December 10, 2013
Q3: 2013 Flow-of-Funds Report – ‘Tis the Season to Be Jolly
I know that being a Debbie Downer gets more face time on cable news, but after looking at the Fed’s latest Financial Accounts of the U.S. report, formerly known as the Flow-of-Funds report, I cannot contain my optimism about the economy’s prospects in the New Year. First and foremost, the report shows that combined Fed and depository institution credit creation is not only accelerating, but is growing at a relatively high rate. Secondly, household balance sheets continue to improve, which implies that depository institutions will be more favorably disposed to lend in 2014. Thirdly, part of improvement to household balance sheets is related to the upward trend in the value of residential real estate. Fourthly, the reduction in federal government spending/borrowing is resulting in “crowding in” of spending/borrowing by other sectors. Thus, the wringing of hands and the gnashing of teeth by unreconstructed Keynesians (i.e., The Consensus) over the deleterious effect of budget sequestration on the pace of economic activity is not occurring. And lastly, federal government borrowing relative to the size of the economy is less than its post-WWII median value.
Okay, let’s go to the charts. Plotted in Chart 1 are the year-over-year percent changes of quarterly observations of the sum of Fed and depository institution credit and nominal Gross Domestic Purchases. Fed credit is defined here as the Fed’s holdings of securities obtained by outright purchases as well as repurchase agreements. The sum of Fed and depository institution credit is advanced by one quarter because the historical relationship between the two series suggests that changes in this particular credit aggregate lead changes in nominal Gross Domestic Purchases. Notice that year-over-year growth in the sum of Fed and depository institution credit has been accelerating since Q4:2012, which is when the Fed’s third round of quantitative easing (QE) commenced. Year-over-year growth in this credit sum reached 8.7% in Q3:2013, the highest since Q2:2006. The median year-over-year growth in this credit sum from Q1:1953 through Q3:2013 is 7.2%. The recent relatively rapid growth in this credit sum suggests that growth in nominal domestic spending will be gaining strength over the next couple of quarters.
Chart 2 shows that the growth acceleration in the sum of Fed and depository institution credit starting in Q4:2012 was due to Fed QE actions. Growth in depository institution credit alone has been decelerating. But with the Fed expected to start “tapering” the quantity of securities it purchases by the end of Q1:2014, won’t the sum of Fed and depository institution start to decelerate then with negative implications for the growth in nominal domestic spending? Not if growth in depository institution credit picks up. Chart 3 shows that starting in this past October and continuing through November, growth in commercial bank credit, the principal component of depository institution credit, has been accelerating. In the eight weeks ended November 27, bank credit increased by almost $58 billion. If this pace of bank credit increase were maintained, it would more than compensate for the anticipated decline in Fed securities purchases early in 2014.
Which brings us to the balance sheet for households. Plotted in Chart 4 are quarterly observations of total liabilities of households and nonprofit organizations as a percent of their total assets. This ratio reached a post-WWII record high of 20.2% in Q1:2009. As of Q3:2013, this ratio had declined 520 basis points to 15.0%, the lowest since Q2:2001. This sharp decline in household leverage in combination with the sharp decline in the debt burden of households (see Chart 5) should make households appear more creditworthy in the eyes of depository-institution lenders.
An improving residential real estate market has played an important role in the improvement of household balance sheets. Chart 6 shows the leverage of owner-occupied residential real estate. After reaching a post-WWII record high of 63.5% in Q1:2009, leverage has fallen to 49.2% in Q3:2013, the lowest since Q2:2007. The rise in the value of residential real estate in the past two years and the absolute decline in mortgage debt following the bursting of the housing bubble are responsible for the leverage decline in owner-occupied residential real estate. Again, this improves the creditworthiness of households. The rise in residential real estate values also reduces mortgage loan write-offs by depository institutions, which should increase their willingness to put new loans on their books.
Speaking of residential real estate, it still looks attractive as an investment, but not as attractive because of the increase in its value and the rise in mortgage rates. Plotted in Chart 7 are quarterly observations of the imputed yield on owner-occupied housing, the effective mortgage interest rate and the differential between the two. The imputed yield on housing is calculated by dividing the Commerce Department’s estimate of the nominal dollar value of imputed shelter services produced by owner-occupied houses by the Fed’s estimate of the market value of residential real estate (then multiplied by 100 to put the ratio into percentage terms). If the imputed yield on housing is higher than the mortgage rate, then one can purchase an asset that currently is yielding more than the cost of financing it. From Q1:1973 through Q3:2008, there has been only one instance in which the imputed yield on housing was above the mortgage rate – in Q2:2003, when the differential was a mere 0.04 percentage points. But after the bursting of the housing bubble, the market value of residential real estate plummeted, boosting the imputed yield on housing. The rise in the imputed yield on housing in combination with the plunge in mortgage rates brought the differential into positive territory starting in Q4:2008, where it has remained through Q3:2013. This positive differential, after having reached its zenith of 3.7 percentage points in Q4:2012, has drifted down to 2.3 percentage points in Q3:2013. At first blush, the narrowing in the positive differential between the imputed yield on housing and the mortgage rate might suggest that the pace of the current expansion in residential real estate would moderate in 2014. But if depository institutions “loosen” their mortgage qualification terms due to the improved balance sheets of households, then “effective” demand for owner-occupied housing could increase.
Remember how the sequestration-induced federal government expenditure cuts and the increase in the marginal tax rate for upper-income households were going restrain nominal total spending in the economy? Well, they didn’t because of “crowding in”. The reduction in federal government spending in recent years along with increased tax revenues has reduced federal government deficits. Those entities that had planned to lend to the Treasury had its deficits been larger, now have some excess funds on their hands. They can either lend to some other entity that will spend – a household, a business, a state or local government, a furiner – and/or they can spend these funds themselves. Either way, the decrease in spending caused by the decline in federal government expenditures and the increase in taxes is offset by increases in other nonfederal government spending. So, as the federal government borrows less, other nonfinancial entities save less, i.e., spend more. This is what is meant by the term “crowding in”. “Crowding out” is the opposite – the government borrows more, other nonfinancial entities save more, i.e., spend less.
This is shown in Chart 8. Notice that as federal government net borrowing decreased precipitously in the second and third quarters of 2013, net lending by the remaining entities in the nonfinancial sectors also fell precipitously. Conversely, as federal borrowing surged in 2009, in part due to the fiscal stimulus, nonfinancial sector net lending also surged. This is an example of “crowding out”. In analyzing the macroeconomic effects of changes in fiscal policy, it is a good idea to “follow the money”. The funds have to come from somewhere to finance an increase in government expenditures and/or a decrease in taxes. Unless there is a net increase credit created by the Fed and the depository institution system, i.e., unless there is a net increase in thin-air credit, it is likely that the funds will come from the nonfinancial sector, which means that the stimulus to demand emanating from the government sector will crowd out spending that otherwise would have emanated from other nonfinancial sectors. And when government spending is cut and/or taxes increased, the fiscal drag on spending emanating from the government sector will crowd in spending from other nonfinancial sectors. Think of the extra money that otherwise would have gone to purchase government bonds “burning a hole” in the pockets of the otherwise lenders.
Lastly, for all you fiscal hawks out there, Chart 9 should make you jolly. Plotted in Chart 9 are quarterly observations of seasonally adjusted at annual rates (SAAR) net lending / net borrowing by the federal government (obtained from the Fed’s Financial Accounts report) as a percent of SAAR nominal GDP. The median percentage from Q1:1953 through Q3:2013 is minus 3.2%. In the second and third quarters of 2013, this ratio was minus 1.3% and minus 1.9%, respectively. With some help from the Fed in getting thin-air credit back up to a more normal post-WII growth rate in order to stimulate nominal GDP, some restraint in federal government spending along with some tax increases and voilá! the federal government deficit relative to the size of the US economy looks quite reasonable in an historic context.
I hope to send out my airing-of-grievances Festivus letter next week. I have a lot of problems with you people!