11 Trillion Dollars Are Missing
May 22, 2010
My post today is about a macroeconomic accounting mystery. Macroeconomic statistics attempt to measure impossibly big, complicated systems. For my dissertation research, I focus on national income statistics (the National Income and Product Accounts, the UN System of National Accounts, etc.) which try to measure the value of the production of all the goods and services in a given region. But I’m also fascinated by some of the companion statistics to national income/GDP like unemployment and inflation, both of which are measure by particular surveys (in the US, the Current Population Survey and the Consumer Expenditures Survey, among others). There are fascinating stories behind and around all of these practices (check out last week’s NYTimes Magazine on the GDP’s history for a starting place, or a new book by Thomas Stapleford about cost of living statistics).
There’s another big macroeconomic data set in the US that gets a lot less attention than the three I listed above: the Flow of Funds Accounts (or FoF) produced by the Federal Reserve Board of Governors. The Flow of Funds attempt to track the movement of money between various sectors and subsectors – banks and other financial firms, large and small industrial corporations, households, and government entities. The Flow of Funds tracks the movement of money by attempting to balance the flows in and out of each sector, and by attempting to balance the buying and selling of every kind of asset (or ‘instrument’). So, for example, F.110 looks at flows of money in and out of US-chartered commercial banks. F.220 tracks the buying and issuing of commercial mortgages. In addition to tracking flows, the FoF tracks levels – L.220 looks at how many commercial mortgages are outstanding in total and which sectors have them, rather than looking at changes.
Like the NIPA, and unlike unemployment and inflation statistics, the Flow of Funds are cobbled together using all sorts of heterogeneous data rather than being based on some single survey. Trying to track all of the money (much like trying to track all the production) is not a task easily accomplished with a sample. And so, because the Flow of Funds are cobbled together, they have some… quirks. My favorite quirk concerns the category “Unidentified Miscellaneous Financial Claim” (F.231 and L.231). What, you might ask, is an unidentified miscellaneous financial claim? We don’t know! Here’s the entirety what the Federal Reserve says in its description of the table from the Guide to the Flow of Funds Accounts:
For many sectors, unidentiﬁed miscellaneous financial claims are determined indirectly as the residual after the total of changes in individual ‘‘identiﬁed’’ asset or liability items for the sector (which appear on other instrument tables in the ﬂow of funds accounts) has been subtracted from the change in total assets or liabilities reported by the sector. For other sectors (nonfarm noncorporate business, the federal government, bank personal trusts and estates, private pension funds, and mutual funds, for miscellaneous assets; and nonfarm noncorporate business, for miscellaneous liabilities), the amount of such claims is obtained directly as the total amount reported by original sources as ‘‘other’’ assets or liabilities.
In most cases, the nature of the items in this category is truly unidentiﬁed. In some cases, however, items that are identiﬁed separately in original documents are included here because the items are not signiﬁcant enough from an analytical viewpoint to be classiﬁed as individual transaction categories. Examples are interest accrued by, prepaid expenses of, and real estate acquired for banking-house purposes by the Federal Reserve System (the major component of the monetary authority sector), as reported in the System’s Annual Report; and the intangible assets of U.S.-chartered commercial banks, as reported in their quarterly reports of condition. [Emphasis added.]
So, to reiterate, the Federal Reserve has not identified most of what’s in this class of instruments. Ok, but it’s a residual category – it’s probably small, right? Enh… you be the judge. In 2007, the “Net Change in Assets” in this category was $1.5 trillion. In 2008, that collapsed to a still large $365.3 billion. The level outstanding at the end of 2009 – $11.7 trillion dollars. There are more than 11 trillion dollars in assets that we’ve identified as existing, but not what kind of asset they are! And this category has grown tremendously – at the end of 1997, for example, there were a measly $4 trillion in outstanding “Unidentified Miscellaneous Financial Claims”.
So readers, I ask you – what could be in this category? What would have made it grow so much in the 90s and 2000s?
One possible type of asset that could be in this instrument is goodwill. Goodwill is an accounting term that attempts to capture the value of a firm purchased by another firm above and beyond the value of the assets of the first firm. E.g. if firm A buys firm B for $2 million, but firm B is only worth $1 million on its books, what happens to the balance sheet of the new firm A+B? A shelled out $2 million in real assets, but seems to have only gotten $1 million in return. There have been a number of solutions to this problem over time – one option was just to write-off the difference as an accounting loss. Another was to put the difference on the books of the new firm as “goodwill”, to represent the value of the purchased firm above and beyond its book value (in terms of ideas and synergies and future earnings potential from a brand name and all that) and then amortize the loss over a number of years. More recent solutions are more complicated, and involve keeping the goodwill. (Check out this 2007 paper by Ding et al. for a bit of the history of goodwill.)
But is that the whole story to this missing $11 trillion? What else could be in this category? And how would we figure it out?