What Happened to The Money Americans Borrowed Before the Crisis?
Just in time for the holidays, the US Federal Reserve released its quarterly Z.1 report. Recently retitled Financial Accounts of the United States (formerly Flow of Funds), the new data have provided a new, and perhaps final, chapter in the great financial crisis: the end of acute deleveraging. We take this opportunity to review the evidence and reflect on the nature of the great financial crisis.
The great financial crisis was not the result of pressures building up from decades of financialization. That process had been running for six decades prior the financial crisis without any event near the great financial crisis’s magnitude. Viewing household credit growth net of nominal GDP, as shown in Figure 1, it is easy to observe the dissimilarity of household credit relative to national income only in the final decade preceding the financial crisis. On a nominal dollar basis, household debt grew below the rate of national income until 2000.
The narrative to date was one of the impact of policy making being either inadequate or ephemeral in the face of the private sector deleveraging, which was supposed to be required to restore equilibrium.
The salient question is, What is the equilibrium value of household debt?
Household credit never eclipsed 20% of assets, as shown in Figure 2. It could be assumed that the violent contraction in this ratio furnishes us with an equilibrium at least between the minimum and maximum values displayed on this chart. It is difficult, however, to presume the notion that credit liabilities comprising just 20% of assets was so far away from equilibrium that it required violent reversion.
Prices of assets contain information on the expectations of buyers and sellers about the future economic value of assets. Liabilities taken out on those assets represent a derivative of this expectation of value — an expectation about the future expectation of economic value. This can be a self-fulfilling prophecy. Prices are set on the margin by these expectations, but the impact of change in price ripples into the assets that weren’t transacted by providing equity growth liquefiable by non-transacting liability growth. Consider a home purchase across the street from your home which raises the perception value of in your neighborhood. This by proxy raises the value of your home, and you can borrow against your new equity.
The information in the price of both the liabilities and the assets made the assumption that $1 of liability growth would result in more than $1 of GDP, which had been at a minimum true for most of the 50 years prior to 2000.
The credit crisis was a marginal utility of credit crisis in the preceding decade. The income generated from household credit fell beneath the growth of credit.
There was $4.5 trillion of GDP growth between 2000 and the end of 2007. Had the marginal utility of credit been on the low end ($1 of credit = $1 of GDP) of historical normalcy, GDP growth would have been 6.95% compounded annually, as opposed to the 4.75% it managed. The gap cumulatively represented $2.67 trillion of income, as seen in Figure 3, or $333 billion per year. This could be alternatively stated as follows: The credit crisis would likely not have manifested had nominal income growth been 46% higher between 2000 and 2008. It is easy to conclude that the valuation of fixed assets would have been justifiable had this income materialized.
Where did this lost income go?
Before this point, household credit growth can be seen keeping pace with net investment, as shown in Figure 4.
Where the household credit was previously circulated into forming real capital, despite the leap in investment in the mid-2000s, the gap between investment and household credit growth, which we will term “uninvested household credit growth,” was massive. This means the marginal propensity to invest collapsed.
We are left to speculate as to what households were doing with their marginal dollar of debt. Coincident to the decline in invested household credit growth was the surge in the trade deficit, shown in Figure 5. If investment using domestic service labor (such as homebuilding) has the highest multiplicative impact on the economy, low utility consumption of foreign production must have the highest dampening effect.
It could be proposed that the growth in uninvested household credit was spent in foreign production. Even if the borrowers were not directly the consumers of foreign goods, the effect in macroeconomic aggregates and accounts remains the same.
Why were the marginal debt dollars drained into foreign accounts?
There are probably many reasons for this, the most obvious of which is shown in Figure 6: Defense expenditures rose almost identically in proportion to the trade deficit during this period. Although most of these expenditures would be domestic (save for raw materials and energy), this displaced the same quantity of production and service capacity into output which was entirely non-consumable by households.
Igor Greenwald has suggested another cause could be found in the strength of the US dollar combined with (and contributing to) a less competitive US labor force.
The true root of the great financial crisis is embedded in the cause of the decline in marginal utility of household credit. As long as this measure remains positive, the rekindling of household credit will remain sustainable. Whether it remains positive relies on trade policy and the competitiveness of our labor.
If you liked this post, don’t forget to subscribe to Inside Investing via Email or RSS.
Photo credit: KeithBishop
You get a somewhat different story from your figure 1 if you use personal disposable income instead of gdp , i.e. (change in HH debt – change in DPI) divided by nominal gdp ( or nominal DPI ).
Debt growth was comfortably below growth in DPI up until the early ’80″s. There was a significant above-zero bump in the mid-80’s which is why you see a positive step-change in the graph of the HH-debt/DPI ratio at that time , though not as severe as the one that occurred in the 2000’s.
The other missing factor is that of distribution. HH debt/income ratios have climbed steadily since the 80’s for the bottom 95% of the income distribution , and not at all for the top 5%. Given the differentials in propensities to consume , this has consequences for consumer demand.