Steve Keen's Debtwatch No 14 October 27 Draft only Steve Keen's DebtWatch No. 14 October 27 An unprecedented level of debt... and its aftermath? Australia enters the 27 election with the highest level of private debt in the nation's history, and with a debt burden that has been exceeded only twice before--in the Depressions of 189s and the 193s. The only two periods in Australia's history that can compare with today are the Melbourne Land Boom of the 188s, and the Roaring Twenties. But the private debt levels reached during those periods of irrational exuberance pale into insignificance alongside the 17% level of today (see Figure 1). Long Term Figure 1 16 14 12 Debt Ratio Trend 1964-Now Trend 188-1892 Trend 192-1932 Debt to GDP: The Long Term View 1 8 6 4 2 188 189 19 191 192 193 194 19 196 197 198 199 2 21 Debt's impact depends on the rate of interest as well as the debt level. As is well known, nominal interest rates reached record levels during the 199s (see Figure 2). Figure 2 Page 1
Steve Keen's Debtwatch No 14 October 27 Draft only 2 Nominal Interest Rates 1 1 188 189 19 191 192 193 194 19 196 197 198 199 2 21 Those historically high interest rates shuffled the debt burden deck, so that the relatively modest debt level of 8% in 199 actually resulted in the highest recorded nominal interest payment burden in Australia's history (Figure 3). 2 Figure 3 Nominal interest payment burden 1 1 188 189 19 191 192 193 194 19 196 197 198 199 2 21 However, there's a second joker in the debt pack: the rate of inflation. Inflation has varied radically in Page 2
Steve Keen's Debtwatch No 14 October 27 Draft only Australia over the last 1 years, from a low of minus 1% in the 189s to a high of 2% in 192 (Figure 4). Figure 4 3 Inflation 2 1-1 - 2 188 189 19 191 192 193 194 19 196 197 198 199 2 21 Conventional economists treat inflation as always and everywhere a bad thing, but in some circumstances, inflation makes life easier. If inflation is high, then an apparently crippling rate of interest is actually fairly easy to service, since inflation drives up incomes faster than the interest rate increases debt. The "real interest rate" is then negative--it's as if you're being paid to borrow money. Discussing such a hypothetical might sound like like discussing the flight plan of a squadron of pigs, but in fact it has happened twice in recent memory--during the 197s, and also from 1946 till 193. Conversely however, falling prices can turn a low nominal rate of inflation into a crushing debt repayment burden. This was why the Great Depression was so horrific, because prices fell by over ten percent per annum in 1931. The inflation adjusted (or "real") rate of interest is thus radically different to the nominal rate. In 192, when inflation hit 24% (courtesy of the Korean War), the real interest rate was minus 2 percent. The real rate of interest in the 189s, on the other hand, was 22 percent: far higher than the 12 percent level of 199 (Figure ). Figure Page 3
Steve Keen's Debtwatch No 14 October 27 Draft only 3 Inflation-adjusted "Real" Interest Rate 2 1-1 - 2-3 188 189 19 191 192 193 194 19 196 197 198 199 2 21 Put all these influences together, and the key indicator of the impact of private debt on the economy is the inflation-adjusted (or "real") ratio of interest payments to GDP. This shows that, while 199 was bad--with a real interest repayment burden of 1 percent--it isn't nearly as bad as today's real interest burden of 12 percent. The only times when the real interest payment burden was greater than it is today was during the Depressions of the early 189s, and 1931--because then, falling prices dramatically increased the burden of relatively modest nominal interest rates (of 8 and 4 percent respectively). Today's debt burden is the heaviest non-depression level Australia has ever experienced (Figure 6). Figure 6 Page 4
Steve Keen's Debtwatch No 14 October 27 Draft only 2 Inflation-adjusted interest payment burden 1-1 188 189 19 191 192 193 194 19 196 197 198 199 2 21 Figure 7 shows the changes over the last seven years, and explains why mortgage stress has become such a key election issue. With the increase in debt levels in the last seven years--from 1% of GDP in 2 to 17% today--and the increase in interest rates (from 8.4 to 9.6%), the effective debt burden has risen by over 7% since 2, from 6.9% to 11.8%. That's 24% above the burden during "the recession we had to have". It's no wonder that the electorate isn't feeling "relaxed and comfortable", despite 17 years of economic growth. Figure 7 1 Inflation-adjusted interest payment burden 1 2 22 24 26 Note: The above research is based on data supplied by the RBA from Ric Battellino's speech, and RBA Research Discussion Paper RDP1999-6 "Two Depressions, One Banking Collapse", by Page
Steve Keen's Debtwatch No 14 October 27 Draft only Chay Fisher and Christopher Kent; and Table PC3-31, page 214 of Australian Historical Statistics (Fairfax, Syme and Weldon Associates, 1987), edited by Wray Vamplew. Subsequent tables and charts are derived from the RBA Statistical Bulletin. The 189 and 193 peaks shown here exceed those in Battellino's speech, because his graph only showed bank credit prior to 193; I have imputed total credit from Fisher and Kent's research into credit shares. Aggregate Data and Trend Growth Rates Debt yet again rose faster than GDP last month, and the ratio is now edging towards 16 percent (see Table One). However, though not too much should be made of a monthly figure, the rate of growth of lending to households has slowed substantially--and for the first time since early 24, the level of personal debt actually fell. (see Table Two). Table One: Aggregated Debt Summary Table One D 1 = 1 2 3 4 6 7 8 9 1 11 12 13 14 1 1 2 "Summary" "Total Private Debt" "Nominal GDP" "Date (levels)" 27.8 27. "Levels ($m)" 16889 1478 "Change Month $m" 12917 692.2 "Change Month %".79.67 "Change $m" 223166 7933 "Change %" 1. 8.18 "Since 199" 8.6.4 "Since 198" 11.98 7.93 "Since 1964" 13.48 9.42 "Date (% GDP)" 27.8 "N/A" "As % of GDP" 17.3 1 "Change Month".13 "N/A" "Change " 6.76 "N/A" "Since 199" 2.9 "N/A" "Since 198" 4.1... Table Two: Disaggregated Debt Summary The rate of growth of mortgage debt also fell to be no greater than the rate of growth of output--again, something that hasn't happened since early 24. While it is obviously too early to tell, this may be the first signs of the long overdue slowdown in household borrowing. Business borrowing, on the other hand, continues to accelerate, which is no bad thing--at least, not if it mainly indicates rising investment for the China boom, rather than M&A activity. Business debt is now a greater proportion of both GDP and Gross Operating Surplus than it was during the 199s--see Figures 3 and 7 below. However, business's debt servicing burden is substantially lower than in 199, since rates are less than half what they were then. Nonetheless, the combination of still-rising aggregate debt and higher interest rates has pushed the aggregate interest payment burden on the economy past the 1% of GDP mark--a level previously only exceeded between March 1989 and July 199. Though the effects will take several months to filter through, next month's data should nonetheless start to show the impact of the recent turmoil on financial markets. They should make interesting reading. Table Two Page 6
Steve Keen's Debtwatch No 14 October 27 Draft only D 2 = 1 2 3 4 6 7 8 9 1 11 12 13 14 1 2 3 "Detail" "Business" "Mortgage" "Personal" "Levels ($m)" 628 88178 148299 "Change Mth $m" 761 799-491 "Change Mth %" 1.23.66 -.33 "Change Yr $m" 12237 9787 23 "Change Yr %" 19.4 12.49 18.41 "Since 199" 4.88 14.71.42 "Since 198" 1.61 14.3 1.44 "Since 1976" 11.1 14.31 11.23 "As % of GDP" 9.67 83.78 14.9 "Change month".7.1 -.98 "Change year" 1.36 3.93 9.4 "Since 199" -.83 9.2 -.42 "Since 198" 3.1 6.2 2.62 "Since 1976" 3.7.77 2.99 D 2 = 1 2 3 4 6 7 8 9 1 11 12 13 14 Debt to Income Ratios Table Two 1 2 3 "Detail" "Business" "Mortgage" "Personal" "Levels ($m)" 628 88178 148299 "Change Mth $m" 761 799-491 "Change Mth %" 1.23.66 -.33 "Change Yr $m" 12237 9787 23 "Change Yr %" 19.4 12.49 18.41 "Since 199" 4.88 14.71.42 "Since 198" 1.61 14.3 1.44 "Since 1976" 11.1 14.31 11.23 "As % of GDP" 9.67 83.78 14.9 "Change month".7.1 -.98 "Change year" 1.36 3.93 9.4 "Since 199" -.83 9.2 -.42 "Since 198" 3.1 6.2 2.62 "Since 1976" 3.7.77 2.99 Debt to Income Ratios Debt to GDP (D2 & G12) Figure 1 Page 7
Steve Keen's Debtwatch No 14 October 27 Draft only 1 Australian Private Debt to GDP 1 196 198 2 s Debt to GDP Regression Figure 2 Australian Private Debt to GDP 1 Debt to GDP Ratio Exponential Fit from 1964 1 196 198 2 Debt Components to GDP Figure 3 Page 8
Steve Keen's Debtwatch No 14 October 27 Draft only 1 8 Components of Australian Debt Mortgage Business Personal 6 4 2 198 199 2 Monthly Growth Rates 6 4 All Debt Business Mortgage Personal Debt Monthly Growth Rates 2 2 26 27 Date ly Growth Rates Page 9
Steve Keen's Debtwatch No 14 October 27 Draft only Debt ly Growth Rates 2 1 All Debt Business Mortgage Personal 2 22 24 26 Date Ratios ly Growth Rates 3 2 Debt Ratios ly Growth Rates All Debt Business Mortgage Personal 1-1 199 2 2 Date Debt to Household Disposable Income Figure 4 Page 1
Steve Keen's Debtwatch No 14 October 27 Draft only 2 1 Household Debt to Disposable Income Mortgage Personal All Household 1 198 199 2 Mortgage Debt to Household Disposable Income Figure 1 Mortgage Debt to Household Disposable Income 1 198 199 2 Debt to Household Disposable Income (the big jump in personal and fall in business debt in 1989 was due to a change in bank classifications of debt types that caused a proportion of business debt to be reclassified as personal). Page 11
Steve Keen's Debtwatch No 14 October 27 Draft only Figure 6 2 Personal Debt to Household Disposable Income 2 1 1 198 199 2 Business Debt to GOS Figure 7 16 Business Debt to Gross Operating Surplus 14 12 1 8 6 198 199 2 Page 12
Steve Keen's Debtwatch No 14 October 27 Draft only Investment Percent Total Housing Lending Housing Finance Analysis Figure 8 "Investor" Percentage of Total Housing Lending 4 GI1 1 3 2 1 199 2 GI1 Construction Percent Total Housing Lending Figure 9 Page 13
Steve Keen's Debtwatch No 14 October 27 Draft only 3 Construction Percentage of Total Housing Lending 2 1 199 2 Investment Construction Percent Total Housing Lending Figure 1 2 Investor Construction Percent of Total Housing Lending 1 1 199 2 Construction Percent of Investor Lending Figure 11 Page 14
Steve Keen's Debtwatch No 14 October 27 Draft only 8 Construction Percent of Investor Lending 6 4 2 199 2 Personal Finance Analysis Figure 12 Credit Card Data Page 1
Steve Keen's Debtwatch No 14 October 27 Draft only 1 Card Limits Card Balances Card Transactions Credit Cards To GDP 199 2 Figure 13 Credit Card Data Credit Cards Usage 3 2 1 199 2 Balances To Limits Transactions To Limits Figure 14 Credit Card Repayments Page 16
Steve Keen's Debtwatch No 14 October 27 Draft only Credit Card Repayments 2 1 Average Rate 199 2 2 Debt components to Income Figure 14 2 1 All Debt Business Household Trend All Trend Business Trend HH Trends in Private Debt 1 198 199 2 Page 17
Steve Keen's Debtwatch No 14 October 27 Draft only Debt to GDP Trends Figure 1 Debt to GDP Ratio and Trends 1 Data Trend since '64 Since '94 Last 1 197 198 199 2 21 Debt to GDP Exponential Growth Correlation Ratios These tables show the approximate exponential rate of growth of debt from various starting dates, and the correlation coefficient between this exponential approximation and the data. The correlation is staggeringly high, especially for a data series which, from an equilibrium point of view, should have no trend, or at worst should move in the opposite direction to changes in the official rate of interest--thus keeping the debt repayment burden constant. Table Three: Exponential Growth Rates & Correlations since 1964 & 1977 Corr77 = 1 2 3 4 "Debt ratios" "All" "All" "Business"Household""Mortgage" 1"Start Date""mid-1964" 1977 1977 1977 1977 "Growth 2 rate" 4.17 4. 3.9.7.77 3"Correlation" 99.11 98.43 73.47 98.1 98.11 Table Four: Exponential Growth Rates & Correlations since 199 Corr9 = 1 2 3 4 "Debt ratios" "All" "Business" "Household" "Mortgage" 1 "Start Date" 199 199 199 199 2"Growth rate" 2.8 -.97 6.81 9.32 3 "Correlation" 96.4-17.37 99.67 99.76 Page 18
Steve Keen's Debtwatch No 14 October 27 Draft only Debt to GDP Linear vs Exponential Regressions 1 Figure 16 Australian Private Debt to GDP 1 Debt to GDP Ratio Exponential Fit from 1963 Linear Fit 196 198 2 Debt Servicing Burden Interest Rates & Payments 2 Figure 17 Interest Rates & Interest Burden 1 1 Average Rate Interest % GDP 196 197 198 199 2 Interest Payment Trends If trends in debt growth continue, then even without any increases in official interest rates, the interest repayment burden on the economy will exceed that of 199 sometime between September 28 and September 29. Page 19
Steve Keen's Debtwatch No 14 October 27 Draft only.2 Figure 18 Interest Repayment Burden Trends.1.1. Debt Servicing by Loan Type 1 Last From '94 198 199 2 21 Figure 19 Debt Servicing Burden Mortgage Personal Business 1 198 199 2 Page 2
Steve Keen's Debtwatch No 14 October 27 Draft only Household Debt Servicing Figure 2 1 8 Household Debt Servicing Burden Mortgage Personal Total 6 4 2 198 199 2 Figure 21 2 1 Household Debt Servicing Burden Mortgage Personal Total 1 198 199 2 Page 21
Steve Keen's Debtwatch No 14 October 27 Draft only It's obvious why high interest rates prior to 199 brought the economy to a standstill when one sees the following graph: the interest servicing charge on business loans peaked at almost 3 per cent of Gross Operating Surplus. Even though business debt has recently started to rise as a proportion of GDP, the debt servicing burden remains in the range that applied in the early 198s. Figure 22 3 Business Debt Servicing Burden 2 1 198 199 2 The debt repayment burden is affected by both the rate of interest, and the level of debt. This chart shows the percentage of GDP that is required to pay the interest on outstanding debt, as a function of average interest rates (the vertical axis) and the debt to GDP ratio (horizontal axis). We are approaching the pain threshold that applied back in 199, when debt servicing consumed 16.7% of GDP. The dramatic rise in household debt in the last thirteen years has almost negated the impact of falling average interest rates. Figure 23 Page 22
Steve Keen's Debtwatch No 14 October 27 Draft only Interest Payment Burden 2 16.7% ('9) 8% 2% 3-7 '7 '8 ' Now 1 1 1 2 Debt (Per Cent of GDP) Income Shares Figure 24 6 6 Income Shares (% GDP at Factor Cost) Wages Profits 4 4 197 198 199 2 3 In the "it's an ill wind that blows no good" category falls the impact of rising debt levels on the share of income going to finance capital. Having shown no trend at all between 196 and 199, it has suddenly blown out in the last seventeen years, to almost four times the previous average level. Page 23
Steve Keen's Debtwatch No 14 October 27 Draft only Somehow I doubt that this is a good thing for the rest of the economy. It is instead a very potent indicator of the extent to which financial commitments are a burden upon the productive sectors of the economy. Figure 2 4 Finance Share of Income 3 2 1 196 197 198 199 2 Figure 26 1 Finance Share of Profits 8 6 4 2 196 197 198 199 2 Debt contribution to Effective Demand Page 24
Steve Keen's Debtwatch No 14 October 27 Draft only Figure 27 3 2 Growth of Debt & GDP Debt Nominal GDP 1-1 198 199 2 Figure 28 3 Annual Change in Debt 2 1-1 198 199 2 Figure 29 Page 2
Steve Keen's Debtwatch No 14 October 27 Draft only 2 Contribution of Change in Debt to Demand 1 1-196 198 2 Figure 3 Change in Debt & Real GDP 2 1 Change in Debt/Nominal GDP Change in Real GDP - 1 197 198 199 2 Ignore for a moment the labels on the next graph, and simply imagine that they were indicators on some medical or industrial gauge. Which series would imply an out of control process to you--the red one or the blue one? Of course, with the bias economists have developed about inflation--and the related blind eye towards debt levels--they ignore the red line, see only the blue line, and worry that this has recently moved up somewhat (even though, over the longer term, it has clearly fallen substantially). Figure 31 Page 26
Steve Keen's Debtwatch No 14 October 27 Draft only 2 1 Inflation vs Debt Debt to GDP Inflation Rate 2 1 1 1 198 199 2 - Dwelling Value and Mortgage Indices Figure 32: Debt trumps prices 6 Dwelling Price & Value vs Mortgage Debt Value of Housing Price of Housing Mortgage Debt 4 2 199 2 Page 27