Shiller vs Shiller

The Shiller interest rate numbers are monthly. Values after 1953 are from FRED maybe. Values before 1953 are calculated from annual numbers. Before 1953 the file contains numbers for January of each year, and calculated values for February thru December of each year. These calculated values produce straight lines from January to January, up until 1953:

Graph #1: Shiller Monthlies

The graph below uses only the January values from the same Shiller data:

Graph #2: Shiller Januaries

OMG!

So I started looking at the Shiller file, and I saw stuff in there like this:

=B21*7/12+B33*5/12

What the... ??

Turns out, Shiller is taking seven twelfths of a previous actual number and adding it to five twelfths of a subsequent actual number, to calculate a monthly value. It is all like that, with a number for January of each year and calculated trend-values for the eleven intervening months, from 1871 all the way up to 1953. After that there are numbers for every month, rather than calculations.

Shiller is figuring points on straight lines that run from January to January. That's why the EconomPic graph appears relatively smooth (up to 1953), then jiggy:

Graph #1: 110 Years of Interest Rates (EconomPic)

In the previous post I wrote

Gotta go with annual, because otherwise I'm just making things up.

Apparently, sometimes it's okay to just make things up.


Okay. But instead of taking averages of Shiller's calculated values, I'm just gonna go with the actual values that are given for January of each year. That's definitely my plan for the numbers up to 1953. After that I could take averages of yearly values. But it would be easier just to go with the January numbers all the way through. We'll see.

Frequency Aggregation

My DPD numbers are annual. I can go back to the 1940s or '50s with quarterly data, but before that it's all annual. So, annual it is.

Shiller's interest rate numbers are monthly. It makes for a spreadsheet with a lot of rows, something like 1600. Excel scrolls down well, but Google Docs goes at its own pace and you have time to see the numbers and notice whether they're increasing rapidly or not. While you wait.

That's not really the problem. The problem is that I want to compare the interest rate numbers to the DPD numbers on the same graph. So I need (or think I need) numbers with the same frequency: All monthly, or all quarterly, or all annual.

Gotta go with annual, because otherwise I'm just making things up.

So naturally I turned to Uncle FRED:

The FRED frequency aggregation feature converts higher frequency data series into lower frequency data series (e.g. converts a monthly data series into an annual data series). In FRED, the highest frequency data is daily, and the lowest frequency data is annual. There are 3 aggregation methods available- average, sum, and end of period.

The average, sum, and end of period aggregation methods all return lower frequency values with the same number of decimal places as the higher frequency values that are being aggregated. For example, a monthly series with values 100.1, 100.4, and 100.9 for the first 3 months of year is averaged to a quarterly value of 100.5 (i.e. 100.467 rounded up to 100.5).

FRED's GDP series numbers are quarterly, and when you convert them to annual, the default Aggregation Method is Average.

FRED's FEDFUNDS interest rate series shows monthly data, and when you convert them to annual, again the default Aggregation Method is Average.

I will therefore average Robert Shiller's monthly numbers to get my annual values.

FRED, I don't know what I'd do without ya.

Longwave (2): Make that three peaks

After I saw the similarities and differences between DPD and the interest rate (see yesterday's post) I decided to plot the two against each other, like inflation versus unemployment on the Phillips Curve graph. So I went back to EconomPic, where Jake has a link to interest rate data from Robert J. Shiller's Irrational Exuberance. When I clicked Jake's link, download started automatically for the Excel file ie_data.xls.

The file contains two sheets with graphs, and one with data going back to 1871: stock dividends and a price index among others, and in column G the interest rate numbers.

A graph of Shiller's interest rate data going back to 1871 shows three longwave peaks. Not two:

Graph #1

The graph also seems to show "secondary" peaks at the lows: one in the neighborhood of 1894, and another at the time of the Great Depression. The first of these falls neatly at the "bottom" between the 1872 and 1921 peaks. The second seems to come before the 1941 low, and belie the notion of secondary peaks. Could be. Or it could be that the severity of the Great Depression -- or the policy response to it, or both -- created additional downward force. I am obviously not prepared to say at this time.

The years between the first peak and the end of the first "secondary" peak match the years of the Long Depression.

Longwave

My thesis in this lecture is that macroeconomics in this original sense has succeeded: Its central problem of depression-prevention has been solved, for all practical purposes, and has in fact been solved for many decades.

The "many decades" to which Robert E. Lucas, Jr. refers in the above excerpt were the decades called "the golden age of postwar capitalism" and the decades called "the great moderation". Those many decades were the comfortable parts of the Kondratieff cycle, which is sometimes called the long wave.


A long-term view of the interest rate on 10-year Treasuries, and a long-term view of Debt-per-Dollar:

Graph #1: 110 Years of Interest Rates (EconomPic)

Graph #2: Debt per Dollar, 1916-2010

Both graphs display cyclic patterns. Each graph peaks twice, though the peaks occur at different times. The interest rate peaks earlier. DPD peaks later.

Since interest rates peaked and began falling in the 1980s, we should have known what was in store for us regarding debt. Since the 1960s we should have known. Heck, our grandfathers should have told us as much. But the people who study such things for a living (like Robert Lucas) were claiming victory over the forces of nature.

Black Friday... Black Saturday...

People don't remember things. Black Tuesday was not a good day. But that's forgotten now. Now a "black" day is for some reason a day to be a shopping fool.

Of course, Black Tuesday was a lifetime ago. You can't really expect people to "remember" it. So I guess if there was a business cycle that lasted that long, you couldn't really expect people to notice it either, huh?

"Those who cannot remember the past are condemned to repeat it"

The Job Guarantee

People call for a job guarantee. Why? Because people need income. But all income is money. And all money is debt, so all income is debt. So income is a problem.


What's wrong with that argument?

The phrase "money is debt" is what's wrong. The phrase does not mean "money is the same as debt". It means "money comes into existence through the creation of debt" or something like that. Whether that is always true is not up for consideration here. Certainly it is true sometimes.


I want a job so that I can receive money. Oh, sorry, that's too direct.

I want a job so that I can receive income, because the stuff I receive as income is money, and I want the money. So far, so good.

But now someone cries out All money is debt and that is supposed to bring the discussion to a grinding halt.

I want a job because I need the money, but all money is debt.

?? What is that supposed to mean?

At best, I have this: I want a job because I need the money. All money comes into existence through the creation of debt. Or, maybe not all money, but a lot of it. Or some of it, anyhow.


The phrase "All money is debt" is irrelevant. Stop using it.

"All money is debt"

You say: All red is color.

I say: We should be using green.

You say: Green is still a color, so there is no difference.

Thanks!

I may not be an economist, but I certainly am a blogger. I probably spend as much time checking my blog stats as I spend checking my email.

I thought today might be a good day to say thanks. Thank you for all the visits, and all the conversations, and all the challenges, and all the kind words.

From Google Analytics, a peek at visitors-per-week since 1 July 2009:

You guys give me confidence.


The other day I looked at the Blogger Stats and noticed something odd in the Traffic Sources. Most of the visitors were coming from Mosler's and from Mike Norman's.

That meant there were links there, to my blog. That interested me: What links? A timely look turned up links in comments posted by Clonal at Mike Norman's and Matt Franko at Mosler's.

The Blogger Stats also showed which posts were getting the traffic:

My most-visited current post showed up third on the list, after two that were targets of Matt and Clonal's links.

By the way, Clonal: The Greatest Scam of the 20th Century is high on my list of best and most important posts, so I'm real glad you linked to that one. And you did a great job of presenting it at Mike Norman's.

The Blogger Stats also showed a change in activity related to those links. Here's daily activity for 16 October to 15 November, showing a major spike:

It's always graphs with me, isn't it.


So thanks again, to everyone. To friends I can identify by name and point-of-view; to people who stop by regularly but have not left their imprint in the comments; and to first-time and one-time visitors who help to keep my Stat trends going up. And special thanks to Clonal, for thinkin and linkin my stuff. And to Matt, for picking up that ball and running with it.

Happy Thanksgiving, all.

(DEBTTLUSA188A)

A little something from our friend FRED:

Central government debt, total (% of GDP) for the United States

Notes:

Debt is the entire stock of direct government fixed-term contractual obligations to others outstanding on a particular date. It includes domestic and foreign liabilities such as currency and money deposits, securities other than shares, and loans. It is the gross amount of government liabilities reduced by the amount of equity and financial derivatives held by the government. Because debt is a stock rather than a flow, it is measured as of a given date, usually the last day of the fiscal year.

Not sure why it only goes to 2009.

I had a book that I used a lot when I was doing a lot of C programming -- Programmer's Problem Solver for the IBM PC, XT & AT by Robert Jourdain. In his Introduction, Jourdain warned that the book might not be easy reading: "The prose in this book is dense, to say the least".

That paragraph from FRED there? That's dense prose. Good stuff.


If something goes up during recessions, and kinda goes flat in a listless economy, then maybe it would go down in a vigorous economy, hey?

If that "something" is something you don't like so much, like central government debt, then maybe you would think restoring vigor to the economy is a high priority.

Averages of Averages

This is how I thought of it at first:

The Double-3: Take a zero-phase 3-year moving average of a zero-phase 3-year moving average of some economic data... I think this would be like assigning weights to the numbers, heaviest at the current value, lightest two years ahead or behind.

When I looked at it in Excel, a Double-5 gave a result more like what I had in mind:

Looking at any point on a trend line, the actual value at that point should have more significance than neighboring points; and more distant neighbors should have even less. I think this is a decent way to generate a trend line for an economic dataset.

To compare these images, use Bloggers Lightbox: Click a graph to activate the lightbox, and use the selector at the bottom of the screen to switch between graphs.

The Suppression of Money

I got numbers for M1 money from the Historical Statistics (1915-1970) and from FRED (1959-2011). The HS numbers are annual, so I decided to use annual values for this whole project.

I got GDP numbers from my Google Docs files. They came from Measuringworth. Again, annual values.

I divided the M1 numbers by the GDP numbers, to see how much spending-money there is in the economy relative to the stuff we buy each year. To see the history of that relation.

Graph #1: Money in Circulation relative to Output at Actual Prices

The blue part comes from the Historical Statistics. The red part comes from FRED. The M1 numbers from the two sources match up well.


I got "percent change from year ago" numbers for the Consumer Price Index from FRED. I used their CPIAUCNS, which goes back to 1913. (The "change from year ago" numbers go back to 1914.) I start the graph at 1915, because that's how far back the M1 money numbers go.

The blue line on Graph #2 shows inflation as "percent change from year ago" of the CPI. The abbreviation "PCYA" is awkward shorthand for "percent change from year ago":

Graph #2: Inflation (annual change in prices) plus trend lines

I put a 5-year moving average line on the graph (in red). And (in yellow) a 5-year moving average of the 5-year moving average values. Something I've been looking at a bit, lately.


For the next graph I got rid of the moving averages. I took the change-in-CPI line (since 1921) from Graph #2 and fitted it to Graph #1. To look for similarity, I scaled and shifted the CPI numbers: multiplied by a number that made the height difference of the CPI about the same as the height difference of the M1/GDP, and added a constant to slide the whole set of CPI numbers up and get them close to the M1/GDP numbers. My goal was to make the two lines close. I got them pretty close, on the left half of the graph. Not so close on the right half.

Graph #3: Combining Inflation from Graph #2 with MRTO from Graph #1

Look at the left half, and look at the peaks of the yellow change-in-CPI line. Those peaks pretty well fall within the limits set by the blue M1/GDP line. As the blue line rises, the yellow inflation peaks also rise, to a high point just at the end of World War II.

After the war the blue line falls, and so do the inflation peaks.

But then, on the right half of the graph the lines don't match up. Inflation spikes up three times in the 1970s while the quantity of money relative to GDP (now shown in red) continues to fall. Then in the Reagan years, inflation is suppressed. Perhaps surprisingly, the M1/GDP line stops declining at the same time. It runs horizontal for 10 or 15 years and shows the bumps that I point out all the time.

Inflation moderated when they stopped suppressing the money.

After the third bump, the M1/GDP decline resumes, reaching a low (ten cents) just at the start of the financial crisis and recession, 2007-08.

Anyone who says "they're printing too much money" has not spent enough time looking at this graph.

But the point is that in the years after the red and yellow lines cross, there really is not much similarity between M1/GDP and inflation. U.S. anti-inflation policy quit working in 1960.

Before 1960, inflation rose and fell within the limits set by the money ratio. Since 1960, the relation between M1/GDP and inflation has disappeared.

This failure on the graph, this loss of trend similarity, has a real-world counterpart. The inflationary peaks of the 1970s occurred while the quantity of money was being suppressed. And then inflation quieted down when they stopped suppressing the money. That's evidence of an astonishing loss of policy effectiveness.

A lot of people feel that inflation is proof that "they must have been printing too much money". But the graph shows that they were not printing too much money. The graph shows that since 1960 M1 money was severely suppressed but the suppression did not restrain inflation. There was a loss of trend similarity. The thing that used to work to fight inflation stopped working. That is what Graph #3 shows.


Why did inflation policy stop working? What caused the loss of trend similarity? What changed?

We increased our reliance on credit.

To fight inflation, the Fed suppressed circulating money. To stimulate growth, Congress encouraged credit use. They suppressed the money and encouraged credit use, and our reliance on credit increased.

Essentially our economy did not change, except that money was suppressed and credit-use was encouraged and we found ourselves using less money and more credit. The thing that changed was the thing we use for money.

The difference between money and credit is that credit is more costly to use. An economy that shifts from a low reliance on credit to a high reliance on credit will find itself with financial costs rising relative to other costs. It will find itself with interest costs rising relative to wages and profits. It will find itself with living standards squeezed, business growth below par, and financial activity as its growth industry.

An economy that shifts from a low reliance on credit to a high reliance on credit will see rising costs that are associated with the cost of interest.

I got interest rate numbers from FRED. I picked their Moody's Seasoned Aaa Corporate Bond Yield (AAA) because the numbers go all the way back to 1919.

Graph #4: Adding Moody's AAA Corporate Bond Yield to Graph #3

Again, I tried to make the lines similar by scaling the green line and shifting it up a bit. After 1960, the green and yellow lines are a very good match: Inflation and interest rates show similar patterns. Before 1960, there is little similarity.

Before 1960, money was dominant. After 1960, credit was dominant. Graph #4 shows the change from reliance on money to reliance on credit.

Graph #5 shows a close-up of the transition:

Graph #5: "X" Marks the Spot

Don't forget: the yellow line shows inflation. Inflation was at its lowest, and stable for a few years in the early 1960s, just at the balance-point between excessive money and excessive credit-use.

On your mark! Get set!! Decline!!!

And if it matters -- and it matters more than anything -- the decline in the quantity of circulating money was the cause of the economic problem, and the decline of income was a result.

18 November 2011

Graph #1: Showing the Decline in the Quantity of Circulating Money since 1946

Graph #2: Showing a Slowdown in Real Income Growth since about 1970

Graph #2, from Krugman, from a post somehow linking "the Reagan Non-miracle" to the slowdown of inflation-adjusted income growth.

Krugman's graph shows a slowdown starting somewhere around 1970. The decline of circulating money began in 1946, a generation before that slowdown. The slowdown of income cannot have been the cause of the decline of money.

Phase Shift

Earlier this month, jim offered some technical knowledge regarding the use of moving averages:

In digital signal processing (DSP) the moving average is classified as a low pass filter (tends to attenuate high frequency components in the data). The moving average is sometimes called a "boxcar" filter.

In DSP the issue of where you reference the input and output time stamp is referred to as causality. If you only use the most recent past input data points to create the current output data point the filter is causal. If you use data points ahead of the current output it is not causal.

If the output current sample point is the middle of the input sample points the frequency response of the filter is said to be "zero phase".

So... If the moving average is plotted at the last year of the years averaged, the result is "causal". If the moving average is plotted at the middle year of the years averaged, the result is "zero phase". The difference is significant enough that people invented names for the different versions.


Graph #1 shows the Federal Funds rate -- monthly numbers in blue, and annual numbers in red -- from the St. Louis Fed:

Graph #1

For Graphs #2 and #3 below, I'm using annual numbers. Because it's easier.

The vertical blue bars on the graphs below show the Federal Funds rate. The thin red line shows the moving average.

You can make the moving average move by sliding your mouse back-and-forth across the number bar below the graph.

Graph #2 is a "causal" representation. Each point on the red line shows the average of the previous N years:

Graph #2: A "Causal" Moving Average

1 YEAR DATA

2 YEAR AVG

3 YEAR AVG

4 YEAR AVG

5 YEAR AVG

6 YEAR AVG

7 YEAR AVG

8 YEAR AVG

9 YEAR AVG

10 YEAR AVG

You can see the peaks mellow and drift to the right as you move the mouse rightward over the number bar. They mellow because they get averaged down, blended in with lower numbers. They drift to the right because each point on the red line is plotted at the last year of the years averaged. You can see the same rightward drift in the start-point, the left end of the red line as you wave the mouse back and forth.

It is as if the mouse pulls the red line rightward, then allows it to spring back to its original shape.


Graph #3 is a "zero-phase" representation. Each point on the red line is plotted in the middle of the period averaged. The average value plotted at the average year, so to speak:

Graph #3: A "Zero Phase" Moving Average

1 YEAR DATA

3 YEAR AVG

5 YEAR AVG

7 YEAR AVG

9 YEAR AVG

11 YEAR AVG

13 YEAR AVG

15 YEAR AVG

Move your mouse back-and-forth across the number-bar below Graph #3. Again you can see the peaks mellow as you move to the right. But this time the mellowing peaks remain centered on the rigid peaks of the blue bar graph.

The "zero phase" version of the moving average makes more sense to me, for the things that I look at. It puts high points where the high points are, rather than dragging them off to the side.

Both versions are useful, no doubt. But the "causal" version appears to be more common; it's the way Excel does moving averages, for example. And the dragging-off-to-the-side thing is an important thing to know, if you plan to work with moving averages.

The red line itself is the same in both versions. Compare the 7-year Causal to the 7-year Zero Phase, or the 5-year Causal to the 5-year Zero Phase, for example. The red lines on the two graphs are identical. Only the location of that line against the background graph is different.

Money and Income (2): A Fallacy of Composition

Income is money. Income is circulating money that passes into your hands. That's why people often say "money" when they mean "income". And it's why some people, when you ask them where money comes from, say money comes from work.

Hey, I want more income, just like everybody else. And yes, when I say I want more income, I sometimes use the words "I want more money". Because income is money.

Boosting income could improve my personal economic situation, and yours. But if we boost our incomes by boosting each other's debt, we have gained nothing. Boost income by boosting money instead of debt, and everyone will be better off.

Insist at this point that money is debt, and you will see nothing.

Boosting someone's income is a microeconomic fix. Boosting everyone's income is still a microeconomic fix, the sum of the parts. The economic problem is a macro problem, excessive accumulation of debt. Boosting income carelessly does not fix the problem.


A FALLACY OF COMPOSITION

It is possible to assert that no individual owes more debt than he can afford -- in 2006 this was still a popular view -- and yet find that in the economy as a whole, the level of debt has become unsupportable. What is true for the individual, or even for every individual, may not be true for the whole.

Before the crisis, people didn't think they had too much debt. Maybe some did. But debt kept growing, so the general consensus must have been that we didn't have too much debt. And everybody thought they could handle it. Yet we had the crisis.

Some people say that everybody was wrong, that all the microeconomic evaluations were wrong: "Banks made bad decisions". I think that's a weak argument. I say the macroeconomic evaluation was never made. Policymakers made bad decisions.

The macroeconomic evaluation compares debt to the quantity of circulating money, out of which debt must be repaid -- and which is reduced by debt repayment.

The macroeconomic evaluation looks like my debt-per-dollar graph:

What is the source of the money that becomes income? The government? Sure. But that's not where I get money. I get money as income, or as a loan. People in general get money as income, or as a loan.

And that is exactly my point. Sometimes we borrow the money we spend. And sometimes we earn it. I'm saying we need to borrow less, and have more unborrowed money that we can earn and spend. So really, I'm saying we need more income. See? But I'm also saying how it must be done.

Others seems not to care where the money comes from, the money that becomes income. I think it makes all the difference.

If the money comes from increased debt, then the economy as a whole does not benefit from that increased income. Neither the debt-to-income ratio nor the debt-per-dollar ratio is reduced. But if the money comes not from increased debt, then the extra income does benefit the economy. Both ratios are reduced.

And if it matters -- and it matters more than anything -- the decline in the quantity of circulating money was the cause of the economic problem, and the decline of income was a result.

Sure, the extremes are different. But...

The Paradox of Thrift is sometimes portrayed as a test of the extremes: "when one person saves" versus "when everyone saves".

See, for example, Wikipedia:
In Keynesian macroeconomics, the "paradox of thrift" theory illustrates this fallacy: increasing saving (or "thrift") is obviously good for an individual, since it provides for retirement or a "rainy day," but if everyone saves more, Keynesian economists argue that it may cause a recession by reducing consumer demand.

See also Tom Hickey:
If one person saves instead of spending it is thrift, but if everyone saves instead of spending, aggregate demand craters and recession ensues.

See also Soojae Moon's homework:
When one person saves, that person’s wealth is increased, meaning that he or she can consume more in the future. But when everyone saves, everyone’s income falls, meaning that everyone must consume less today.

See also Krugman in Macroeconomics:
When one person saves, that person's wealth is increased, meaning that he or she can consume more in the future. But when everyone saves, everyone's income falls, meaning that everyone must consume less today.

See also Krugman in Essentials of Economics:
When one person saves, that person's wealth is increased, meaning that he or she can consume more in the future. But when everyone saves, everyone's income falls, meaning that everyone must consume less today.

For a lot more of this, see Google.


The Paradox of Thrift is often expressed as "when one person saves" as against "when everyone saves". This is an example of testing at the logical extremes.

Considering extreme cases is a useful technique. With the Paradox of Thrift, it shows different outcomes at the different extremes. But it only shows the outcomes at the periphery. It misses the central conclusion. In the case of the Paradox of Thrift, it shows different outcomes for "one person" and "everyone". It looks away from the fact that there must be a turning-point somewhere between those extremes.

Considering the effect of increased saving for only "one person" and "everyone" is a simplification. There is a vast distance between "one person" and "everyone".

Surely it is not a problem if two people save?

What if there is one person *not* saving?

Perhaps 17 is the magic number??

There is no magic number. Let me restate the Paradox of Thrift: On a sliding scale of "how much is saved" there is a point beyond which increased saving reduces income, which reduces saving. Beyond that point, saving more becomes self-defeating for everyone because it undermines economic growth (as Tyler Durden acknowledged). The outcome is different on the low side of the turning-point than it is on the high side.

Really, the Paradox of Thrift states that there is a tipping point at which there is a change in outcome even though there is no change in behavior. We save a little more, and our wealth increases. We save a little more, and our wealth increases. We save a little more, and our wealth increases. And then suddenly, we save a little more and our wealth declines.


And for the record, in the case of extreme income disparity it may be possible to have only one person saving, who saves so much that it creates the effect that is said to result when everyone saves.


Some people reject the Paradox of Thrift. Okay. For them, lets consider instead the Paradox of Taxes: On a sliding scale of tax rates, there is a point beyond which increased rates reduce tax revenues. Raising tax rates then becomes self-defeating.

The Paradox of Taxes may be familiar to you as the Laffer Curve. The Laffer Curve says that there is a tipping point at which there is a change in outcome even though there is no change in behavior. Tax rates are raised, and revenues increase. Tax rates are raised, and revenues increase. Then suddenly, tax rates are raised, and revenues fall. The Laffer Limit has been reached.


Come to think of it, the Paradox of Thrift works on a Laffer Curve, too.

The long-term growth of savings

What is not said is that the long-term growth of savings is central to the process by which we arrive at those abnormal times.

15 November 2011

I got numbers for M1 and M2 money from the Historical Statistics (Bicentennial Edition) and from FRED. The HS numbers cover the years 1915 to 1970. The FRED numbers cover 1959-2010. Some overlap. Annual numbers.

M1 is circulating money. M2 is circulating money plus money in savings. (M2-M1) is money in savings. I did a graph of (M2-M1) relative to GDP:

Graph #1

Big discrepancy between the old numbers and the new. I didn't look for a reason. No matter the reason, that discrepancy is disturbing. But I figure, there may be different standards of measurement for the two number sets, but there must be internal consistency within each set. I put the numbers on two different graphs.


The Early Years:

Graph #2

Accumulated savings bottomed out in 1918, relative to GDP, then rose despite everything until 1932. Then downtrend to 1943, interrupted only in 1938 by the recession-within-depression. Then, after a war-related hump that ends in 1951, a general increase in savings until 1968.

That 1951-68 increase corresponds quite well to the famous golden age of post-war capitalism.


The Recent Years:

Graph #3

Graph #3 shows the latter part of the strong increase in savings that accompanied the golden age. This graph shows the uptrend ending a bit earlier, in the mid-1960s. Savings then runs roughly horizontal until 1991, when it drops until 1995. In the "macroeconomic miracle" years after 1995, savings again increased rapidly. Savings continued to increase even after the miracle ended.


These graphs show that Savings-per-GDP increased
 • during the Roaring '20s
 • during the Great Depression
 • during the golden age
 • during the macroeconomic miracle years following 1995.
Savings increases when the economy is doing well (and also when GDP collapses).

Savings growth stagnated from the mid-1960s to 1991, and then actually fell. Stagnant savings growth corresponds to stagnant economic growth.

Savings growth declined during the FDR years that preceded the golden age; during the early 1990s before the macroeconomic miracle; and savings is in decline at present. Savings decline is a necessary prerequisite for the savings increase that accompanies growth.

So much should be obvious. In addition, the growth of accumulated savings is part of the monetary imbalance that causes Depressions and stagnations and financial crises like ours.

The economy's response to accumulated savings is essentially identical to its response to the excessive accumulation of debt. The long-term growth of savings is central to the process by which we arrive at abnormal times.

Normal Durden

From Guest Post: The Paradox of Thrift — Debunked, submitted by Tyler Durden:

  • Savings channeled into new capital investment actually boost growth.
  • Savings are only harmful when used to repay debt or other non-productive purposes.

In other words, saving is not the problem. Repayment of debt is the problem.

Apparently, if you borrow the money for productive purposes, it's okay to welch on the debt


The whole argument of the post is that the Paradox of Thrift does not apply in "normal" times:

Keynes was correct in his observation that high level of savings caused a shortfall in national income, but we need to remember that he was writing in the 1930s — in the middle of the Great Depression... What Keynes observed was an anomaly caused by the financial crisis.

Under normal conditions, however, the paradox of thrift does not apply...

If the entire nation saves, there is no effect on national income provided savings are channeled through the financial system into new capital investment.

This misconception that the paradox of thrift applies in normal markets has done immense harm to the economy...

Actually, the post entangles analysis of economic forces with the self-interest of savers in the attempt to create a more convincing argument:

This misconception that the paradox of thrift applies in normal markets has done immense harm to the economy and eroded the savings of the middle-class and retirees.

But such entanglements only make the argument weak.


Note the opening of the post:

Ever since John Maynard Keynes popularized the Paradox of Thrift, economists, central bankers and politicians have labored under the misapprehension that high levels of savings are bad for the economy and inhibit growth.

And this generalization, presented without evidence:

For three generations, central bankers attacked savers by artificially reducing interest rates — in the belief that lower savings would boost demand and stimulate the economy.

Three generations. I suppose this means "since Keynes". Since 1936 would be 75 years, okay. But let's go back farther. Let's go back 110 years. From EconomPic, here's a look at the 10-year Treasury yield, a measure of interest rates. Note that the dates on the graph are month-and-year, not month-and day. This graph starts in September 1901:

From FDR to Reagan, interest rates were on the rise. Since Reagan, since Supply-Side economics was put in place, interest rates have fallen relentlessly. There is no 75 years of central bankers artificially reducing interest rates. And -- until the crisis forced changes in the backstory -- the story was that since Reagan, inflation was under control, money was sound, and interest rates were right where John Taylor said they should be.

But now the story's different.


In abnormal times, the post says, "Not only will national income fall when savings are used to repay debt, but it falls rapidly... The impact on national income — as evident from the 1930s — can be devastating."

So, in abnormal times the Paradox of Thrift holds true, according to the Borg-like Durden. What is not said is that the long-term growth of savings is central to the process by which we arrive at those abnormal times.