degrees of freedom. Or, better yet, enough information at sufficiently frequent mark-to-market intervals that statistically robust methods like continuous rather than discrete probability distributions can be used to meaningfully assess volatility, be it actual, or implied. ShadowStats facilitates the use of longer term historical data for trading insights.
Some assumptions that must be confirmed in order for ShadowStats proxy indicators to be valid:
- the underlying data used to calculate discontinued metrics remain available
- the methodology used to calculate the discontinued metric was not proprietary to the original source, or otherwise undisclosed.
The second thing I liked about ShadowStats, at least what I could view in the non-subscriber section, was some interpretation of results that wasn’t full of doom-and-gloom hyperbole (although there seems to plenty of that in other sections of the site). About the M3 chart above:
Here we show year-to-year growth as a measure of the changing money supply. A downward slope in this growth curve does not necessarily mean that the money supply is dropping. Only if the curve goes below zero does that show money supply having contracted over a full twelve months. Also, for money supply changes over periods of less than a year, such need to be viewed on a seasonally adjusted basis. Unadjusted change over short periods may show changes that are little more than regular seasonal variations. Short-term changes also may run counter to year-to-year change, as seen in the latter part of 2009.
* Some might wonder, “What is this M-3?” which echoes my own thoughts. I recall the M1 and M2 measures of money supply from Fabozzi and my macro economics coursework at the Wharton School. I do not recollect a third metric for gauging the money supply. Nor do I remember M-3 from my days of proprietary trading in U.S. government bond futures and CBOT options-on-futures.