The ‘Wealth Effect’ and Debt – Two Sides of the Same Coin – Part 2

In our last post we covered the background of the wealth effect and how housing and stocks to drive economic growth. In this post we will cover the following topic:

1. How does Debt impact the Wealth Effect?
2. Why is understanding Credit Cycles and Business Cycles are Important ?
3. A Deeper Dive – The Wealth Effect and Economic Growth in the US – 1975 to 2021
4. Why are Interest Rates and Debt-Service-Ratio Important to Watch for an Economy reliant on the Wealth Effect/Debt Growth?

1) How does Debt impact the Wealth Effect and Economic Growth?

What is the Long-term Credit Cycle?

The credit cycle refers to the self-reinforcing interactions between perceptions of value and risk, risk-taking, and financing constraints. Typically, rapid increases in credit drive up property and asset prices, which in turn increase collateral values and thus the amount of credit the private sector can obtain until, at some point, risk appetite reduces due to change in conditions where the debt can no longer be serviced or bankers are no longer willing to supply credit (i.e. increasing unemployment/bankruptcies, pandemics which reduces the labor supply, natural disasters, interest rates are raised beyond the neutral rate of interest, war/tensions, etc.).  

Historically, this mutually reinforcing interaction between financing constraints and perceptions of value and risks has tended to cause serious macroeconomic dislocations. Risk assets (Corporate Equity/Stock and Real Estate) prices tend to follow the credit cycle on the way up and down.

The Bank of International Settlement (BIS) has done a fair amount of work in this area and can be described by the following:

• Financial cycle/Credit Cycle peaks tend to coincide with banking crises or considerable financial stress. During expansions, the self-reinforcing interaction between financing constraints, asset prices, and risk-taking can overstretch balance sheets relative to income/cash flows, making them more fragile and sowing the seeds of the subsequent financial contraction.

• The Credit Cycle can be much longer than the Business Cycle. Business cycles as traditionally measured have tended to last up to eight years, and financial cycles or Credit Cycles around 30–60 years since the early 1980s. The difference in length means that a financial cycle can span more than one business cycle. As a result, while the financial cycle peaks can usher in recessions, not all recessions are preceded by such peaks….we cover this later on.

• There is a strong link between the pace of debt accumulation and subsequent debt service, which in turn has a large negative effect on growth.

• Credit cycles describe the changing availability—and pricing—of credit. When the economy is strong or improving, the willingness of lenders to extend credit, and on favorable terms, is high.

• When the economy is weak or weakening, lenders pull back, or “tighten” credit, by making it less available and more expensive, contributing to asset values and further economic weakness and higher defaults.

---

So How Does One Measure The Credit Cycle?

• A common method of measuring potential output is the application of statistical techniques that differentiate between the short-term ups and downs (cycle) and the long-term trend (structural) based on demographics, debt, and technology/productivity changes.  

• The trend is interpreted as a measure of the economy’s potential output and the cycle as a measure of the output gap relative to this trend. The Hodrick-Prescott (HP) Filter is one popular technique for separating the short from the long term. We have used this technique in estimating Long-term Credit Cycle and Business Cycle. We used a smoothing factor of 400,000 for quarterly data in determining the trend which is consistent with BIS research.

• We used an HP Filter to de-trend the data (separate the trend from the cycle) using between 1-8 years for the business cycle using Real GDP (light blue line) and 30-60 years for the long-term Private Credit Cycle (orange line). This is a similar method used by central banks/economists.  

• The Private Credit Cycle (Table 1) is measured by taking:
  • Real Private Credit Growth (Household and Corporates) – Table 2;
  • Real Residential Real Estate Prices – Table 3;
  • Real Estate Price-to-Rent – Table 4;
  • Real Estate Price-to-Income – Table 5; and
  • Private Credit-to-GDP ratio – Table 6. 

However, there is a boom-bust cycle in which the boom part of the cycle is characterized by the higher risk appetite of households and corporates which increase and lever up their balance sheets through debt to grow their returns on capital. At some point, risk appetite is reduced (as interest rates rise) as debt service overwhelms underlying cash flows. In response, the fiscal authorities run large deficits (light-grey-line) during the bust part of the cycle, to allow the private economy (household and corporates) to de-lever and bring debts back in line with cash flows (see Table 1)

Table 1 – Long-term Credit Cycles – 1890 to 2021 – United States

Table 2 – Real Private Credit Growth (Household and Corporates)

Table 3 – US Real Residential Real Estate Prices

Table 4 – US Real Estate Price-to-Rent

Table 5 – US Real Estate Price-to-Income

Table 6 – US Private Credit-to-GDP ratio

What is a Business Cycle?

• Business cycles are a type of fluctuation in aggregate economic activity in market-oriented economies. Business cycles are recurrent expansions and recessions in an economic activity affecting broad segments of the economy. The business cycle goes through four major phases: expansion, peak, contraction, and trough. 

• The Central Bank helps to manage the cycle with monetary policy, while fiscal authorities use fiscal policy. A nation’s central bank influences the business cycle by targeting inflation and unemployment with targeted rates and other unconventional policies such as quantitative easing. It uses tools designed to change interest rates, lending, and borrowing by businesses, banks, and consumers, which in turn drives real GDP higher over time.

• The central bank lowers its target interest rates to encourage borrowing in attempts to end a contraction or trough to increase declining real GDP and deflation. This is called expansionary monetary policy because they are attempting to push the business cycle back into the expansionary phase.

• To keep the economy from growing too quickly (real GDP and inflation) and take on too much debt, the central bank raises its target interest rates to discourage borrowing and spending. This is called “contractionary monetary policy,” because the bank is trying to contract economic output to keep expansion under control.

• Expansionary fiscal policy aggregate demand by increasing government spending or lowering taxes—can be used to close a negative output gap. We would have witnessed fiscal stimulus during the latest pandemic which was used to close a negative output gap (real GDP below the potential GDP), reduce the unemployment rate, and increase inflation.    

• The goal of fiscal and monetary policy is to keep the economy growing at a sustainable rate while creating enough jobs for everyone who wants one and is slow enough not to increase inflation.

• The output gap is an economic measure of the difference between the actual output of an economy and its potential output. Potential output is the maximum amount of goods and services an economy can turn out when it is most efficient—that is, at full capacity

• The output gap measures the degree of inflation pressure in the economy and is an important link between the real side of the economy—which produces goods and services—and inflation. All else equal, if the output gap is positive over time so that actual output is greater than potential output, prices will begin to rise in response to demand pressure in key markets. Similarly, if actual output falls below potential output over time, prices will begin to fall to reflect weak demand.­ See table 7 –which shows the US Output gap over time. 

Table 7 – Real GDP ($B) versus Potential Real GDP (CBO) ($B) and Output Gap (RGDP)

• The unemployment gap is a concept closely related to the output gap. Both are central to the conduct of monetary and fiscal policies. The nonaccelerating inflation rate of unemployment (NAIRU) is the unemployment rate consistent with a constant rate of inflation

• Deviations of the unemployment rate from the NAIRU are associated with deviations of output from its potential level. Theoretically, if policymakers get the actual unemployment rate to equal the NAIRU, the economy will produce at its maximum level of output without straining resources—in other words, there will be no output gap and no inflation pressure.­ Table 8 shows the Output gap from an unemployment perspective. The Federal Reserve’s dual mandate is “stable prices” and “maximum employment,” referring to inflation and unemployment. It sounds complicated but means ensuring that the prices you pay for goods and services remain relatively stable over time and that everyone who wants a job in the U.S. economy can find one. Typically, when the output gap (RGDP and unemployment) becomes positive, this means the economy is performing above potential output and higher inflation above the target of 2%, which is what we are seeing in early Q1/22. To reduce demand and slow the pace of economic growth, interest rates are expected to be raised.  

Table 8 – Unemployment Rate vs NAIRU and Output Gap (Unemployment Rate) 

• The Federal Reserve Bank of San Francisco has recently looked at an alternative metric to measure the economic output gap (RGDP and Unemployment) and has suggested direct measures of the degree of labor market tightness, such as the vacancy-to-unemployment ratio (V/UR), provide superior inflation forecasts for prices and wages (Table 9).

This is an alternative to the statistical measures of output gaps the inflation-targeting central banking typically relies upon. The V/UR ratio represents the number of job vacancies, or demand for labor, relative to the number of unemployed individuals, or supply of labor or marginal cost of labor. As both unemployment and inflation are lagging indicators, raising the Fed Funds rate to a level that reduces the V/UR ratio by increasing unemployment to reduce inflation back to the target of 2% and potential GDP.

Table 9 – Job-Vacancy vs Unemployment Level

In Table 10, we show the various business cycles documented by Economic Cycle Research Institute across various countries since 1948. We discuss in the next section how central banks use various tools to influence monetary policy and influence the business cycle. 

Table 10 – Historical Business Cycles across Several Countries – 1948 to 2020 (ECRI)

How do Central Banks Influence the Business Cycle with Monetary Policy?

Central banks vary the short-term interest rate (overnight lending between banks known as interbank lending) based on the business cycle as measured by the output gap whether it is positive or negative. In influencing the business cycle, Central Banks use three primary tools:

1. Setting the current overnight interbank rate

2. Forward guidance to influence the expected future interest rates

3. Quantitative Easing (QE) or Quantitative Tightening (QT) – targeted purchases or sale of long-maturity bonds and other long-duration assets on the central bank balance sheet. This is how central banks influence the term premium. Longer-term bond yields, tend to be set by market demand and supply of bonds based on long-term growth and inflation expectations. However, as central banks buy a larger proportion of domestic sovereign bond markets via QE, interest rates will no longer be solely set by the markets. Currently, the US Fed Reserve owns about 30% of the sovereign bond market.

• Long-term Maturity Bond Yield = Expected Central Bank Interbank Interest Rate + Term Premium

• By varying the current interest rate, the expected path of future interest rates, and term premium, central banks influence long-maturity bond yields. Long-maturity yields, influence a variety of borrowing rates (the most popular mortgage rate in the US the 30-year term) and asset valuations¹ across the economy, which impact aggregate spending, employment, and inflation via the wealth effect as we discussed in Part 1.

In Table 11, we reviewed the impact of long-term inflation and growth expectations (structural) versus short-term interest rate expectation (cyclical) by performing a regression on various points on the yield curve (2-year, 5-year, 10-year) and expectations (inflation and real GDP over the next 10 years) and 3-mth T-bill rates (one-year forward).  You will note that 10-year yields are determined by long-term inflation and growth expectations, whereas shorter-term yields such as 2-year rates are much more influenced by short-term expected interest rates (cyclical) based on the current monetary policy stance.  

Table 11 – Yields (10-Year, 5-Year, 2-Year) vs. Long-term 3mth rate, RGDP and Inflation Expectations (10-year)

Longer maturity bonds depend on long-term inflation and growth expectations

• The term premium is the extra return required to hold to maturity a long-maturity bond versus rolling over short-maturity debt (e.g.3-month T-bills). The term premium is largely driven by economic factors beyond the current monetary policy stance, particularly over long-term horizons such as long-term expectations (inflation and real GDP growth), as well as savings/investment by households, and demand/supply of sovereign debt.  

• Term Premium = Inflation Uncertainty + Safe-Haven Savings + Net Demand/Supply of Sovereign Debt 

• QE influences bond yields by reducing term premium and QT increases term premium. The term premium is generally counter-cyclical – both inflation uncertainty and investor risk aversion tend to be higher in recessions than in booms – a flare-up in safe-haven demand can further suppress term premium and bond yields over the near term.

• Structural impacts are impacted by demographics which future impact will growth rates.

• Other structural impacts include savings vs. investment, and saving glut/inequality. Structural elements impacts both the natural rate of interest and term premium.

• The declining US working-age population (Table 12) will affect long-term yields/term premiums in 2 ways: 1) lower growth potential will lower neutral rate, and 2) higher demand for safe assets (bond vs equities) given that retirees tend to reduce their equity market risk. The shift in supply/demand balance will weigh on risk-reward for holding long-term.  On the other hand, the number of retired households is expected to rise and begin to consume rather than save, thereby contributing to a lower saving rate and may lead to an increase in the natural rate of interest.

Table 12 – Term Premium vs Growth in Working-age Population

How Could We Anticipate Central Bank Monetary Policy?

Simple Monetary Policy Rule – The Taylor Rule – Setting the Overnight Interbank Rate

• The Taylor Rule (TR), which is a simple formula that Economist John Taylor devised to guide policymakers. It calculates what the federal funds rate should be, as a function of the output gap and current inflation.

Central Bank Interbank Interest Rate = Natural Rate of Interest + Current Inflation Rate + 0.5 (Inflation Rate – Inflation Rate Target) + 0.5 (Output Gap) where:  

• Inflation Target = 2%
• Inflation = Core Personal Consumption Expenditures (Excluding Food and Energy)
• Output Gap = the percent deviation of real GDP from a Potential GDP (based on Congressional Budget Office Potential Output and BT model based on 10-year business cycle)  (discussed above)
• Natural Interest Rate = long-run equilibrium interest rate or neutral real rate, is the rate that would keep the economy operating at full employment and stable inflation (based on Laubach-Williams r-star model).

• The TR captures the intuition the central bank should set a higher interest rate when inflation exceeds its target and a lower interest rate when inflation is below its target. The TR prescribes a real interest rate above r-star when the inflation gap or output gap is positive, and a real interest rate below r-star when the inflation gap or output gap is negative. There are several variations of the TR depending on the inputs and assumptions.

• In Table 13, we have provided our application of the estimate for the Fed Funds Rates based on the Congressional Budget Office estimate of potential RGDP. You can see that the Fed Funds Rate has followed the Taylor Rule pretty closely over time and there have been times where the Federal Reserve has lagged slightly. We note the inflation gap in 2021-2022 is significant given the fiscal stimulus related to COVID-19, and supply chain issues, though the output gap has only started to close in the last quarter and there are signs the economy is starting to slow into H2/2022 from the demand side and the supply side, we are seeing easing in the ports and shipping rates (Baltic Dry Index).

• However, the Russia/Ukraine conflict (including various sanctions) and China’s no-COVID policy could potentially be a further source of higher than target inflation going forward in the supply chain. We saw a similar deviation from the Taylor rule that last time we saw high inflation in the late 1970/the early 1980s and Fed had to overcorrect. However, we have much higher debt-to-GDP now relative to this time.

• Looking back at the 1940s to 1950s (see Table 14), coming out of WW2 the debt level was similar to current levels coming out of the COVID-19 pandemic, and you will note that real interest rates were negative (inflation greater than interest rates). This financial repression allowed to get rid of the high levels of debt through keeping interest rate levels deliberately below the rate of inflation. We believe this will be a similar strategy used this time around as well.

Table 13 – Taylor Rule Fed Funds Rate vs. Fed Funds Rate – Q1/1961 to Q4/2021

Table 14 – Real Interest Rates vs Debt Levels

At high levels of debt, financial repression (setting rates below of rate of inflation) has been used to reduce leverage (Debt-to-GDP)  

We’ll discuss how interest rates influence debt-service, asset valuations, and credit/business cycle further in section 4.

2) Why is understanding Credit Cycles and Business Cycles are Important in an unanchored Fiat Currency System – A Review of the US – 1975 to 2021?

So why don’t we just track the business cycle?

As we mentioned in our last post, in 1971 (Type 3 system) the monetary system was delinked from gold and the US dollar and US Treasuries backed by the full faith and credit of the United States has continued to play a central role in the international monetary system during the latter part of the 20^th century due to the country’s economic strength and the stability of its political and judicial system.

As we noted fiat money could be created in unlimited quantities through the private banking system to support the real economy or by the government issuing debt and monetizing the debt via quantitative easing, if foreign countries were not willing to buy the debt in periods of crisis, to support the de-levering of the private sector.  

As a result, given the nature of unanchored credit-based systems, business cycles are much longer than pre-1971, and credit cycles do not build quite as high (see Table 15). Generally, the Credit Cycle peaks at about 7-10% above trend like it recently have in 1979, 1988, and 2007. By the end of 2021, US Credit Cycle was about 10% above trend.

In all cases, the Credit Cycles tend to reverse course as the Federal Reserve raises the Fed Funds rate. These peaks are also associated with equity market peaks in the S&P500 before large drawdowns within the next 12 months or so. Troughs have existed in 1993 and 2012.

Peaks in the credit cycle are associated with the Federal Reserve’s lagging the Taylor Rule Federal Funds Rate by a significant deviation of 2-3% for a period that allowed an asset price bubble to form in stocks and real estate.    

Table 15 – US Private Credit Cycle vs U.S. Real GDP – 1890 to 2021

So you may ask what has changed since 1971 that has supported increasing the length in the Business Cycle and Credit Cycles. Well, several things:

• Financial markets were liberalized starting in the 1980s. Without sufficient prudential safeguards, this change is likely to have allowed greater scope for the self-reinforcing interactions at the heart of the credit cycle to play out.

• Inflation-focused monetary regimes became the norm in reaction to high inflation of the late 1970-early 1980s. This led central banks to downplay the role of monetary and credit aggregates in their frameworks.

• The combination of growth in the working-age population (see Table 16), further participation from women, supply squeeze for oil, and liberalization of credit in 1970-the 1980s were potentially drivers for the high inflation during 1970-1980s.

• Central banks had little reason to tighten monetary policy if inflation remained low due to globalization (lower labor and goods costs), even as financial imbalances built up (including high house prices to income). 

• This focus on inflation targeting may have also reduced real wage growth and may have increased asset prices and the wealth effect, further exacerbating wealth inequality. Also, tax policy across many countries focused on investing in assets relative to labor.

• Push towards globalization: The entry of China and former communist countries into the world economy, alongside the international integration of product markets, and technological advances, boosted the global supply of labor reduced local labor negotiating power. Given the capitalist nature of many of the Western economies, moving production to low-cost jurisdictions maximizes corporate profits, though hollowing out manufacturing sectors. This generated returns to shareholders and households that owned stocks.    

• Demographic changes: Working-age population as % of the population (see Table 16) in the West first increased until about 1990s and declined thereafter, causing economies to rely more on capital/wealth effects and investments in technology to improve productivity to drive Real GDP growth forward rather than labor.  Structural factors such as regulation and taxation responded to support this movement away from labor towards capital. Slowing labor supply has also reduced natural rates over time…more on this later.  

• Further banking regulation (Basel Committee on Banking Supervision) was introduced to differentiate risk-based lending (lower credit risk weights generally for loans collateralized by residential real estate) and support the system in market stresses (government debt treated as a highly liquid asset) to maintain the existing system of lending and borrowing.

Table 16 – Working Age Population (Age 15 to 65) as % of Total Population

Coupled together financial booms could build up further and a turn in the credit cycle (i.e. increasing unemployment or bankruptcy, pandemic, etc), rather than rising inflation and the consequent monetary tightening, would trigger an economic downturn. This also appeared to have a lengthening effect in the business cycle.  So many market participants focus on central bank tightening and inflation cycles on determining the length of the business cycle, without considering the credit cycle and impact of the interest rate changes on debt-service ratios.

3) A Deeper Dive – The Wealth Effect and Economic Growth in the US – 1975 to 2021

To understand the wealth effect and the credit cycle, we need to take a deeper dive into the US wealth distribution.

At a high level, the US relies on a higher proportion of household spending at almost 70% of GDP and is much more unequal from the perspective of both wealth and income, as the Top 10% wealth has 70% of the country’s wealth and more wealth is in financial assets relative to real estate (see Table 17) than other countries we’ve reviewed.

As we noted in Part 1 Table 2, the ratio of Mean Net Wealth to Median Net Wealth was 7.0, which far surpassed the average of countries reviewed and income inequality was the highest with a 0.40 Gini coefficient.

Table 17 – US Household Net Worth and Wealth Distribution – 1987 to 2021

Through this inequality, the average wealth of the US has been maximized which is consistent with the American system of capitalism to maximize profits and capital gains of financial assets supported by capital markets. Lowering the cost of capital (through lower natural interest rates), has lowered the barrier for risk-taking within innovation and technology in the hopes of achieving outsized financial net wealth. The US continues to lead in technology and innovation as measured by patents and this is supported by underlying capital markets. A symbiotic link has existed between wealth creation and job creation. When businesses prospered, employment expanded and communities thrived.

The bottom 50% of net wealth in the US derives a larger proportion of their wealth in real estate assets (see Table 18) and changes have been volatile over time given the higher proportion of leverage used in Real Estate. The Top 10% of net wealth derives its wealth from stocks and has generally exhibited less volatility over the last 40 years (see Table 18).

Table 18 – Top 1% and Bottom 50% Net worth Percentiles vs Equity and Real Estate Prices 

The wealth effect via real estate is shown in Tables 19 and 20, as we note that Real Home Price Index (HPI) returns are strongly negatively correlated with 30-year mortgage rates (-0.74) and 30-year treasury bonds (-0.83) (which is the most common mortgage term for housing the US). Since 1990-early 2000, we note as the labor share of GDP declined in the US due to outsourcing, Real Estate has appeared to be more important to drive consumption forward. Mortgage Rates have followed 30-year Treasury rates and much of the decline in real interest rates have been the general decline of natural interest rates….more on this later.         

In Table 20, we note that as Real HPI is strongly positively correlated with Consumption as % of GDP (0.84), so it thought as Real Estate prices increase net wealth increases and consumption increases (consuming domestic services and a combination of domestic/imported goods). We note that home price increases appear largely on the demand side (lower rates for mortgages and wealth effect) rather than constraining supply (as population growth outstrips housing supply) as the population-to-dwelling ratio has remained relatively steady (slightly declining over the last 20 years) over the past 20 years we have data on (Table 21) and housing supply follows the business cycle largely. This dynamic is important to understand as many countries in the West rely on immigration for economic growth.       

Over the last 30 years, inequality has increased and there has been a shift of 10% in total net worth from the bottom 90% to the top 10% (see Table 17). Gains in net worth during 2020 and 2021 were the largest for the least wealthy (bottom 50% of wealth distribution).

Table 19 – US Real Housing Prices vs 30-Year Real Mortgage Rates

Table 20 – US Real Housing Prices, Real Disposable Income (DPI), and Consumption as % of GDP  

Table 21 – US Price-to-Income/Rent vs dwelling supply metrics

Looking at the Credit Cycle and Business Cycle together (Table 22), we note that during the late 1970s and 1980s, the credit cycle and business cycle tended to follow one another closely as labor/working-age population was a larger contributor to GDP growth over this period (see Table 23) as consumption and income shares for a large proportion of the country was better aligned. Since the year 2000, the Top 10% of Wealth has grown from 60% to 70% and more private consumption appears to be driven by this small cohort as the labor share has declined over the same period. This has led to a higher debt burden placed on the bottom 50% for housing and consumption.     

Table 22 – US Long-term Credit Cycle and Business Cycle – Q1/1976 to Q4/2021

Table 23 – Labor Compensation as % of GDP – 1976 to 2021

We note that income, consumption, and wealth were more evenly distributed during this time. Over the past 40 years, the US has increased wealth and income disparities and both wealth and income have flown to the top of the distribution, resulting in a higher debt burden placed on the bottom 50% for housing and consumption.  

Throughout the 1990s-2000s, globalization was the focus through outsourcing the manufacturing base to countries with ample labor supply (China and other emerging markets) at a lower cost-reducing labor compensation as % of GDP since 2000 and increasing profits for those that owed financial assets (stocks). In a healthy economy, companies continually are born, fail, expand, and contract, while new jobs are created and others are destroyed. Over the last 30 years, competition has been reduced and industries have been consolidated with large firms taking a larger share of the overall profit pool.

This lead to potentially relying on the top 10% wealth for a greater proportion of spending and investing. This reliance on a small proportion of the population for consumption and indebtedness of the bottom 90% for housing and consumer consumption potentially weakens Real GDP trend growth going forward. The top 20% of income distribution consumes 40% of the US total annual resources and can save almost ~20% of their income. This is particularly concerning, rather than financing investment in productive capacity, the savings have been associated with dissaving/debt accumulation. The U.S. economy requires political commitment and resolve to protect the robust competition that spurs productivity growth and improves living standards, even when well-resourced interests resist. We discuss this more later on…  

You may notice that the recession 2000-2001 did not have a typical peak in the credit cycle, but something else was going on here. Looking at table 24, US equity markets reached a peak (tech bubble) during the time (2000-2001) and consumption patterns (grey line) appear to follow the equity deviations from 2001 and onward.  It is very difficult to bifurcate the impact of the stock market relative to real estate impact on consumption though.

Table 24– US Long-term Credit Cycle/Equity Markets and Private Demand (Consumption & Investment)

In Table 24, we also note that the deviation in 2007-2008 of credit cycle/real estate coincides with the Great Financial Crisis (GFC). High levels of mortgage debt can therefore aggravate downturns and increase economic volatility. Indeed, downward pressure on house prices can adversely affect economic activity via wealth effects that lower consumption and via bank balance sheets and reduced new lending. During the US subprime crisis, non-recourse arrangements allowed borrowers with negative home equity, especially those that had bought to let or resell rather than occupy, to strategically walk away from their mortgage debt.

To get the economy (real GDP) going post GFC, the Federal Reserve lowered interest rates and flooded the markets with liquidity, and bought financial assets (bonds) in an attempt to lower long-term interest rates causing financial assets (stocks) to rise, and increase the confidence/risk appetite of investors and consumers. This was helpful for those Americans rich enough to own stocks and continued to increase the wealth gap within the country. The borrowed money was essentially interest-free, so investors and corporate borrowers took advantage of the situation and drove profits and stock prices up. The money did not trickle down to the bottom of the wealth or income distribution, so wealth and income gaps continued to grow.

Overall the American system appears to maximize average net wealth per citizen rather than median wealth, and favors financial assets (stocks) relative to housing which indirectly impacts economic output – this is consistent with a capitalist economy.

The US has a capitalist market economy that utilizes both markets and government regulation and intervention to distribute resources, where the vast majority of value-producing assets are under private ownership and are operated to maximize profits and capital gains from financial assets (stocks).   

A recent paper (February 2021) by Mian, Staub, and Sufi called the “Saving Glut of the Rich” highlighted the large rise in savings of Americans in the Top 1% of income or wealth distribution over the past 40 years, rather than financing investment in productive capacity, the savings have been associated with dissaving/debt accumulation by the non-rich in the distribution via housing debt and consumer credit and the government.

• The research finds that lenders indirectly used the top 1% saving deposits to finance borrowing by the bottom 90%, enabling the rich to benefits from debt repayments.
• The analysis also suggests that the top 1% of households in the US may have as much influence as emerging market economies in fueling the debt of the bottom 90% of the wealth distribution.
• Income distribution has been consistent over time, with high-income (top 20%) earning about 50% of the income (Table 25), and income outstripping consumption (Table 26), has enabled substantial savings by high-income earners over the last 30 years. The gains and savings experienced by the wealthiest households mainly reflect higher rates of return of capital and dividends (stocks), which have largely outpaced returns on real estate and wages over the past 40 years.

A proportion of these savings appears to have been placed in the banking system through time/savings deposits funding growth in household debt/real estate prices, as well as stocks.

Table 25 – Growth in Household Income by Quintile

Income across all cohorts grew ~4% annually and distribution remained similar

High-income/wealthy households have accumulated substantial financial assets that are direct claims on the US government and household debt (mortgage and consumer credit). We have summarized this comparison on Table 27 – Banking Net Asset by Wealth Distribution, and note that the Top 10% of the Wealth Distribution effectively benefits from the debt repayments, and the share of the claims via financial assets (time/savings and checking deposits) by the Top 10 grew by ~6% since 1989, which is in-line with high-income earners (top 20%) being able to save an additional 6% of income (see Table 17).

Table 26– Income, Consumption, and Savings by Quintile

Savings increased for high-income earners dramatically since 1989, at the expense of lower consumption/lower Real GDP, as aggregate consumption-to-income declined from 98% to 82% in 2020. The increase in savings has found its way into the banking system (see Table 27) and financial assets such as stocks. 

A large rise in inequality generates a ‘savings glut of the rich’ which has pushed the economy into a debt trap characterized by 1) Low-Interest Rates, 2) High Debt Levels, and 3) Output below Potential (Real GDP below Potential Real GDP).

These elements have been observed in the current system and we will discuss the implications in the next section Natural Rate of interest which discusses the decline of interest rates over the last 40 years. This decline in rates has improve asset valuations as well.

These elements have also been discussed in the context of the rise of political populism and polarization which has particularly intensified coming out of the Great Financial Crisis given the large wealth and income gaps. Typically when wealth and income gaps become large as they are now, these have historically devolved into civil wars a Ray Dalio has warned given the amount of political polarization and differing views on how resources in the economy may be allocated (labor vs. capital).

Coming out of the COVID-19 crisis, many countries have begun to question the strategy of globalization and supply chain risk/security in the face of rising international political tensions and the potential for conflict. It appears that re-weighting the US capital allocation more towards labor relative to capital and investment in domestic productive capacity (such as computer chip manufacturing and rebuilding domestic infrastructure) may help reduce the wealth gaps over time and potentially raise the natural interest rates and reduce supply chain risks, however, this may be at the determinant to the Top 1% of wealth holders. The increase in inequality has reduced potential Real GDP over time as well (Table 28) and the neutral rate of interest….   

Table 27 – Banking System Net Asset by Wealth Distribution

Table 28 – Real GDP vs Total (Public and Private) Debt Trends

Note: Real GDP Trend-based 10-year business cycle through the application of a statistical filter. Real GDP Potential from US Congressional Budget Office (CBO) estimates retrieved from FRED Federal Reserve.

4) Why are Interest Rates and Debt-Service-Ratio Important to Watch for an Economy reliant on the Wealth Effect and Debt to drive economic growth?

a) What is the Neutral or Natural Rate of Interest?

The natural rate of interest, also called the long-run equilibrium interest rate or neutral real rate, is the rate that would keep the economy operating at full employment and stable inflation and is said to be operating at trend Real GDP growth over a five to ten year period.

It is a function of the economy’s underlying characteristics (labor, capital, productivity) and is not “set” by the Federal Reserve, but in some sense can be thought of as the Fed’s target interest rate when the economy is at full strength. 

Understanding the real natural rate (or r-star) matters because it affects how the Fed steers interest rates, which impacts asset values (important to countries that rely on the wealth effect to drive growth).

The Fed may temporarily set the benchmark Fed funds rate, the rate at which banks borrow from each other overnight,  below or above the natural interest rate to stimulate or cool the economy, but will ultimately rely on its estimates of the natural rate to decide the direction rates should go and the level rates should reach. Since households and businesses make investment and savings decisions now based on what they think interest rates will be in the future, the Fed’s views on the natural rate have immediate importance.

The Natural Interest Rate or R-star is thought to be unobservable. Economic theory implies that the natural rate of interest varies over time and depends on the trend growth rate of output – higher trend Real GDP growth would increase the natural rate of interest. The level and variability of the estimated natural rate of interest depend on variables used to approximate the underlying inflation expectations. Below are a few simplifying formulas for the above concepts: 

• Output Gap (Real GDP) = Potential Real GDP/Trend GDP less Actual Real GDP
• Output Gap (Unemployment Rate) = Actual Unemployment Rate less Non-accelerating inflation rate of unemployment (NAIRU)
• Nominal Neutral Rate = Real Neutral Interest Rate (R-star) + Expected Inflation (over the long-term)
• Positive Output Gap = Accommodative Monetary Policy (Real Fed Funds Rate below R-star)
• Negative Output Gap = Tightening Monetary Policy (Real Fed Funds Rate above R-star)

The Laubach-Williams (LW) (2003) model uses data on real GDP, inflation, and the federal funds rate to extract trends in U.S. economic growth and other factors influencing the natural rate of interest. You will note since 1960 rates have declined. As Williams’ notes, there are numerous potential influences on the natural rate of interest, including, but not limited to, fiscal policy, technological change, and demographics. As of November 2020, LW natural rates of interest are no longer being published.

In Table 28, we show that when the real Fed Funds rate (orange line) moves above, the natural interest rate (blue-line), the business cycle reduces actual GDP below potential GDP to reduce inflation back to target and increase unemployment. You can see that Corporate Debt Service Ratios tend to peak (along with widening credit spreads) during these tightening cycles when the Real Fed Funds Rate is above the Natural Rates. As we are about to embark on a rate tightening cycle, we hope to establish early warning indicators. We’ll look at this in the next section.      

The natural rate of interest has been on the decline since the 1960s. Many reasons for the bond bull market/declining rates have been cited including:

• Demographics (decline in working-age population/labor supply) – we covered this earlier,
• Global savings glut (as creditor countries such as China, Germany, and Japan have invested excess US dollars in government bonds lowering the rate of interest),
• Wealth inequality/credit creation increasing by higher savings by higher-income households, and
• Reduced productivity over time (average from 1947-2018 has been 2.1%, but since 2005 productivity has been 1.3% and declining to 0.8% from 2010 to 2018)   

Table 28 – Real Interest Rates vs Business Cycle vs Debt Service Ratio – 1977 to 2020

However, reduction in real interest rates has been something going on for the last 800 years as documented by the Bank of England’s working paper “Eight centuries of global real interest rates, R-G, and the ‘suprasecular’ decline, 1311–2018” which covered 80% of advanced countries.

Capital accumulation/debt accumulation trends have been proposed for what has caused the decline over time rather than demographics, indifference to the central bank or fiscal policies, commodity-backed regimes (i.e. gold), and other factors. The paper also suggests that expansionary monetary and fiscal policy responses designed to raise real interest rates from current levels may have at best a cyclical effect (See Table 29).  

Table 29 – Real Interest Rates over the last 800 Years (1311 to 2018)

As Mian, Staub, and Sufi note, the demand for US dollar-denominated safe assets comes almost as much from the top1% in the US as demand from the rest of the world (China, Japan, and Germany).

One area that has not been examined in great detail has been the impact of Top 1% savings and debt accumulation by the bottom 90% since the 1980s and higher total debt-to-GDP ratio (public and private). We note in Chart 28 the Real GDP Trend growth has declined by 100 bps from 1976 to 2021 (or 75% of 1976 trend growth) as Total Debt-to-GDP has expanded by 2x.   

b) Why is the terminal rate lower each rate cycle?

Recessions in 2000, 2008, 2020 are all associated with Fed rate hiking cycles and generally rise 2-3% above trend to reduce the level of economic activity to bring Real GDP back in line with Potential GDP/Trend over roughly 2 years (the last cycle is the exception), which pushed debt-service ratios (DSR) up significantly from trend (see Table 30) as households and corporates fail to anticipate higher rates and may be over-levered. Interest rates are cut and government debt increases (fiscal impulse) and is used to de-lever the private sector (see Table 31).

As a result of these actions, Total Debt-to-GDP has increased over time and debt is only reduced when it is repaid or written off. We propose that the terminal rate is lower each rate cycle given the level of debt which keeps increasing to drive growth forward. You can see that Fed Fund Rate Trend has a strong negative correlation with the Total Debt-to-GDP ratio. The flip side of the higher household and government debt is the saving glut mentioned earlier. As more debt is taken on growth becomes weaker.      

Table 30 – Private Long-term Credit, Short-Term Credit (DSR) and Fed Funds Rate Cycles

Table 31 – Private Credit Cycle vs. Fiscal Impulse

Table 32 – Fed Funds Rate vs Total Debt-to-GDP

You can see that the Fed Funds rate has followed CPI deviation from the trend which is consistent with inflation-focused monetary regimes (see Table 33).

The historic hiking cycles generally are 2 years. Current market pricing (Overnight Index Swap curve) has the Fed Funds rates reaching terminal by end of 2023 and following roughly the same historical pattern of a 1.5 to 2% hike from a recent trend of 0.5% Fed Funds Rate, potentially suggesting a very short tightening cycle this time around.

Table 33 – Consumer Price Index Inflation (CPI) and Fed Funds Rates

While higher-income households tend to have higher debt loads, debt payments as a proportion of household disposable income are larger for lower-income households. We tend to watch changes in debt-service ratios and changes in interest rates rather than output gaps (Real GDP vs Potential Real GDP or unemployment gaps) alone as a determining factor of the changes between business cycle – growth and recession, given the level of debt within the economy held by a large proportion of the population in most countries and we don’t have separate debt-service ratios for different points of the wealth distribution. 

We have noticed a recurring pattern reviewing runs to 2000, 2008, and 2020 recessions in the US. When short-term real rates rise above long-term real rates, private debt-service (households and corporates) levels tend to increase above trend reducing the growth rate of the economy as employers begin to lay off employees to maintain profit margins, and in a high debt economy, interest rates need to be lowered or fiscal stimulus provided directly to indebted customers as we saw during COVID-19 in 2020.    

In Tables 34 to 36, we have attempted to estimate the business cycle to better understand the interaction of interest rates, Long-term Private Credit Cycle, Fiscal Impulse, GDP, and Private Debt Service Ratio.

• Estimated Real GDP (blue-line) and unemployment gaps (orange-line) using a statistical filter assuming a 10-year business cycle trend.

• We have compared the Real Fed Funds Rate and the Real Long-term Equilibrium rate when tends to follow the output gaps.  

• Real Fed Funds Rate (grey-line): We have used the 3-month Treasury yield as a proxy for the Fed Funds Rate less Long-term Average of Survey of Professional Forecasters from the Philadelphia Fed (over next 5-10 years), to help us map out the future path of the Federal Reserve. Expectations are usually in the market before the central bank moves.

• Long-run Equilibrium Rate: We have used the 10-year Treasury rate using a statistical filter assuming a 10-year business cycle less Long-term Average Inflation from the Survey of Professional Forecasters from the Philadelphia Fed (over next 5-10 years). Given the long-term nature, we believe this would pick up underlying trend factors of labor, capital, and productivity.  

• We’ve discussed the Credit Cycle (yellow-line) and Fiscal Impulse (light-blue) and note that oscillation between the two impulses, appears to have an impact on the output gaps, as the government comes in with their balance sheet (typically during the recession) when the private sector decrease.  

• Private Debt-service Ratio Deviation from Trend (green-line): Bank of International Settlements (BIS) data which began disclosing private DSRs beginning in 1999 which can be compared across countries. We have applied using a statistical filter assuming a 10-year business cycle to determine the trend, extracted the cycle (or deviation from trend). It takes 2-3 quarters of the Real Fed Funds Rate above the Long-run Equilibrium rate to reduce growth (output gap below zero) and results in a recession.

• Corporate Credit Spreads (BBB) Deviation from Trend (purple-line): We have applied using a statistical filter assuming a 10-year business cycle to determine the trend, extracted the cycle (or deviation from trend). We note that the purple-line generally follows the Private Debt-service Ratio Cycle green-line as the market anticipates increases in credit risk.    

Table 34 – US Output Gap, Credit Cycle/Fiscal Impulse and Real Rates

Table 35 – US Private Debt Service Ratio vs Interest Rates

When short-term real rates rise above long-term real rates, private debt-service (households and corporates) levels tend to increase above trend, which tends to lead to recession.  Therefore, the slope of the yield curve (yield curve inversion) tends to be quite predictive by leading about 12-months.

We intend to measure other select G20 nations to determine if similar patterns and cycles emerge over time so that we could use these as early warning indicators to reduce risk assets before recessions.  

Join us for our next series of posts on the Canadian economy and how housing contributes to growth.

Appreciate any questions or feedback you have or other research topics you may suggest.

Please contact us at beowulftreasury@gmail.com.

BT