A recent article presenting an essay by Jim Welsh contained a graph showing the declining marginal return to GDP of increasing debt over the past 42 years. Jim says the graph was supplied to him by a third party, who attributed the data to Antal Fekete.
Professor Antal Fekete
is an Austrian School economist associated, since 2005, with the Intermountain Institute for Science and Applied Mathematics. He retired from a distinguished career as a university professor in 1993. He has appeared recently as a guest on CNBC.
Prof. Antal Fekete argues that the creation of excess new money in the current situation is deflationary, because we have entered an era of negative marginal returns on added debt. In this condition, the more debt created, the lower the GDP growth (or the greater the GDP shrinkage).
The key to understanding the problem is the marginal productivity of debt, a concept curiously missing from the vocabulary of mainstream economics. Keynesians take comfort in the fact that total debt as a percentage of total GDP is safely below 100 in the United States while it is 100 and perhaps even more in some other countries. However, the significant ratio to watch is additional debt to additional GDP, or the amount of GDP contributed by the creation of $1 in new debt. It is this ratio that determines the quality of debt. Indeed, the higher the ratio, the more successful entrepreneurs are in increasing productivity, which is the only valid justification for going into debt in the first place.
Conversely, a serious fall in that ratio is a danger sign that the quality of debt is deteriorating, and contracting additional debt has no economic justification. The volume of debt is rising faster than national income, and capital supporting production is eroding fast. If, as in the worst-case scenario, the ratio falls into negative territory, the message is that the economy is on a collision course and crash in imminent. Not only does more debt add nothing to the GDP, in fact, it causes economic contraction, including greater unemployment. The country is eating the seed corn with the result that accumulated capital may be gone before you know it. Immediate action is absolutely necessary to stop the hemorrhage, or the patient will bleed to death.
The year 2006 was the watershed. Late in that year the marginal productivity of debt dropped to zero and went negative for the first time ever, switching on the red alert sign to warn of an imminent economic catastrophe.
The graph from Jim Welsh’s essay does not show the decline to zero, perhaps because the data in the graph is quarterly and the dip had too short a duration to show.
Plotted below is the marginal return data from the Jim Welsh article graph. The following table shows the nine occasions that a rise of more than 10% occurred in the marginal GDP return. This happened for all six recessions, as well as three times not closely associated with a recession. The first five recessions (1969 through 1991) all showed dramatic increases in the marginal GDP returns for increased debt. Since 1991, the 2001 recession showed a delayed and more gradual rise in marginal return; the current recession has not shown any significant response yet.
Perhaps the current recession will have an associated increase in marginal return. If it is delayed and gradual, as in the case of 2001, it could be taken as further evidence of the poor quality of debt. Click to enlarge
In the Welsh essay, a linear trend line is drawn that projects to zero marginal GDP return for an increase in debt, occurring in 2015. I have explored other trend lines and find that they all have similar correlation with the data, with R-squared values around 0.75, which is a good correlation, by my definition.
The linear trend line has an R-squared value of 0.75. One could just as well fit a quadratic trend line and obtain the graph below with an R=squared value not significantly different from the linear trend line:
Click to enlarge
If the quadratic trend line is extended, it does not go to zero, but asymptotically approaches about 12% in future years. This is shown in the next graph:
Click to enlarge
There has been a clear pattern of decreasing ability of increasing debt to grow the GDP over the past 40+ years. The use of debt has shown productivity improvement coming out of recessions, but the effect was weaker for the 2001 recession. We do not know yet what the result will be for the current recession.
I would speculate that the reason debt has greater productivity coming out of recessions is that the weakest users of capital have been “weeded out” by the recession and only the more efficient remain.
I would argue that if the trends of the last 40+ years continue, we are just as likely to asymptotically approach some low marginal return rate of return for added debt (quadratic functionality) rather than go to a zero productivity (linear functionality).
The question that is obvious is: What policy actions could be undertaken to change the declining trend of productivity of debt? One focus that I have discussed before is increasing the use of debt for developing the means of production of goods and services of utility and decreasing the use of debt to finance consumption and higher leveraged debt instruments.
These relationships involving the marginal productivity of debt are new to me and I hope to return with further analysis as I learn more.