One of larger banks to have undergone a substantial downshift in its outlook on the market and the economy, is Bank of America, which continued the dour mood this morning when in note by its economist Ethan Harris it again pointed out that US consumer spending has “settled at a lower trend.”
Here are the details:
Consumer spending has settled on a lower trend: Consider these three facts: (1) household wealth is nearly back to peaks during the housing bubble as a percentage of disposable income; (2) interest rates are at historical lows; and (3) initial jobless claims have touched the lowest level since 1973. That looks like a recipe for spectacular consumer spending. But, yet, real consumer spending has only averaged a 2.0% annual rate since the recovery started, which is nearly half the pace of the prior 10-year average. What happened? The recovery in wages has been slow and the savings rate has increased. The 4-quarter moving average of the personal savings rate (derived as disposable income less outlays) has reached 5.9% over the last two quarters. An alternative measure of the savings rate, which is measured from household net worth, has seen even more impressive gains. As we argue, the savings rate is likely to remain elevated, limiting the potential upside to consumer spending due to a variety of cyclical and structural reasons.
While spending has declined, savings – that bogeyman of Keynesian economics – continue to rise. As BofA writes, there is a long list of reasons for consumer spending to have undershot the growth in disposable income, boosting the savings rate.
The bank addressed the top theories, starting with the cyclical factors (that are potentially reversible) and ending with the structural.
Below is BofA’s full breakdown of these 5 secular factors:
A Model Consumer
We can illustrate the first two factors using a model of consumer spending. At its simplest, consumer spending should be a function of growth in disposable income and changes in net worth. Using the specifications in Pistaferri (2016), which was presented at the most recent Boston Fed economics conference, we estimate real per-capita growth in consumer spending as a function of real per-capita disposable income and net worth, through 3Q2007. As we show in Chart 3, the “base model” predicts a higher trajectory of consumer spending than the realized rate. In other words, actual consumer spending was weaker than the model had expected based on the historical relationship between consumption, income and wealth prior to the Great Recession. Based on this model, real per-capita consumer spending should have increased an average of 2.2% yoy over the recovery (since 3Q09) versus the actual gain of 1.2% yoy.
Taking it a step further and adding in a one-quarter lag of leverage, defined as the ratio of debt to disposable income, improves the fit, but still leaves a more robust trajectory for consumer spending. Following Pistaferri, we add in a one-quarter lag of consumer confidence: now the model closely tracks the actual data. This highlights the importance of diminished animal spirts in containing consumer spending and boosting savings. The good news is that these factors are partly cyclical. As we move further from the crisis, consumers will be increasingly influenced by current trends in income, wealth and debt levels. All else equal, this would support stronger consumption and lower savings.
The consumer over the lifecycle
The challenge is that there are strong structural forces pushing up savings, countering the potential cyclical improvement. As Vice Chair Fischer addressed in his recent speech to the Economics Club of New York, the aging of the population naturally increases household savings. Turning back to our macroeconomic classes, the basic assumption for the consumer is that he or she wants to “smooth” spending over his or her lifetime based on actual and expected income. This will result in higher savings when income is high relative to the average of the lifecycle (otherwise known as permanent income) and lower savings when there is a shortfall in income.
Income tends to peak as people approach retirement, but this is also a period when they have a high propensity to save. We illustrate the typical behavior in Chart 4, which shows that savings increases as individuals approach middle-age through retirement. The Baby Boomers (born between 1946 and 1964) are aged between 52 and 71, which is a period of heightened savings. Of course, as these boomers continue to age, their saving rate should dip a bit and could eventually curtail the upward pressure on the saving rate.
A structurally higher savings rate contributes to lower trend growth and reduced longterm rates. Indeed, according to research from the Federal Reserve Board and outlined in Vice Chair Fischer’s latest speech, the increase in the savings rate due to population aging can account for a 75bp reduction in the equilibrium fed funds rate relative to its level in the 1980s. This compares to the 120bp simulated drag from a decline in potential GDP growth and another 60bp drag from under-investment in capital. Of course, we have to be careful reading too much into point estimates, but the simulation illustrates the relevance of the increase in the savings rate to the future path of economic growth and interest rates.
Other stories: safety nets and inequality
Another driver of higher saving is that households are likely increasingly aware of the swelling budget deficit in the US, implying future tax hikes, entitlement cuts or both. It has been hard to ignore the shutdown of the government, debt ceiling showdowns and the fiscal cliff. These events have raised awareness of the difficulties in balancing the budget and supporting future entitlement programs. Moreover, low interest rates and concerns about the health of pensions further contribute to higher savings.
The rise in income and wealth inequality has also contributed to the upward trajectory in the savings rate. Those in the top tier of the income and wealth distribution tend to have a higher propensity to save given fewer budget constraints through the life cycle. As we see in Chart 5, there has been a shift in income toward the top 5% of the income cohort, which boosts aggregate savings rates.
The most important drivers for consumer spending are income and wealth creation. An increase in both will naturally drive spending higher. However, there are limits to the upside for consumer spending given the structural changes prompting greater savings. It is this structural limit that the world’s central banks – and soon governments – continue to fight against.