The PSR and private debt to GDP ratio are, indeed, negatively correlated (Chart 2). The correlation should not, however, be perfect since the corporate sector is included in the private debt to GDP ratio while the PSR measures just the household sector. We used the total private sector debt ratio because the household data was not available in the years leading up to the Great Depression.
The most important conceptual point concerning the divergence of these two series relates to the matter of the forgiveness of debt by the financial sector, which will lower the private debt to GDP ratio but will not raise the PSR. The private debt to GDP ratio fell sharply from the end of the recession in mid-2009 until the fourth quarter of 2013, temporarily converging with a decline in the saving rate. As such, much of the perceived improvement in the consumer sector’s financial condition occurred from the efforts of others. The private debt to GDP ratio in the first quarter of 2014 stood at 275.4%, a drop of 52.5 percentage points below the peak during the recession. The PSR in the latest month was only 1.7 percentage points higher than in the worst month of the recession. Importantly, both measures now point in the direction of higher leverage, with the PSR showing a more significant deterioration. From the recession high of 8.1%, the PSR dropped to 4.8% in April 2014.
Historical Record
The most recently available PSR is at low levels relative to the past 114 years and well below the long-term historical average of 8.5% (Chart 3). The PSR averaged 9.4% during the first year of all 22 recessions from 1900 to the present. However this latest reading of 4.8% is about the same as in the first year of the Great Depression and slightly below the 5% reading in the first year of the Great Recession.
In Dr. Martha Olney’s (University of California, Berkeley and author of Buy Now, Pay Later) terminology, when the PSR falls households are buying now but will need to pay later. Contrarily, if the PSR rises households are improving their future purchasing power. A review of the historical record leads to two additional empirical conclusions. First, the trend in the PSR matters. A decline in the PSR when it has been falling for a prolonged period of time is more significant than a decline after it has risen. Second, the significance of any quarterly or annual PSR should be judged in terms of its long- term average.
For example, multi-year declines occurred as the economy approached both the Great Recession of 2008 and the Great Depression of 1929. In 1925 the PSR was 9.2%, but by 1929 it had declined by almost half to 4.7%. The PSR offered an equal, and possibly even better, signal as to the excesses of the 1920s than did the private debt to GDP ratio. Both the level of PSR and the trend of its direction are significant meaningful inputs.
John Maynard Keynes (1883-1946) correctly argued that the severity of the Great Depression was due to under-consumption or over-saving. What Keynes failed to note was that the under-consumption of the 1930s was due to over-spending in the second half of the 1920s. In other words, once circumstances have allowed the under-saving event to occur, the net result will be a long period of economic under-performance.
Keynes, along with his most famous American supporter, Alvin Hansen (1887-1975), argued that the U.S. economy would face something he termed “an under-employment equilibrium.” They believed the U.S. economy would return to the Great Depression after World War II ended unless the federal government ran large budget deficits to offset weakness in consumer spending. The PSR averaged 23% from 1942 through 1946, and the excessive indebtedness of the 1920s was reversed. Consumers had accumulated savings and were in a position to fuel the post WWII boom. The economy enjoyed great prosperity even though the budget deficit was virtually eliminated. The concerns about the under-employment equilibrium were entirely wrong. In Keynes’ defense, the PSR statistics cited above were not known at the time but have been painstakingly created by archival scholars since then.
Implications for 2014-2015
In previous letters we have shown that the largest economies in the world have a higher total debt to GDP today than at the time of the Great Recession in 2008. PSRs also indicate that foreign households are living further above their means than six years ago. According to the OECD, Japan’s PSR for 2014 will be 0.6%, virtually unchanged from 2008. The OECD figure is likely to turn out to be very optimistic as the full effects of the April 2014 VAT increase takes effect, and a negative PSR for the year should not be ruled out. In addition, Japan’s PSR is considerably below that of the U.S. The Eurozone PSR as a whole is estimated at 7.9%, down 1.5 percentage points from 2008. Thus, in aggregate, the U.S., Japan and Europe are all trying to solve an under-saving problem by creating more under-saving. History indicates this is not a viable path to recovery. [reference: Atif Mian and Amir Sufi,. House of Debt, University of Chicago Press 2014]
Japan confirms the experience in the United States because their PSR has declined from over 20% in the financial meltdown year of 1989 to today’s near zero level. Japan, unlike the U.S. in the 1940s, has moved further away from financial stability. Despite numerous monetary and fiscal policy maneuvers that were described as extremely powerful, the end result was that they have not been successful.
U.S. Yields Versus Global Bond Yields
Table one compares ten-year and thirty-year government bond yields in the U.S. and ten major foreign economies. Higher U.S. government bond yields reflect that domestic economic growth has been considerably better than in Europe and Japan, which in turn, mirrors that the U.S. is less indebted. However, the U.S. is now taking on more leverage, indicating that our growth prospects are likely to follow the path of Europe and Japan.
With U.S. rates higher than those of major foreign markets, investors are provided with an additional reason to look favorably on increased investments in the long end of the U.S. treasury market. Additionally, with nominal growth slowing in response to low saving and higher debt we expect that over the next several years U.S. thirty-year bond yields could decline into the range of 1.7% to 2.3%, which is where the thirty-year yields in the Japanese and German economies, respectively, currently stand.
Van R. Hoisington
Lacy H. Hunt, Ph.D.
