"The big problem is not cost-push inflation, but cost-push decline."

 

Why the cost of finance must be reduced

I think maybe I had too many graphs in Wednesday's post. Today, I repeat the thoughts I couldn't put into words until those graphs were done.

First, allow me to remind you that the cost of finance is driven by two factors: the rate of interest, and the size of debt on which interest is paid.

 
The growing cost of finance was the prime mover, the initial cause that set in motion not only the wage-price spiral of the 1960s and '70s, but also our subsequent decline.

 

The cost of finance can create cost pressure and cost-push inflation as surely as OPEC can. People still today talk about OPEC as a source of inflation, and wages as a source of inflation. But nobody ever points the finger at finance and says here is the cost that drives cost-push. But it is. Finance is the root of the problem. Excessive finance.

It is all well and good to say business has to raise prices because wages have gone up, and to say labor needs an increase because prices have gone up. But the question that must be asked is "Which came first?" We have accepted the story of the wage-price spiral for half a century and more, without ever insisting that there must have been a first cause.

There must have been a first cause. But it cannot be wage hikes and it cannot be price hikes, because in the accepted story each of these is caused by the other. There must be some other cause, some initial cost increase that set the price spiral in motion.

That initial cost was the cost of finance.

The rising cost of finance in the years after the second World War contributed to the "creeping" inflation of 1955-57, the inflation that troubled Samuelson and Solow (1960). The cost of finance continued to increase, creating the cost pressure that led to rising wage demands in the latter half of the 1960s.

Finance is our primary growth industry. The growing cost of finance is endless. But the growing cost of finance is a source of cost-push pressure. It drives prices up and economic growth down.

Inflation is not the worst part of this problem. The cost pressure is the worst part. The cost push. Cost-push slows economic growth. The big problem is not cost-push inflation, but cost-push decline. And since the growing cost of finance is endless, the result is long-term economic decline. 

Long-term economic decline is indistinguishable from the decline of civilization.

The study of credit and economic growth

Graph #1: Slide the ScrollBar Right to See More, or just Click the Graph

 

 

 

 

 

It has to go up, or everything I thought I knew about the economy is wrong.

Here I develop an example of financial cost after World War II and consider the growth of that cost in subsequent years. The growing cost of finance is a source of cost-push pressure. It drives prices up and economic growth down. Moreover, the growing cost of finance is endless, because finance is our primary growth industry.

The growing cost of finance was the prime mover, the initial cause that set in motion not only the wage-price spiral of the 1960s and '70s, but also our subsequent decline.

 
For manufacturing, the FRED data on Average Hourly Earnings goes back to 1939. For the private sector in total, it goes back only to 1964.

Graph #1: Average Hourly Earnings before 1980

Where they overlap, the two data series run close. (Less close after 1980, but still close.) I will take the manufacturing data as an estimate of the average "Total Private" wage before 1964.

The FRED data for interest paid by the household sector starts in 1946. It's a big number -- billions, not dollars per hour. But I only want to see how fast it goes up, compared to hourly earnings. So dollars vs billions is not a problem.

Graph #2: Average Hourly (blue, red) and Interest Cost (green) Comparison

I added interest cost (green) to the graph. If I take the green line and multiply it by 0.86 and divide it by 1.735, green and blue will be exactly equal in 1946. I'll do that, I'll make them both 0.86 in '46. Plus I'll zoom in by cutting off the years after 1965:

Graph #3: Interest Cost (green) and the Average Hourly (blue, red)

Interest cost goes up a lot faster than the average wage in these early years. But in 1946, when the average hourly earnings amounted to 86 cents, people didn't spend their whole paycheck just to pay interest. There wasn't that much debt in those days, for one thing, and interest rates were low.

I should make the green line less. Household debt amounted to 20 percent of personal income in 1946. And the mortgage rate was, say, 4%.

So I can picture my dad's 86-cent hourly wage, and figure his debt was 20% of that, 17 cents of every dollar he earned in 1946. And the interest on the mortgage was 4% of 17 cents, or two-thirds of a penny. So now I want to divide the green line by something to make the 1946 value equal to two-thirds of a penny. That's what Dad paid for interest, out of every dollar he earned.

On graph #3, both lines have the value of 86 cents in 1946. If I divide 86 by 0.67 (which is two-thirds of a penny) I get 128.36. So I want to go the other way and divide the green line by 128.36 to reduce the 1946 value to two-thirds of a penny:

Graph #4: Household Interest Cost per Dollar of the Average Wage

It worked! The green line, now low on the graph, has a value of $0.00670 in 1946: two thirds of a penny.

The interest cost doesn't look like a big number, does it, on graph #4.

Let me look at it another way: interest cost as a percent of average hourly earnings. It will go up. It has to go up, or everything I thought I knew about the economy is wrong.

Graph #5: Household Interest Cost per Dollar of the Average Wage, Percent

Household Interest Cost
as a Percent of
Average Hourly Income

The green line goes up, as expected. It goes up faster than the average wage. It had to, if finance is the true source of the cost-push pressure. It had to.

So now I can say that the rising cost of finance in the years after the second World War contributed to the "creeping" inflation of 1955-57, the inflation that troubled Samuelson and Solow back in 1960. I can say also that in the years after the second World War, the rising cost of finance created cost pressure that led to rising wage demands in the latter half of the 1960s.

"Wage-push inflation" they called it. Labor got the blame. Nobody looked into it enough to notice that the rising cost of finance was putting pressure on wages. Nobody had our back. Nobody looked out for consumers. Nobody. 

Eh, that was all a long time ago. I remember, and I resented the "wage-push" nonsense because it meant they were blaming me and people like me for the inflation, people just scraping by. But I don't hold a grudge. That's not what this essay is about.

This essay is about the cost of finance -- a cost that can create cost pressure and cost-push inflation as surely as OPEC can. But people still talk about OPEC as a source of inflation -- and wages as a source of inflation -- and nobody, far as I can tell, nobody ever points the finger at finance and says here is the cost that drives cost-push. But it is. Finance is at the root of the inflation problem. Excessive finance.

It is all well and good to say business has to raise prices because wages have gone up, and to say labor needs a wage hike because prices have gone up. But the real question is "Which came first?" We have accepted the story of the wage-price spiral for half a century and more, without ever insisting that there must have been a first cause.

There must have been a first cause. But it cannot be wage hikes and it cannot be price hikes, because in the accepted story each is caused by the other. There must be some other cause, some initial cost increase that set the price spiral in motion.

That initial cost was the cost of finance.

Inflation is not the worst of the problem. Worst is the cost pressure, the cost push. Cost-push slows economic growth. The real problem is not cost-push inflation, but cost-push decline. And the problem is not once upon a time long ago. The problem is long-term decline, driven by the long-term growth of finance now and for the past 60 years.

Interest as cost

Total interest paid is equal to total interest received. Some people will tell you it doesn't matter how much we pay as interest, because we receive exactly the same amount of interest. But who is this "we"? Unless net interest is zero for every one of us, there is a redistribution of income. So it does have an effect. It does matter.

And even if it was net zero in every instance, it can still have an effect on the economy: 

• If 2% of our income is interest and 98% is income from productive activity, we got a lot of output with little financial cost.
  
• But if half our income is interest and half is income from productive activity, then we didn't produce much, and we had to pay a lot for it to cover the financial cost.
  

Between individuals, incomes vary. But for the economy as a whole, if interest is a high percentage of our income, then the nation's output will be inexplicably low. Inexplicably, because we choose not to see the high level of interest as a problem. Low, because interest is not a payment for the creation of output.

"Adversity"

CBS has a self-promotional TV commercial they play during the Army football game. The theme is adversity. They make it sound like losing a football game is the height of adversity.

That's not it. That's a bad day. Adversity is more like this:

"An essential feature of the law of motion" of the rise and fall of civilizations

Or, the fall anyway.

Financialization in a Long-Run Perspective: An Evolutionary Approach, by Alessandro Vercelli

On the view of Marx:

What we call financialization is nothing but the process through which exchange value becomes independent from and gains predominance over use value. In this sense, financialization is not just a symptom of the basic contradictions of capitalism but is an essential feature of the law of motion of capitalism.

Even a "long" century is short, compared to a Dark Age

Giovanni Arrighi: The Long Twentieth Century

From the Preface:

With the advent of the Reagan era, the "financialization" of capital, which had been one of several features of the world economic crisis of the 1970s, became the absolutely predominant feature of the crisis. As had happened eighty years earlier in the course of the demise of the British system, observers and scholars began once more hailing "finance capital" as the latest and highest stage of world capitalism.

Arrighi based his book on Fernand Braudel's interpretive scheme:

In this interpretive scheme, finance capital is not a particular stage of world capitalism, let alone its latest and highest stage. Rather, it is a recurrent phenomenon which has marked the capitalist era from its earliest beginnings in late medieval and early modern Europe. Throughout the capitalist era financial expansions have signalled the transition from one regime of accumulation on a world scale to another. They are integral aspects of the recurrent destruction of "old" regimes and the simultaneous creation of "new" ones.

Arrighi mentions "Braudel's notion of financial expansions as closing phases of major capitalist developments", and refers to

the current financial expansion, in the course of which the structures of the now "old" US regime are being destroyed and those of a "new" regime are presumably being created.

Presumably, he says. In the Epilogue, Arrighi considers "three possible outcomes":

• the old centers may succeed in halting the course of capitalist history
• the old guard may fail to stop the course of capitalist history
• capitalist history would ... come to an end ... by reverting permanently to the systemic chaos from which it began ...

Permanent systemic chaos, of course, is the Dark Age.

The other part of the debt problem

I came upon a site called Business Writing Services. The site offers to write an essay or term paper or research paper for me. It has a "Live Chat" option and two "Order Now" buttons. Needless to say, it is not my kind of place.

https://www.businesswritingservices.org/business-finance/372-factors-that-influence-the-cost-of-finance

But there is this, under the heading "Factors that Influence the Cost of Finance":

Effects of taxation – Debt finance is cheaper by the amount equal to tax on interest and this means that debt finance will entail a saving in cost of finance equivalent to tax on interest.

It describes the precise amount that debt financing can be less costly than some other methods of paying for things. They're talking about reducing your tax by taking advantage of the tax deduction. Anymore, this applies mostly to business.

"Debt finance is cheaper by the amount equal to tax on interest". That's the line that struck me. You borrow, you pay interest, you deduct the interest from your taxable income, and the tax you pay is less as a result.

See, you couldn't do that if you stole the money, or printed it yourself. You also couldn't do it if you worked to earn that money or issued more stock in your company. The tax advantage arises because you borrowed the money.

The tax code creates economic behavior which increases the use of credit and the growth of debt.

Felix Salmon says 

Just as America’s obesity problem is largely a function of the ubiquity of cheap high-fat food, America’s debt problem is a result of the ubiquity of cheap easy-access credit.

Salmon should know better. "Cheap easy-access credit" is only part of the debt problem. The other part is that policy encourages both the use of credit and the ubiquity of access to it.

This is from memory, but I think consumers could deduct all their interest costs from taxable income until around 1990. After that, only home mortgage interest was deductible. And according to Wikipedia, as of 2018 interest on home equity loans is no longer deductible.

Most of the interest deductions for consumers are gone now. The increase of debt got a boost because of those deductions but, unfortunately, the debt didn't go back down again when they took the deductions away. And now things are so bad that it's almost impossible to avoid going deeper in debt.

Most of the interest deductions for business still exist. Business still gets a tax break for using borrowed money. Again, this encourages the growth of finance and the growth of debt.

If you have economic policies that encourage borrowing, but you don't also have economic policies that encourage the repayment of debt, then debt will grow unnaturally fast and will reach an unnaturally high level. Even now, by the way, debt is still going up.

When policymakers eliminate tax breaks that encourage borrowing, rather than replacing them with tax breaks that accelerate the repayment of debt, they are not fixing the problem.

They raise and lower interest rates all the time. But debt just goes up

So I said "As long as debt grows faster than GDP, the cost of finance increases." Then I acknowledged that "interest rates also play a role in the cost of finance."

But how big a role? How much of interest cost is due to interest rates? And how much to the level of outstanding debt? Sure, as interest rates rise, more of the cost is due to interest rates. As rates fall, less of the cost is due to interest rates. As the accumulation of debt increases, more of the cost is due to the size of the debt. And as the accumulation falls in size -- Oh! Does that ever happen?

Not since 1937:

Graph #1: Percent Change in Domestic Nonfinancial Sector Debt (Annual Data)

The Excel sheet shows a -0.03% change between 1936 and 1937: For every $100 of debt in 1936, there was three cents less debt in 1937.

There has been no decrease in Domestic Nonfinancial Sector debt since that time.

As long as debt grows faster than GDP, the cost of finance increases.

As long as debt grows faster than GDP, the cost of finance increases.

Debt-to-GDP, 1834-2020

I make a statement too general to be supported, in order to draw your attention to a less general statement that may demand more of your time and attention. Obviously, interest rates also play a role in the cost of finance.

Arthurian theory in brief: Financial cost-push

As a rule, the cost of finance is always rising.

To the extent that finance is nonproductive, it increases the cost but not the volume of output.

This creates cost-push inflation or, if inflation is prevented, cost-push pressure that must find relief by some other means.

Cost-push pressure which cannot be relieved by inflation is relieved by slowing growth.

To the extent that finance is nonproductive, it slows economic growth.

Long-term growth of finance creates long-term slowing of economic growth.

Long-term slowing of economic growth is indistinguishable from the decline of civilization.

The Extents of Finance

Finance creates cost.

To the extent that finance is non-productive it creates cost-push pressure.

To the extent that finance is productive at the micro level by the extraction of rent, it creates cost-push pressure.

To the extent that finance is 100% productive, the growth of finance beyond its economies of scale creates cost-push pressure.

And to the extent that financial income comes from the nonfinancial sector, it creates cost-push pressure. 

Cost-push pressure slows economic growth.

One and one

From mine of 1 November:

Rising interest rates are not the only avenue by which the costs of leverage increase. As long as credit grows faster than output, the growth of credit itself increases that cost, even when interest rates remain unchanged.

 

From 6 November:

Debt-to-GDP, 1834-2020

The graph shows the size of nonfinancial debt as a multiple of the size of GDP since the 1800s:

• 1883: Debt about half the size of GDP
• 1928: Debt about equal to GDP
• 2002: Debt about twice the size of GDP
• 2020: Debt about three times the size of GDP
  

As long as debt grows faster than GDP, the growth of debt increases the cost of finance (unless interest rates fall enough to offset that increase).

Rules of thumb: Debt always grows faster than GDP. And the cost of finance is always rising.

The Problem with S1 E2

The Problem with Jon Stewart: I watch because it's Jon Stewart. I watched episode two a second time to take some notes. Second time, I noticed how good that episode is. The guests are great.

Not sure who said it. One of the guests:

"There is a slow erosion of the principles of democracy"

Not slow enough. And you won't agree, but the reason for the erosion, specifically, is that the economy is so bad. People can live with many things, but they can't live with unlivable living conditions. So when a guy comes along with Trump's skillset, they jump on his bandwagon. We think things get worse. They know things have been getting worse for a long time and they are anxious, some of them desperate, for a solution. Trump provides a clear alternative to the status quo. Thus the attraction. 

If you want to beat Trump, you also must provide an alternative to the status quo. The same old policy is not an alternative. Nor is more of the same. Not even if you add the comedic twist.

The voters who supported Trump (and even those who didn't) lived through the covid recession, most of them; not ten years before that, we had the financial crisis and that whole economic disaster from which we never recovered; farther back in time, there was the inflation of the 1970s -- and general economic deterioration since that time.

If you have money for cigarettes, you cannot understand the situation of a co-worker who doesn't have money for cigarettes. He's not likely to sit down and explain it to you, either. But there is a limit to how much he can take. Your best option is to be sympathetic, because you may be the next one to find yourself without cigarette money. Ours is a continuing economic decline. Nomadland.

Jon Stewart: "Everybody wants to ban shit that they don't agree with. And how do you square that?"

You square it, Jon, by fixing the economy. People will think you're going off-topic. But the economy is the root of the problem. Everything else is consequences. Fix the economy so that the government can recede, "a distant ship, smoke on the horizon". 

When people can afford their own medical insurance, the government doesn't have to make arrangements. Isn't that what people really want? When living standards are rising, the "socialist" social spending becomes less and less necessary. That spending can recede. As it recedes, it becomes less objectionable to those who object to it. That is how you bring polarized people together: by fixing the economy.

We have to solve the problem, not the consequences. To wait until we're starting to have insurrections, and then say insurrections are the problem, no Jon, that's not right. Not even close.

31 minutes into the episode, Bassem Youssef says

"The death of democracy in America could only be summarized in one word: money."

That's what I said, except I called it the economy.

Maria Ressa jumps in after Bassem and expands the thought:

"Look, it is about power and money, all across the board. Power and money. So think of -- the pieces are this, right:

• Someone seeds a meta-narrative that is kind of half true.
  
• It becomes a virus of lies and infects real people.
• Then those real people spread it. 

That is a diseased system. And that is what we are living in."

You picked up on that, Jon, and brought it back to money:

"It's like we're fighting two pandemics, two viral things. And it's interesting. I originally said oh this is like the fable of the three bears. It's really scarface. It's: first you get the money, then you get the power."

"First you get the money, then you get the power." There's a word for that, Jon: financialization. Economist Thomas Palley writes:

In effect, there is a politics of financialization that goes hand-in-hand with the economics.

It's all tied together. 

When we fix the economy, we change the flows of money and power. We just have to be sure we fix it right. Neither Democrats nor Republicans have the right plan.

A year ago, a month ago -- Same difference? No.

CPI and PCE -- Same difference? Yes.

Download as PDF

The first graph shows CPI, the well-known measure of inflation.
The second graph shows PCE, the inflation measure preferred by the Fed.
(Both graphs show the same difference: Blue is high, Red is low.)

The blue line shows the change in prices from a year ago.
The red line shows the change in prices from a month ago.
(Both are figured every month.)

Note: Blue and Red are the default colors,  not a political statement.

CPI Inflation since June 2019:

Graph #1: https://fred.stlouisfed.org/graph/?g=ICPC

 

 

PCE Inflation since June 2019:

Graph #2: https://fred.stlouisfed.org/graph/?g=ICPF

 

 

 

The change in prices a year ago was important a year ago.
What's important now is the change in prices now.

Debt to GDP 1834-2020

This you know:

Graph #1: Debt-to-GDP, 1946-2020

Dunno about you, but I want to know two things: What happens next? 

And what happened before.

I can't answer the first question. I can tell you that my guide to understanding what's going to happen is the cycle of civilization: "Civilizations die by suicide" and all that.

This post addresses the second question.

The debt data comes from A neverending debt trap by Steve Keen, from 2013. I looked at it some years back, and even some years before that. But it took me till now to look at it as debt-to-GDP.

The GDP data is from Measuringworth.

Graph #2: Debt-to-GDP, 1834-2020

One could ask: When did financialization start? The 1980s? The '70s? The '60s? I don't think so. The Civil War? Maybe.

One could quote Benjamin Friedman from 1986:

One of the most striking features of the U.S. financial system during the post-World War II era -- but not since 1980 -- has been the stable relationship between debt and economic activity... Moreover, except for the depression of the 1930's, the debt ratio was also fairly stable, and trendless, during the pre-war period extending as far back into the nineteenth century as available data permit.

Stable and "trendless" before 1980? No. Before the Great Inflation of the 1960s and '70s, there is a persistent upward trend going back almost to the start of the graph.

Friedman's assertion is backed up by the supporting evidence of this footnote:

See Friedman (1980, 1982) and Goldsmith (1985).

Evidence? References are not evidence. References only create the appearance of evidence. Eh, but I looked. In the first of those three references, Friedman writes:

Throughout their history, but more so during the twentieth century and especially in the years since World War II, the American financial markets have undergone a shift away from direct transactions between nonfinancial borrowers and lenders toward the intervention of financial intermediaries.

That sounds to me like an example of increasingly greater financialization. Friedman says this was occurring throughout the history of American financial markets. Since the beginning.

I do not see that the statement from Friedman (1980) supports the view that financialization and/or the debt ratio were "trendless". There was persistent increase in all aspects of finance, at least since the Civil War, and no doubt since the inception of the financial markets.

On top of all the other cost of finance...

At CFO: Metric of the Month: Total Cost of the Finance Function (May 11, 2015) by Mary C. Driscoll:

... CFOs are still expected to do more with less, however. In other words, automate on the cheap, move more work into shared services centers, cut headcount wherever possible, and drive down the cost of financial operations.

Yet gauged in terms of Total Cost of the Finance Function as a Percentage of Revenue, the first APQC Metric of the Month (an ongoing CFO feature),  many organizations are, sadly, still wasting hundreds of thousands or even millions of revenue dollars on inefficient finance operations (Figure 1).

This data reflects the relative cost profiles of 543 organizations that completed an APQC benchmarking assessment for the finance organization as a whole. Total cost includes personnel, systems, overhead, and any other costs necessary for day-to-day operation of the finance organization. The data shown in Figure 1 are calculated as follows: total finance cost divided by total business entity revenue, which is then (for display purposes) multiplied by 100.

I'm thinking these financial cost figures apply to nonfinancial businesses. So, at the median, the cost of overhead attributable to financial operations comes to 1.2% of revenue. These are internal costs, as opposed to the external cost of interest on a bank loan, for example. Hey, you have to have finance people in the company, to deal with the finance people you're doing business with.

So if I were to figure the total cost of finance, I'd start with the financial sector, add to it the financial dealings of the nonfinancial business sector, and then on top of all that cost add this 1.2% of nonfinancial business sector revenue, to cover the cost of the finance organizations within the nonfinancial businesses.

I don't think I'm double-counting anything.

Inflation: The "Percent Change from Year Ago" Lag

The red line measures price change from a month ago. The blue measures price change from a year ago. These are the default colors. Nothing to do with politics.

The high red peak in 1974 is not near 10%. The red line is measured by the numbers on the right side of the graph. That red peak shows prices increased a little over 1% from the month before.

The blue peak shows prices increased a little over 10% from the year before.

Graph #1: From-Month-Ago (red, right scale) vs From-Year-Ago (blue, left scale)

That red peak occurred in June, 1974. The blue peak appears eight months later, in February 1975. After June of 1974, inflation was falling. The red line shows a definite down-trend, and the line is not even very jiggy until it approaches a bottom in the latter half of 1975.

The blue line runs downhill parallel to the red but several months later. The "percent change from year ago" numbers lag the monthly changes shown in red.

Notice that later on the graph, where the blue line peaks at the end of 1980, the blue line again lags the red. That's the way the "from year ago" calculation works.

I expect it to work the same way now, lag and all.

Recent numbers show that from-year-ago inflation increased from 1.5 in February 2021 to 3.5 in May, and 3.6 since June. You should note, however, that some of this increase was due to deflation dropping out of the from-year-ago calculation in March and April 2021. That deflation came with the covid recession of March and April 2020. The deflation, by dropping out of the calculation, contributed to the from-year-ago inflation in the Spring of 2021.

Those below-normal monthly numbers were replaced by the above-normal numbers of March and April 2021. These higher numbers also contributed to the increase this past Spring.

Since April 2021, when the monthly inflation number peaked, the monthly increase in prices has been getting smaller. As of September 2021, monthly inflation is back in the normal range -- or close to it, depending on how you define "normal" inflation. The important point is that the current, month-to-month rate of inflation has been falling for five months, and prices are now rising at an unremarkable rate. 

Unfortunately, the from-year-ago calculation shows inflation to be much higher than recent monthly numbers show. The same thing happened during the inflation of the 1970s and 1980s. Also, as we saw at that time, the from-year-ago inflation develops a lag when the monthly inflation rate trends downhill. It takes a year to get all the old, high numbers out of the from-year-ago calculation. If the lag turns out to be 8 months again this time, the from-year-ago number will run high until December 2021, even if the monthly numbers continue to fall. Expect the media to make the most of it.

The Primary Principle of Policy

The ideas that credit use is good for growth and that interest rates can be used to manage growth, which together form the primary principle of policy, are undermined by the fact that the effects of accumulating debt are nonlinear.

"increased demand for credit can drive up interest rates"

Noah Williams, Financial Instability via Adaptive Learning, 2015

I have to take something out of context. 

In his PDF, Williams sets the stage by giving a brief summary of Minsky’s financial instability hypothesis. I skip to the point where he writes

the growth of credit expands

He then follows up, saying

Eventually, the growth of demand for credit leads to highly levered positions which can be difficult to sustain if there is an increase in the cost of debt service.

Well said. The funds that go to service debt are largely a transfer of income from the nonfinancial sector to the financial sector. This transfer of income hinders economic growth.

Williams attaches a footnote to that last quoted sentence, footnote 1:

Minsky emphasized how increased demand for credit can drive up interest rates and so increase the costs of high leverage. In this paper interest rates are constant, so this channel is shut down...

 In three bullet points we find a tidy endogenous package:

• "the growth of credit expands"
• "increased demand for credit can drive up interest rates and so increase the costs of high leverage"
• "highly levered positions ... can be difficult to sustain if there is an increase in the cost of debt service"

I accept this as a given. But rising interest rates are not the only avenue by which the costs of leverage increase. As long as credit grows faster than output, the growth of credit itself increases that cost, even when interest rates remain unchanged. The three bullet points can therefore be reduced to one:

• "the growth of demand for credit leads to highly levered positions which can be difficult to sustain"

Everyone on the internet points out the effect of rising interest rates. None of them point out that the same effect is created by the increase of outstanding credit.