Leigh Harkness on debt saturation

In a follow up to his previous post please enjoy Leigh Harkness’s latest guest post on “Debt Saturation”.

Many years ago, I tried to identify the relationship between money and inflation. I could not find a general rule for all situations, but for certain countries who adopted “pure” float, I found that inflation was equal to the square root of the growth in bank credit over the growth in real GDP. That simple formula provided a reasonably close explanation of inflation in Australia for domestic products such as services, as shown in Figure 1.

Figure 1. Comparison of Australian Services Price Index and the modelled price changes

The formula also explained most inflation since 1995 for the US economy, as shown in Figure 2.

Figure 2. Comparison of US CPI and modelled prices

The relationship between bank credit, GDP and prices brought me to consider a much more pertinent issue: the relationship between the growth of bank debt and GDP. This relationship can be interpreted as the relationship between domestic debt and our capacity to service that debt. In Australia, the ratio of outstanding bank credit to nominal GDP is 2.7 times the level it was in 1984. Figure 3 compares bank credit, GDP at current prices and the ratio of bank credit to GDP for each quarter since 1980 relative to the level in December 1984.

Figure 3. Australia, bank credit and GDP

Figure 4 presents the same ratios for United States of America since 1985, using 1995 as the base year.

Figure 4. USA, bank credit and GDP

In my earlier blog, I explained how the current account deficit was equal to the growth of bank credit in countries such as the US and Australia. Foreign debt finances a large part of the current account deficit in these countries. Therefore, we can consider that the debt to GDP line also reflects the additional burden of servicing foreign debt. In Figure 4, it would appear that the ratio of bank debt to GDP has hit a ceiling in 2008. It was at that point the US economy experienced a major financial and economic crisis: the GFC.

If there is such a ceiling, it would mean that the growth of bank credit in the US would be constrained in future from increasing any faster than GDP. However, according to the formula for inflation under a pure floating exchange rate system, bank credit can only rise in proportion with GDP if there is no inflation. If bank credit in the US is to rise in proportion with GDP, then it will be rising at a much slower rate than it has in the past. In that case, we may see the US economy perform in a similar manner to the Japanese economy, with slow economic growth, slow growth of bank credit and large fiscal deficits intended to stimulate the economy.

Also, we could see US interest rates remaining low, as in Japan. As the debt burden would be constraining credit growth, it would not be necessary for the Federal Reserve to raise interest rates to constrain credit growth so as to constrain inflation. It is evident from Figure 3 that Australia has not yet reached its debt ceiling. The ratio of debt to GDP continued to grow in Australia, despite the GFC in the USA. Eventually (and possibly soon according to some of the posts on this site) the burden of the growing debt in Australia will reach or exceed the debt servicing capacity of economy. At that point Australia, too, will experience a debt crisis.

The Australian economy may just touch the debt ceiling, and then experience a decline in the rate of growth of bank credit. However, such a decline is likely to reduce the rate of economic growth and could cause a recession. A recession would reduce the capacity of businesses and people to repay their debts, and that could cause a financial crisis similar to that experienced in the US.

Latest posts by __ADAM__ (see all)


  1. Australia imports Chinas RMB credit converted to USD and we export dirt.

    Thank the PBOC monetary policy

  2. Hi Leigh, interesting analysis.

    You might also be interested in the relationship between money supply and house price inflation.

    Research has shown that the Australian median house price has not deviated by more than 14% from ten persons share of M3. In other words, there is a strong relationship between per capita M3, and house price growth. See here for details…

    Median House Price = 10x per capita M3

    This relationship has existed for about 50 years.



    • Shadow,

      Whilst not an expert, given that we live in a system of fractional reserve banking of course there will be a direct link between M3 and house prices.

      Whenever a bank creates a loan they also create a deposit adding to the stock of M3.

  3. Great article.

    Have you done any research on the “marginal productivity of debt” in Australia?

    I’ve seen some research on this measure for the US, showing that it crossed or was close to crossing the zero bound a year or two ago, after a slow decline over the last 30 years.

    It would be very interesting to see a graph of marginal productivity of debt in Australia.

  4. Leigh, a few questions from me:

    1) is your relationship definitely that inflation is EQUAL to or PROPORTIONAL to the square root of bank credit growth divided by real GDP?

    2) for the ” real GDP” term, how do you get that figure? Do you adjust nominal GDP for inflation? If so, doesn’t that make your relationship an iterative solution, where inflation is a function of inflation (as it is found on both sides of the relationship). Nothing wrong with that by the way – it just requires an iterative solution.

    3) which CPI data do you refer to here? Raw CPI, trimmed CPI, or other?

    It would be interesting to see your modeling using various types of inflation data.


  5. lets consider basic identity:

    G- GDP
    I-inflation (CPI)
    M- money supply (broadest, M3)

    when the only tool you have is hammer…(I mean “increasing money supply”)
    it has to hit somewhere. It could slow velocity, but under normal conditions velocity is quite inelastic, so expanding money supply either:

    1)increases GDP
    2)increases CPI
    or both

    in the long run sustainable GDP increase equals real productivity growth (fully monetized and privatized by issuing new credit to that extent). Anything in excess goes to increase “asset prices more than they otherwise would be” (equity, mate, wealth effect). When this spils over- it increases CPI as well.

    Call this “economic management”, set a target spil over rate (2-3%) and don’t question morality 🙂
    or even better- ask the banks to lend more to keep things going 🙂

    • The basic identity is MV = PT where T means transactions (not GDP) and P is price.

      Transaction include not only the purchases of products but the servicing of domestic and foreign debt.

      When you create one unit of M domestic debt increases by M and foreign debt increases by M; a total increase in debt for the economy of 2M.

      Increasing the relative size of debt service costs in the economy does not equate to sound economic management.

  6. You say “I found that inflation was equal to the square root of the growth in bank credit over the growth in real GDP.” Don’t forget the X-factor that economists keep looking for. Once they found X, they would start looking for Y.

  7. Interesting thesis Leigh.

    Not an expert on this, but how would these assertions play out if one puts this into context of something like a dynamic modelling as done by Steve Keen. That theory as per my knowledge serves me, also advocates a saturation point in terms of Debt/GDP ratios.

    One question: If one considers there to be a saturation point, should one draw conclusions as to the ratio being constrained in some manner as two variables are inherently inter-connected? If yes, does it then follow(in a very vague interpration) that inflation is somehow constrained not to breach some arbitary levels.

    Also Steve has advocated that change in change of debt provides a reasonablly predictive relationship for GDP. Would your graphs distill a similar relationship (with lagged time periods)?


    • Our high levels of income in Australia did not grow under a pure floating exchange rate system, reliant on bank credit. It was generated at a time when money was created from export incomes, and when the bulk of bank credit was supported mostly by national savings, not national debt.

      Floating the exchange rate so that bank credit became the only source of monetary growth has resulted in slower economic growth. Growth in GDP decreased and inflation has increased.

      Before the float, say from 1969 to 1983, real average weekly wages in Australia increased on average by 2% per annum. Since 1983, under our current monetary system, they have increased at an average rate of 0.7% per annum. This is despite massive improvements in technology and extensive growth in mineral exports that should have given us substantially higher real incomes.

      In the US, the real wages of workers are still 12% lower than they were when the US floated their dollar in 1973.

      I am not saying that we need to fix the exchange rate but we have major problems in our monetary system. It is unstable and unsustainable. It is not capable of providing the appropriate market signals to sustain an industrial economy. If our economy were an aeroplane, it would not be able to take off.

      We have been fortunate that our extensive mineral resources have allowed us to increase real incomes despite the growth of our debt. But I am reluctant to say that an increase in debt will increase our real incomes.

      If you are interested to look further at this topic, you may like to look at my paper on debt and unemployment in America available at http://www.buoyanteconomies.com/EconomicProblemsOfUS.htm .