Tuesday, January 16, 2018

Anna Zabai of BIS on Household Debt



Household debt: recent developments and challenges (PDF) by Anna Zabai of BIS. From the abstract:

Financial institutions can suffer balance sheet distress from both direct and indirect exposure to the household sector... In a high-debt economy, interest rate hikes could be more contractionary than cuts are expansionary.

The second part of that excerpt is interesting: asymmetry in monetary policy. Also, by implication, in a low-debt economy interest rate hikes could be less contractionary than cuts are expansionary. This could perhaps mean that there is a middle ground where hikes and cuts are symmetrical. In other words, a Laffer curve for debt and growth, again.

The first part of the excerpt exposes the paper's focus on household debt. This is myopia. In our most recent crisis, household debt was the problem (let's say). Therefore, household debt is always and everywhere the problem, and is the only private-sector debt problem worthy of discussion.


Debt lets households smooth shocks and invest in high-return assets such as housing or education, raising average consumption over their lifetimes. However, high household debt can make the economy more vulnerable to disruptions, potentially harming growth.

Here's the thing: The cost of debt is potentially harmful to growth. Therefore, debt is potentially harmful to growth. Therefore, household debt is potentially harmful to growth. I don't have a problem with saying household debt is potentially harmful to growth. But I have a serious problem with a failure to say that not only household debt is potentially harmful to growth. And it should be obvious that the problem is cost. But it doesn't seem to be obvious -- and not only in this BIS paper.

In order to assess the implications of elevated household debt levels, it is crucial to have a sense of whether households can bear the resulting debt burdens without resorting to large adjustments in consumption should circumstances worsen.

Good sentence.

One might add that the implications of elevated business debt levels can be assessed by its effect on the cost of value added, and probably on the price of traded shares of stock.

One might add also that the implications of elevated financial business debt can be assessed by the risk of onset of financial crisis.

One might add a word on public debt also. Because public debt has a cost, too. However, public debt growth (at least since Keynes) is generally a response to problems arising from private debt. The notion that the economy can be improved by reducing public debt (while ignoring private debt) is most foolish.

If all the noise about the Federal debt was redirected to the concept of reducing the need for Federal debt by reducing the amount of private debt, our economic problems could be solved in a matter of minutes.

Aggregate Demand & Irony:

From an aggregate demand perspective, the distribution of debt across households can amplify any drop in consumption. Notable examples include high debt concentration among households with limited access to credit...

Without limited access to credit, the high debt concentration would get even higher. Therefore, increasing access to credit is not a solution to this problem. Unfortunately, increasing access to credit is a standard tool of policy. That's what got us in trouble in the first place. Isn't it obvious?

Pure Irony:

Monetary policy is likely to have asymmetrical effects in a high-debt economy, meaning that interest rate hikes cause aggregate expenditure to contract more than cuts would cause it to expand (Sufi (2015)). This is because credit-constrained borrowers cut consumption a lot in response to interest rate hikes, as their debt service burdens increase. However, they do not expand it as much in response to cuts of equal magnitude. They prefer to save an important fraction of their gains so as to avoid being credit-constrained again in the future...

(Bold added. Irony in the original.)

A little something for the supply side guys:

From an aggregate supply perspective, an economy’s ability to adjust via labour reallocation across different regions can weaken if household leverage grows over time.

Here ya go:

A growing body of evidence points to the existence of a “boom and bust” pattern in the relationship between household debt and GDP growth (Mian et al (2017), Lombardi et al (2017), IMF (2017)). An increase in credit predicts higher growth in the near term but lower growth in the medium term.

Of course! And regarding the second sentence there, note that "An increase in credit" is an addition to debt. The reason you get "higher growth in the near term" is that the increase in credit translates into extra spending (in the near term). The reason you get "lower growth in the medium term" is that the extra spending has dissipated by then, and you're just left with the extra cost of the additional debt. Isn't it obvious?

This boom-bust pattern appears to be robust across different samples. Table 2, following Mian et al (2017), takes a first stab at exploring the relationship between household debt and GDP growth by looking at correlations.

A first stab? Maybe it isn't obvious.

The first row confirms the existence of a boom-bust pattern. Higher debt boosts growth in the near term but reduces it over a longer horizon.

Is it obvious now?

Elevated levels of household debt could pose a threat to financial stability, defined here as distress among financial institutions.

Just a reminder: Our recent crisis was related to household debt (let us assume). But that does not mean that other private sector debt is not also a problem.

Their thought continues:

In most jurisdictions, this is chiefly because of sizeable bank exposures. These exposures relate not only to direct and indirect credit risks, but also to funding risks.

The problem arises, in other words, not only from the risk of household defaults, but also from problems with the banks' own funding. So you can blame the households for borrowing too much. But you must blame the lenders for putting themselves at risk. And you must blame policy, for putting our economy at risk.

Irony, again:

Financial stability may also be threatened by funding risks (Table 1, column 7). In Sweden (as in much of the euro area), banks fund mortgages by issuing covered bonds, which are held primarily by Swedish insurance companies and other banks. This network of counterparty relationships could become a channel for the transmission of stress, as any decline in the value of one bank’s cover pool could rapidly affect that of all the others.

So it goes.

Okay. We finally get to private debt other than household:

This discussion suggests that household-based credit measures could be good predictors of systemic banking distress, much like broader credit measures (eg Borio and Lowe (2002), Drehmann and Juselius (2014), Jordà et al (2016)).

Borio. Borio is my buddy. (I don't think he knows it, though.)

The next sentences are worth repeating:

Among these [credit measures], the credit gap – defined as the difference between total credit to GDP and its long-term backward-looking trend – and the total DSR are of special interest. While the credit gap is typically found to be the best leading indicator of distress at long horizons (eg Borio and Drehmann (2009), Detken et al (2014)), the total DSR provides a more accurate early warning signal closer to the occurrence of a crisis (Drehmann and Juselius (2014)). Going forward, establishing the predictive performance of an appropriately defined “household credit gap” and of the household DSR seems especially relevant.

The "credit gap" is an interesting concept. Expect to see more on that, on this blog.

The conclusion of the paper is not satisfying. Well, I take that back. This part is interesting:

Monetary policy could have asymmetrical effects in an economy with high levels of household debt, meaning that an interest rate hike would be more contractionary than an equally sized rate cut would be expansionary.

I missed it before: To get equal up- and down-effects from changes in rates, you need smaller rate hikes and bigger rate cuts. What they are saying is that high levels of debt push interest rates down. High debt levels may also reduce the "natural" rate of interest, if there is such a thing.

That's interesting.

The rest of the conclusion, maybe not so much:

Central banks and other authorities need to monitor developments in household debt...

Macroprudential instruments such as loan-to-value caps (on the borrower side) or credit growth caps (on the lender side) are designed to force borrowers and lenders to internalise the impact of large credit expansions on the probability of a systemic crisis, thereby aligning private and social incentives. If these measures do succeed in stemming household credit growth, thus containing debt levels, they would also afford central banks greater future room for manoeuvre in setting monetary policy.

With apologies to Anna Zabai: Too intent a focus on household debt. Not enough concern with the "broader credit measures". I understand that the paper is specifically about household debt. And if the lessons learned here are woven into a tapestry of concern for debt in general, there is much of value here. But please don't forget about the rest of the debt.

Please don't forget about the rest of the debt.

Monday, January 15, 2018

US Population back to 1929

POP, FRED's number for Total US Population, only goes back to 1952. But FRED does show both Real disposable personal income and Real Disposable Personal Income: Per Capita, annual data, going back to 1929.

Divide the one by the other. Multiply by a million to make the units match POP. Now you've got US Population all the way back to 1929:

Graph #1: US Population back to 1952 (red) and back to 1929 (blue)
It's not perfect. Usually on "comparison" graphs like this, the red line is centered on the blue. In this case, the bottom edges of the lines are aligned. FRED's POP number is ever so slightly less than my number. Rounding, maybe. But hey, it's close enough for me. For now.

Show it as "percent change from year ago" values, and there is the baby boom! Wow:

Graph #2: Growth Rate of US Population since 1930; Baby Boom highlighted
The growth of RGDP relative to RGDP per Capita shows a similar pattern, but only starts between Batman's ears. You don't get the full effect.

Sunday, January 14, 2018

Households and Hedge Funds

See Note 1

Saturday, January 13, 2018

"The eye that looks ahead to the safe course is closed forever."

I've been reading Tim Taylor more, lately. I think I finally realized how good he is. From his 1 Jan post, quoting Adam Smith:

We trust the man who seems willing to trust us. We see clearly, we think, the road by which he means to conduct us, and we abandon ourselves with pleasure to his guidance and direction... We are afraid to follow the man who is going we do not know where.

And from his 1 Jan 2015 post, quoting John Courtney Murray:

The immediate situation is simply one of confusion. One does not know what the other is talking about. One may distrust what the other is driving at. For this too is part of the problem—the disposition amid the confusion to disregard the immediate argument, as made, and to suspect its tendency, to wonder what the man who makes it is really driving at.

In sum, we will not follow a man if we don't know where he is going. We fear he wants to take us somewhere we don't want to be.

In his sidebar, John Cochrane links to Writing Tips for Ph.D. Students (PDF) where he says

A good joke or a mystery novel has a long windup to the final punchline. Don’t write papers like that — put the punchline right up front and then slowly explain the joke.

I understand. Cochrane says to start out by telling people where you are going to end up. I think it's great advice. It reminds me of a something from a book my parents gave me back in 1956, Exotic Aquarium Fishes by William T. Innes. I was seven. It was my favorite book for years and years. The book is gone now; I lost it when we moved. I'll paraphrase.

Innes quoted an old preacher explaining why his sermons were so good:

First I tell 'em what I'm gonna tell 'em. Then I tell 'em. Then I tell 'em what I told 'em.

I still remember that advice. I understand perfectly well. But I cannot do it. I can't put the punchline up front.

I like to present a chain: First this, then this, then that. The "this" and "this" should be obvious, and the "that" is where they lead me. But "that" comes at the end, because it is the destination. It's where I'm getting to. It's not where I'm starting from. Can't help it.

Start with things you know, and keep going until you get to something you don't know. Then see if you can figure it out. That's Descartes. To me, that's science.

If I start with the answer, then you'll know where I'm going right from the start. Maybe you'd trust me more and fear me less, as Adam Smith suggests. Maybe.

I don't think it's fear. Maybe distrust: You don't know where I'll take you, and you don't want to end up in a place you don't want to be. You wouldn't want to be stuck having to say something nice about Donald Trump, for instance. And if I don't put the punchline up front, there is always a chance that you might agree with something I say, and only later find out it says something nice about Trump.

Heaven forbid.

Maybe you wouldn't want to agree that the explosion of Federal debt started long before 1980, or that the gap between productivity and pay opened up well before 1974. That could take you out of your comfort zone. But the economy doesn't care about your comfort zone. And me, I'm with the economy.

If the explosion of Federal debt didn't start well before 1980, the numbers would show it. If pay didn't start falling behind productivity until 1974, the numbers would show it. So would I.

You want to go with 1980 for the explosion of debt, and 1974 for the productivity-pay gap, yeah, that's fine. Look up somebody who puts his answer first and then provides all the evidence that fits.

I'm not disagreeable on purpose. I just tell you what the numbers tell me.

Conservatives spend their time understanding how the economy works. Liberals don't.

Conservatives think the economy is a system, and that if you understand the operating principles you can shoot down the liberals. Liberals secretly think the conservatives are right. They think that if you understand the operating principles of the economy you are led inescapably to conservative conclusions.

In order to avoid the conservative conclusions, liberals avoid trying to understand the economy altogether. They go so far as to deny that the economy is a system. They pretend you can pass laws to create whatever outcomes you want. Need higher wages? Raise the minimum wage. More hours of daylight? Pass a law to make the sun set later.

Here's the thing. Yeah, the liberals are wrong. But they are wrong in thinking that if you understand the operating principles of the economy you are led inescapably to conservative conclusions. That's what they have wrong.

Conservatives get to conservative conclusions because they are conservative. If liberals would begin by accepting that the economy is a system complete with operating principles, and follow the logic where it leads, liberals would get to liberal conclusions as surely as conservatives get to conservative ones.

And then, the discussion would at least have a logical base.

Now really: If I told you the ending at the start, would you have read this thing?

Friday, January 12, 2018

Two sides to every story // Not Investment Advice

Jeremy Grantham:

Profit margins are probably the most mean-reverting series in finance, and if profit margins do not mean-revert, then something has gone badly wrong with capitalism.

If high profits do not attract competition, there is something wrong with the system and it is not functioning properly.

Grantham probably means you should expect profit margins to come down, which will hurt stocks.

To me, the important thing he's saying is there is something wrong with the system.

Grantham was quoted in an article dated October 2015, now more than two years ago. Here's corporate profits as a percent of GDP thru the third quarter of 2017:

Graph #1: Corporate Profits as a Percent of GDP
Not the best context for profit, but I have nothing else handy
No mean reversion yet. Profits are half again as high as they were in the 1950s and '60s. Twice as high as in the '80s and '90s. No mean reversion yet.

How long do we have to wait for the mean to revert, before we can safely say there is something wrong with the system ?


More recently, Grantham is still pushing the same story. Bloomberg, 3 Jan 2018, reports:

Grantham cited the recent acceleration of U.S. equity prices, a concentration of leadership in stocks and growing media coverage of events such as bitcoin’s surge and Amazon.com Inc.’s success as signs that the final phase of a bubble could be coming in the next six months to two years.

Another two years?

I grabbed some stock market performance data from FRED:

Graph #2
It's not price/earnings ratios or anything like that. It's just numbers from FRED. But if you look at where the purple and green lines were from 1975 to about 1987, and draw a straight line thru that, you come out right on top of the red line. Purple and green are now well below the red.

That doesn't mean a lot, because if the economy deteriorated you would expect the green and purple to be lower than what an old trend predicts. Still, by this measure, stocks are not outrageously high.

The graph uses a log scale. That's why the trends are straighter than you might expect.

The only "bubble" I can see on the graph shows up as purple gaining on green between 1995 and 2000. I don't see anything like that happening now.

Benefit of the doubt? Maybe purple is gaining on the other lines in the last few years. Maybe. It could be like the first couple years after 1995 when that bubble was just getting started.

Hey, I'm not a stock market guy. (Buy green, sell purple?) But my impression is still that, except for the one bubble in the 1990s, it's pretty much a straight-line increase. The recent years don't look like another bubble to me.

It looks to me like Grantham is repeating the mantra that gave him name recognition. Bloomberg:

Grantham, 79, is best known for his accurate prediction in 2000 that U.S. stocks would lose ground for the next decade.

Me, I'm sticking with my prediction of economic vigor, the thing that's going to give me name recognition. Financial costs are down, and that frees up money to be spend on things other than debt service. I expect that money to be spent on stuff that counts in GDP and puts people to work. You know, like a normal economy when it is doing well.

I could be wrong. Grantham could be right. But if we don't get economic vigor like we had in the latter 1990s (perhaps along with a bubble like we had in the latter 1990s) then Grantham is definitely right: something has gone badly wrong with capitalism.

Thursday, January 11, 2018

Labor Share as Index and as Percent

Labor Share as INDEX:

Graph #1

Labor Share as PERCENT:

Graph #2

Looks a lot higher as an index, no?

Wednesday, January 10, 2018

"Inequality is a symptom"

Remember when we mixed metric and English units back in 1999 and crashed our Mars orbiter? The universe didn't care.

At Quartz: A Nobel Prize-winning economist thinks we’re asking all the wrong questions about inequality, 27 December 2017:
America is trying to come to terms with its economic inequality. Does inequality spur growth or kill it? Is it a necessary evil—or necessarily bad? Angus Deaton, an economics professor at Princeton, and the recipient of the 2015 Nobel Prize in economics, is asked questions like these all the time—and he doesn’t see the point.

“These are questions I am often asked,” Deaton writes in a column (paywall) for Project Syndicate. “But, truth be told, none of them is particularly helpful, answerable, or even well posed.”

Deaton believes the biggest misconception about inequality is that it causes certain economic, political, and social processes. But that’s backward. Economic inequality is a symptom ...

That's right. For people, inequality may be a problem. But for the economy inequality is just the way things worked out, given the policies and policy errors (GASP! Did I really say that?) of the last 40 years and more.

28 Feb 2011, The Usefulness of Good and Bad:

unemployment and foreclosures are problems for people. Not for the economy. If you want to fix the economy, you have to look at the economy’s problems.

You have to look for the things that produce the results you don't like. You have to track down causes. Why do you think I have so many posts saying "the debt didn't 'explode' after 1980", and so many posts about the gap between productivity and compensation, and stuff like that? Because that explosion and that gap and that stuff typically started long before 1980, long before 1974, long before the dates people usually identify as the start of those problems.

Sure, because people identify it as a problem when it becomes a problem for people. But if you want to fix it, you have to look back in time to when it was first a problem for the economy. You have to see the birth of the problem, and then you know what to fix.

21 May 2016, Self-correcting? What do you mean by that?:

I say things like: Jobs? You think 'jobs' is the problem?? Okay. But it's our problem. A problem for people. It's not a problem for the economy. If you want jobs from the economy, you have to give the economy what it wants.

I say things like: The economy does not care about inflation or unemployment. Those are not problems for the economy. They are problems for people. For the economy, they are simply ways to correct imbalances.

I say things like: We should use policy to keep the ratio of private debt to public debt at a low level, a level where the economy constantly wants to grow vigorously.

The economy doesn't care that you don't like inequality and unemployment and slow growth. It doesn't care about such things. If you want to fix those things, you have to do it by tweaking things the economy does care about.

Tuesday, January 9, 2018

Wages and Salaries arising from Government Employment

BEA data from Table 1.12 at FRED, National Income by Type of Income: Compensation of Employees. In particular: wages and salaries arising from government employment, as a percent of wages and salaries arising from all sources:

Graph #1: Wages and Salaries arising from Government Employment
From 14% of total in the late 1940s, rising to 21% in the early 1970s, but falling since that time, and below 17% as of 2014. Government's share of wages and salaries has been falling since 1975.

That's gonna be all levels of government, not just Federal.

Monday, January 8, 2018

"Never Confuse Them Again"

Heh, look at me, I'm trying to get people to see the differences between private debt and public. Fat chance. There are still people, a lot of people apparently, who can't tell the difference between an existing debt and an addition to debt.

It's a good thing I write for my own satisfaction. Sometimes there doesn't seem much hope of achieving anything else.