It is widely believed that TARP and the Fed, through purchases of nearly $2.5 trillion of non-performing loans, mortgages, Treasury bonds and other assets (collectively termed Quantitative Easing [QE]) prevented economic collapse and lowered the cost of debt financing by lowering interest rates. Fed Chairman Bernanke explicitly stated that the objectives of QE are a wealth effect on consumption from higher asset prices and lower interest rates intended to spur investment and indirectly increase employment. At the same time, the Fed remains committed to managing to an approximate 2% inflation target while expanding the Fed's asset purchase program at a rate of $40-$85 billion per month until the unemployment rate reaches a satisfactory level.
This article shows that the actual effect on the economy slows real wage and employment growth, thus restraining GDP growth and negates a wealth effect through the following mechanism:
- GDP is a return on the economy's total asset base.
- Wherein capital and other asset formation normally occurs in proportion to GDP growth and where growth is the required return on capital formation.
- The long run return on capital and labor is equivalent.
- QE has created an abnormal (excess) supply of new capital which necessarily lowers its available real return from GDP growth. This requires that real wage growth also stagnate because low capital returns cannot accommodate real wage growth. Stagnating real wage growth, normally equal to productivity growth, depresses real GDP growth.
- Worse, the 2008-09 recession should have resulted in deflation as both the non-performing assets and the associated false real wage growth were wrung out of the economy. Fiscal and monetary policy prevented price and wage deflation, leaving a wage level that cannot accommodate full employment. The consequences are impaired GDP growth and high structural unemployment and fiscal deficits.
- QE has caused low real yields on bonds, thus reducing expenditures from interest income, and causes an under-valuation of the stock market by forcing returns normally expected to come from earnings growth to derive from dividend yield: see my Seeking Alpha article here. This is an anti-wealth effect, and the opposite of the Fed's stated goals.
- The negative real yields may impair, not spur investment if investment is made in expectation of a required real return rather than merely low real cost of capital.
- Finally, the Fed cannot simply "exit" QE because selling bonds to mop up the liquidity it created still leaves the excess capital in the form of new bonds on which the government must pay interest and principal through further money creation.
QE is a purely nominal attempt to solve a real economy problem. Its intended effects have not been proven before its launch: e.g. Is investment better spurred merely by low interest rates, or by adequate prospects for real return? And: Have QE, TARP and unemployment benefits policies maintained real wages instead of allowing them to fall with GDP, and thus prevented a return to full employment?
GDP: a Return on Assets
US GDP is reported as $15.6 trillion at the 2nd quarter of 2012 nominal run rate. If real GDP were growing at a normal 3% (the sum of 1% population growth and 2% per capita productivity growth), it would grow about $480 billion. (Coincidentally, this is the same figure as the volume of announced annual Fed QE3 mortgage purchases at $40 billion per month.)
But how does GDP arise? From labor and all of the other assets that enable and extend its productivity. If GDP itself is viewed as a return on assets, what can be surmised about the size of the asset base? One method would be to capitalize GPD by its long term nominal growth rate of about 5.2% (3% real and 2.2% or so inflation). This results in an estimate of about $300 trillion.
Very few attempts have been made to assess the assets of the U.S. economy. Dr. Rutledge provides a partial tally of U.S. assets as of 2008. The United Nations also attempted a tally. A more complete accounting is attempted below.
The Fed Funds Flows report provides a partial balance sheet of the US economy. I have calculated other major asset classes while trying to avoid double counting of debt and equity:
- Total credit market debt of: $55 trillion (credit is both an asset and liability)
- Total Stock Market (Equity) value of: $25 trillion
- Labor: Compensation to employees is running at about $8.6 trillion pre-tax. If this is valued at the average of the SP500 forward P/E of 13.4 and the inverse 30-year Treasury yield valuation, the resulting value of labor is approximately $204 trillion. Since compensation is an earnings stream, valuing the earnings stream to labor should be akin to valuing a security that is more stable in earnings than the stock market but a bit more risky than a long term Treasury.
- Intellectual Property: Now about $11 trillion (very little of this is thought to be reflected in stock prices to the extent IP is owned by corporations)
- Public Infrastructure: $2 trillion
- Foreign Direct Investment In Production: $3 trillion net of securities and stock holdings already counted above
- Non-Corporate Property Plant and Equipment owned free and clear of debt: $10 trillion at least from Fed Funds Flows balance sheet data
This totals about $310 trillion of assets deployed in the US economy (with some stock market value and debt related to foreign assets generating foreign GDP and vice-versa). Normal nominal GDP would run about 5.2% (3% real and 2.2% inflation) which, if applied to the total asset base, produces a calculated nominal GDP of $16.1 trillion - precisely the current run rate heading into the third and fourth quarters of 2013. Note that labor's share of the asset base is about 2/3 and capital about 1/3.
GDP Growth, Capital Formation, Return on Capital and Labor
GDP growth is a function of increase in population engaged in productive activity and the rate of per capita productivity growth which is enabled by the non-labor assets of an economy - principally those listed above. Normal real GDP growth of 3% (nominal 5% to make the math easy) would result in asset growth wherein GDP growth represents an adequate expected return on new assets: a $16 trillion economy growing at a 5% nominal rate would produce an increase in nominal assets valued at about ($16 trillion x 5% growth)/5% nominal return on assets = $16 trillion.
An empirical validation can be derived from Fed Funds Flow data. From 2004-5 for example, GDP grew $.74 trillion. Total credit outstanding grew $3.4 trillion and stock market value increased $1.7 trillion for total financial capital growth of $5.1 trillion. Total expected asset creation would be $.74 trillion capitalized at a 5% long term GDP growth rate or $14.8 trillion. The growth in capital (about 1/3 of total assets) was in line with predictions. This rule holds nicely over time as well. Returns on capital and labor remain equivalent through arbitrage between them.
Effect of QE on Interest Rates and Inflation Expectations
QE has demonstrably raised nominal interest rates by increasing inflation expectations while forcing negative real rates. How does this phenomenon affect wealth creation and resulting spending, capital investment, job formation and real wage growth?
The chart below compares the expected 5-year real return from TIPS, change in real GDP and expected inflation computed as the difference between 5 year nominal and TIP Treasuries.
click to enlarge images
Clearly, deflation was expected by markets in late 2008 while real rates measured by 5 year TIPS remained above 1% during the latter ¾ of the official recession - indicating that the recession itself was not the cause of the subsequently negative real rates which did not occur until after GDP growth had converged to a real 2% trend.
QE alone dramatically changed market expectations. The Cleveland Fed provides the below interactive chart of the Fed's balance sheet level and composition.
The TARP and Fed asset purchases accomplished at least three things: 1) the fall in real GDP was mitigated by preserving the capital that would and should have been devalued due to lack of return, and added new capital (the asset purchase capital); 2) expected deflation was reversed to inflation; and 3) the real expected rate of return/interest was driven negative.
In terms of marginal capital creation in the economy, the TARP and QE asset preservation and new capital (e.g. reserves) creation activity should be looked at in relation to normal, market-driven capital formation. Between TARP and QE, $2.5 trillion of assets where purchased with newly created money, mostly in the form of bank reserves. An argument can be made that almost all of the $2.5 trillion in purchased assets are so impaired in prospective return that no private market exists for them that would not require a discount so great that it would bankrupt its original owners - mostly banks. Charitably, I assume that the market value of these assets is ¼ of its face value or about $.6 trillion.
Thus, net of an assumed capital write down of at least $1.9 trillion (much more had not TARP and the Fed offered its backstops) total capital creation was $4.4 trillion (the sum of $2.5 trillion in new money and the $1.9 trillion asset write down that was prevented). To put it another way, instead of markets eliminating at least $1.9 trillion of unproductive capital, government entities stepped in, creating a net $2.5 trillion addition to capital.
The $15.6 trillion economy growing at a real 2% (4% nominal) or $.62 trillion, would generate about $12.5 trillion in new assets - 1/3 or $4.1 trillion of that financial. Had non-performing assets been written off, the purely financial capital portion would have been reduced by at least $1.9 trillion instead of an excess $4.4 trillion. Normal return on capital would be 1/3 of the $.62 trillion nominal GDP growth/$4.1 trillion in capital or a 5% pre-tax return and 4% after-tax, leaving a 2% real return. The excess capital drives the pre-tax return to 2.4% and the after-tax real return negative. The shorter the maturity of capital, the deeper the negative real yield or expected return.
The chart below shows the cumulative effect of excess capital on the real return on capital in the economy. The second line shows the amount of GDP growth available as a return on capital. The yellow line shows the total excess capital as defined by the sum of TARP and the change in the Fed's balance sheet since before QE as well as the estimated capital write down that was not taken due to QE. The green line shows private capital formation and the red line shows the real pre-tax return on capital. As applicable, the Fed Funds Flows schedule from which data are taken is shown at the far right.
Comparing the first and fourth charts (below) the negative real yield as a function of maturity is easy to see with even long-term 10-20 year TIPS yields below zero; and fully agrees with the calculated real return on new and excess capital formation provided above.
The Fed's new QE initiative adds to excess capital at $85 billion per month through the Fed's $40 billion monthly mortgage and mortgage-backed security purchases and operation "twist" long term Treasury purchases which at the high end is about 25% more than a 4% nominal growth economy would generate in financial assets.
Real Economy Consequences of QE
If investment growth is a function of some required real return instead of merely low real cost of debt capital, then QE most certainly impairs capital formation; and by extension, related employment. Gross private fixed capital formation, chart below, remains at historically low levels in relation to GDP.
QE has not reduced the nominal cost of capital (long term interest rate) by driving up actual and expected inflation (below chart), but has definitely eliminated and driven negative the actual and expected real return. The fundamental question is: does a negative expected real return on marginal capital cause more investment and thus hiring than a positive return? So far, there is no evidence of this.
Another predicted depressant effect is on real wages and employment. Normally, real per capita GDP growth and disposable income track closely; and indeed must. Only since the 2008 recession has the former lagged significantly below disposable income. This of course is incompatible with return on capital which comes from the spread between real GDP and real wages - unless reduced employment is the source of the needed cost reduction.
The chart below shows real per capita GDP materially lagging below real disposable income for the first time since at least 1960 - because real wages were not allowed to fall. It also shows real disposable income stagnating - both as predicted.
Furthermore, on a per hour worked basis, real compensation is stagnating since the recession, average hours worked per employed person are falling and real GDP per hour worked is rising in an effort to close the return on capital gap. Employment in terms of both hours worked and real compensation are suffering.
Even more concerning, the civilian labor force participation rate among those aged 16-64 is plummeting to 32 year lows; causing understatement of unemployment rates. It is notable that prior to 1980, much of the low participation rate was caused by women in this age range gradually entering the workforce. The plunge in the participation rate happened after the recession and the onset of QE while real GDP grew.
QE will continue to hold real yields low. TIPS yields will remain depressed in order of term structure - the shorter the term the lower (or more negative) the yield. Gold and oil are both held up by low real yields on fiat instruments such as stocks and bonds. GDP growth will be far below its 3% potential as both real wages and labor participation rates are depressed by the low return on capital created by the Fed. This means low stock returns capped by low top-line driven EPS growth and depressed valuations stemming from low real yields.
This is currently a global scenario with the US, EU, China and English economies all "supported" by historically aggressive monetary policies with China adding massive fiscal projects to the mix. The result is employment and real wage attrition and high running deficits, low growth and knife-edge recession risks uncushioned by healthy employment rates and government balance sheets.
A full discussion of an alternative solution is beyond scope for this article. The principal elements would include:
- Stop doing what causes harm i.e. QE
- Reverse QE in a controlled fashion, impacting the entities and persons which originated and owned the bad assets while protecting savings and liquidity. Restore the real return on capital.
- Permit market-driven price and wage deflation. Note that in deflation, the real purchasing power of a dollar of income rises, which mitigates the real effects of wage deflation.
- Lower the tax rate on profit from marginal investment which results in employment for an extended period. Lower tax rates on dividend and interest income which raises asset prices and causes a spending wealth effect.
- These actions may lower nominal interest rates while increasing real rates as inflation expectations fall and the balance of capital to GDP growth is realigned.
The actual effects of QE appear to be the opposite of stated goals in that QE appears to have a depressing effect on asset prices, real wage and normal employment growth. The latter two restrict GDP growth and thus employment growth and asset returns.