Please see the below article from Invesco received late yesterday afternoon:
- Covid-19 saw net savings of the US household sector surge to 20% of GDP
- However, this increase was offset by an increase in net borrowing by both the corporate and government sectors
- Is it plausible to argue that high savings will drive a strong economic recovery?
We believe the recovery following the Covid-19 pandemic will be stronger than, for example, the recovery following the GFC, and stronger than consensus forecasts. However, we do not sign up to the theory that this will be driven by excess savings.
High ‘savings rate’
A popular argument currently is that the relatively high ‘savings rate’ across advanced economies will provide powerful fuel for the economic recovery when the Covid-19 pandemic is finally under control.
On its face, this argument seems plausible; if people have forgone spending in the past and have accrued excess savings these can surely be spent in the economy when it is safe and legal to do so. We know that money has a fairly stable relation to income (known as the income velocity of circulation), but is the same true for savings?
There are two problems with the ‘excess savings will be spent’ argument; first, there is an empirical problem when the data is analysed and second there is a theoretical problem when the argument is considered rationally. Here, we focus on the US economy due to the availability of the data, but the analysis is equally as applicable to any economy.
First, we need to define which ‘savings rate’ we are talking about to be consistent. One can choose either an aggregate savings rate for the whole economy (either gross or net of capital consumption i.e. depreciation) or a sectoral (most often, household) savings balance.
An aggregate savings rate reflects overall investment rates (as opposed to consumption), whereas a sectoral savings balance reflects net borrowing and lending between sectors of an economy. These two different definitions are fundamentally different understandings of what ‘savings’ consist of and how they affect economic growth.
Both types of savings rates will be analysed empirically and theoretically.
The aggregate (or gross) savings rate
Gross savings rates have a moderately positive, statistically significant correlation with both real GDP growth and nominal GDP growth. Between 1950 and 2020 in the US, higher gross savings rates were associated with higher growth rates in both real GDP and nominal GDP.
Furthermore, net savings rates also show a moderately positive, statistically significant correlation with both real GDP growth and nominal GDP growth.
For example, an increase of 1% in the gross savings rate relative to GDP roughly corresponds to a 46 basis point (bps) increase in real GDP growth, and an 81bps increase in nominal GDP growth (measured on a percent change on previous year basis).
The rationale behind this relationship is simple; if a person or economy chooses to invest their capital (either directly or indirectly via financial markets) in productive endeavours instead of consuming goods or services, this increases the productive capital stock and boosts potential economic output.
The personal (or household) savings rate and sectoral balances
In stark contrast, personal (household) savings rates have zero to low positive, statistically significant correlation with both real GDP growth and, to a lesser extent, nominal GDP growth. Between 1960 and 2020 in the US, higher personal savings rates were not associated with higher growth rates in either real GDP or nominal GDP.
To understand why this is the case, consider Figure 1 which contains data taken from the US Financial Accounts.
Figure 1. US sectoral balances (% of GDP, 4-quarter moving average)
The chart tracks the net lending/borrowing between the four key sectors of the US economy, measured as a percentage of GDP (household, corporate – financial and non-financial — government, and external.
The household sector here represents personal savings; the corporate sector represents net lending/borrowing of corporations both financial and non-financial; for the government sector we show the fiscal balance (whether in surplus or deficit); and the overseas (or external) sector is represented by the current account balance (with the opposite sign to show capital inflows or outflows).
Net lending/borrowing across the whole economy (including foreign savings flows) must ultimately sum to zero, as any borrowing (a liability) must be matched by lending (an asset), and in double entry bookkeeping the two must balance.
When Covid-19 first hit the global economy in early 2020, the net savings of the US household sector rose to a historically high level of 20% of GDP. Simply put, spending by households collapsed as social restrictions were introduced and household savings rose commensurately.
Net savings vs net borrowing
However, this increase in net savings by households was reflected in an increase in net borrowing by both the corporate sector and the government sector. Corporates drew down credit lines in the dash-for-cash phenomenon experienced between March and May 2020.
At the same time, the government issued an incredibly large amount of securities to pay for the fiscal response to Covid-19, including unemployment benefits and public health spending relating to the virus.
To understand the effect on overall spending, whenever a particular sector increases its spending, we also need to know how this new spending is funded. If, for example, the government sector increases its spending, and this spending is funded entirely by less spending from the household sector (as was the case in 2020), overall spending is unchanged.
The same argument can be made for other sectoral balances. It is this key insight that undermines the argument that excess savings in the household sector will power the economic recovery.
As previously mentioned, there is nevertheless a relatively strong correlation between gross/net savings and real/nominal GDP growth. In contrast to personal savings, there has not been an increase in gross or net savings ratios, as evidenced in Figure 2.
Figure 2. US gross and net savings (% of GDP)
Since 1950, gross and net savings rates have closely tracked each other (albeit with a difference of about 16-17 percentage points) and have tended to fall as the US has developed into a more consumption-led economy.
The Covid-19 impact and recovery
Prior to the Covid-19 pandemic, gross and net savings ratios were stable at 20% of GDP and 3% of GDP respectively. Both gross and net savings ratios fell in the second quarter of 2020 to very low levels, which had only ever been reached or exceeded (on the low side) in the period following the GFC.
Since then, there has been a recovery in both savings ratios, but only to pre-Covid-19 levels of around 20% and 2.5% of GDP respectively. Therefore, there has been no material increase in gross or net savings during Covid-19, and the corresponding increases in real or nominal GDP growth cannot be expected to materialise.
In conclusion, analysing both aggregate savings and net sectoral savings yields little evidence that the economic recovery following the Covid-19 pandemic will be driven by excess savings.
We do believe, however, that the recovery following the Covid-19 pandemic will be stronger than, for example, the recovery following the GFC, and stronger than consensus forecasts. The stronger recovery will not be driven by excess savings, but by excess money balances relative to income (and technically by base effects also).
This will manifest firstly in strong real economic growth in 2021-22, and then, ultimately, in higher nominal growth. We believe this will also include a period of inflation in 2022-23 higher than the figures the US Federal Reserve is currently projecting.
Please continue to check back for our regular blog updates.