Wednesday, August 31, 2016

Closing out "Household Debt Service" month with an update

FRED says the Debt Service data sets were updated "20 hours ago". Data for Q1 2016 has been added. From ALFRED, this graph shows the new release (red) with the previous release (blue) behind it. Not much has been changed, and the new data value is down slightly from the previous number:

Monday, August 29, 2016

Household borrowing hit a hard limit four years before the bottom fell out

Take household debt and look at the "change from year ago" values: That's household borrowing, in billions of dollars. That's the blue line.

Take household debt service payments "as a Percent of Disposable Personal Income" and convert it to household debt service payments in billions of dollars. That's the red line.

Look at what happened after the 2001 recession:

Graph #1: Household Borrowing in Billions of Dollars (blue) and
Household Debt Service Payments in Billions of Dollars (red)
Household borrowing (blue) rose sharply until it reached the ceiling set by debt service payments. Then it fell back. This happened in 2003. It happened again in 2004 and again in 2005 and again in 2006. It is as if lenders would lend no more than they could get paid back in a year... or borrowers would borrow no more than they could afford to pay in a year.

In 2007, after the fourth attempt, household borrowing went into free fall. You know what happened after that.

Sunday, August 28, 2016

Household Debt and Debt Service

After spending almost this whole month on debt service, I still don't know enough about the relation between debt service and debt owed. Here are the givens:

Graph #1: Household Debt (blue) and Household Debt Service Payments (red)
A small decline in debt since the crisis; a big decline in debt service. But the blue line uses the left scale and the red uses the right. So you can't look at the two lines and see relative sizes or anything useful like that. About all we can learn from this graph is that the Debt Service data doesn't start until 1980.


Here are the same two data sets, the one relative to the other:

Graph #2: Household Debt relative to Household Debt Service Payments
Well, debt is high, relative to debt service. High and rising. I don't like that too much. The only good thing I can say is that the increase seems to be slowing -- for what that's worth.

But this graph shows billions of dollars of debt, relative to debt service as a percent of income. The units are billions of dollars relative to percent. The units are screwy. I don't really know what the graph is telling me. Don't put too much stock in this one.

When I started on this little project this morning I didn't have my coffee yet. It was four in the morning, and even the dogs were still asleep. I took a sip of coffee and looked at what I was working with: debt in billions on the one hand, and debt repayment as a percent of income on the other. Apples and oranges.

I took the debt series (household debt) and made it "percent of income" like the debt service series:

Graph #3: Household Debt as a Percent of Disposable Personal Income
It's striking, that big decline since the crisis. Reminds me of the big decline in debt service. I wonder how the two compare -- household debt (relative to income) and household debt service (relative to income).

It's okay to compare 'em now, because both data sets are "relative to income". In fact, when I divide "debt-to-income" by "debt-service-to-income" income cancels income, and I'm left with "debt relative to debt-service", billions relative to billions. It shows how much debt we owe, for each dollar of debt service paid:

Graph #4: Dollars of Household Debt for every Dollar of Household Debt Service Paid
So, about six dollars of debt in 1985 for every debt service dollar paid... about eight dollars in the mid-1990s... about 10 dollars before the crisis and almost eleven after; but it has dropped a bit since 2013. Still very high, though. And with interest rates on the low side, this graph says households owe an awful lot of debt. An awful lot of debt, compared to what we're paying back.

Now it gets interesting.

Instead of looking at household debt as one big painful lump, let's look at how much it changes from year to year. And lets compare that to how much we pay as debt service in dollars each year, billions of dollars.

Graph #5: Change in Household Debt in Billions of Dollars (blue) and
Household Debt Service Payments in Billions of Dollars (red)
Let me talk thru how to read this graph. Take 1980 as an example, just when the red line starts.

In 1980 we paid just about 200 billion dollars in debt service -- interest and principal on the debt we owed that year. I don't know how much was interest and how much was principal. Let's say half and half. So we paid 100 billion interest that year, and we paid off 100 billion of debt. That's what the red line says.

That same year, 1980, the blue line shows that household debt increased by just under 200 billion dollars. See it on the graph? The blue line, just where the red line starts.

So in 1980 we paid off 100 billion of debt, and our debt increased by about 200 billion. What does this mean? It means we actually borrowed about 300 billion in 1980, but since we paid off 100 billion the net increase is 200 billion -- and the graph shows the net increase.

It makes sense to show the net increase, certainly, but it's also important to know that we actually borrowed more than that.

Second, take the 1990s as an example. The blue line is jiggy in the 1990s but if you imagine it smooth it is pretty much parallel to the red line during the period. So the net increase in household debt (blue) increased during the period, but so did debt service (red).

It looks like we were staying ahead of our debt in those years. But we were not staying ahead. Around the time of the 1991 recession, the net increase in debt was about half as much as debt service. But by the time of the 2001 recession, the net increase in debt was about two-thirds as much as debt service. The net increase in household debt gained on debt service during the decade.

Next, look at the years just before the crisis, where the blue line jumps up and runs neck-and-neck with the red -- and then suddenly falls, creating the crisis.

During those neck-and-neck years, the net increase in household debt was equal to our debt service payments. I'm not going to try to put into words how troubling this seems to me. I'm just going to say that this graph shows the problem that created the crisis.

Look at 2006 there, where the blue line rises above the red, then turns and starts to fall. In round numbers we paid about 1300 billion in debt service, and the net increase in household debt was about 1350 billion. The net increase in debt was more than we paid in debt service.

We were definitely not "staying ahead" that year. And the next thing you know, there was a crisis.

Saturday, August 27, 2016

Debt Service Projection thru 2020

One more time.

I grabbed all three Debt Service series from FRED, the three that I know about anyway. Looking only at the years since the crisis: since 2008 on the graph. Blue is the total of red and green:

Graph #1: Measures of Debt Service
Where will Debt Service go next?

There's a little "V" just at the end of 2012 in all three lines on Graph #1. After the "V", from Q1 2013 to Q4 2015, the lines show little disturbance to the trend. I took those years after the "V", 12 data points on each line, and used them as a basis for trend lines in Excel.

I went looking for the highest R-squared values for different kinds of trend lines, and the sixth order polynomials were highest every time. So I went with that, but when I extended the trends out into the future, the numbers got crazy.

I guess the sixth order line had too many turns in it, and the data after the last turn determined the path of the trend line. Projecting TDSP for nine quarters, the trend line went to more than 100% of disposable income. That's crazy. Things are bad, but not that bad. Look how flat the blue line is on Graph #1, after the "V". There's no way this line will climb to 100 percent in two years.

So I did it again, using the second order polynomial trend, which I've used before. This time each trend line predicts a smooth continuation of actual recent data:

Graph #2: Debt Service 2013 thru 2015 with Trends thru 2020
There is no guarantee that recent trends will continue, of course. But if recent trends do continue, the actual data will look like what this graph shows.

By this measure, MDSP (debt service payments on mortgages) bottoms out in fourth quarter 2016. So if you've recently heard that new home sales surged 12.4 percent and existing home sales had four consecutive months of impressive increases , well, I guess mortgage debt service payments should start going up soon. Graph #2 seems to offer a reasonable forecast.

But it is hard to see much on Graph #2 because it only starts in 2013. So I backed up the start-date to 1980. Now we can see the debt service forecast in the context of actual experience:

Graph #3: Debt Service 1980 thru 2015 with Trends thru 2020
Consider just the blue debt service line and its future projection. We see a large bowl shape that runs from about 2008 to 2020. Let me put this in terms you'll understand:

  •  Most of the downtrend of that large bowl was President Obama's first term.
  •  The bottom of the bowl was most of Obama's second term.
  •  The black uptrend part of the bowl is the first term of our next President.

What does this mean? Hold that thought.

Again, consider the blue line. There is a smaller bowl shape from 1991 to 1995 in the blue line.

  •  In 1991 and 1992 we have the downtrend of the small bowl.
  •  In 1993 and early 1994 we have the bottom.
  •  In latter 1994 and 1995 we have the uptrend of the small bowl.

The 1991-92 downtrend is related to the recession that helped George H.W. Bush lose the Presidency to Bill Clinton. The bottom of the bowl was a period of debt relief. The 1994-95 uptrend is evidence of the increased credit use that brought vigor to the U.S. economy for the rest of the 1990s.

There is no guarantee, of course, that the uptrend in debt service which is starting now will stick to a trend that brings vigor to the economy the way the 1994-95 uptrend did. But hey: Our bowl had its downtrend in Obama's first term. Our bowl had its bottom in the endless, listless economy of Obama's second term. And now our bowl has the start of its uptrend.

I can see the uptrend. Excel can see the uptrend. And I think you can see it, too.

Oh, and the uptrend means vigor. That's what it means.

// the Excel file

Thursday, August 25, 2016

Bezemer and Hudson, Quotable & Relevant

From Finance is Not the Economy by Dirk Bezemer and Michael Hudson:
Mathias Drehman and Mikael Juselius (2015) report that debt-service ratios are an accurate early warning signal of impending systemic banking crises, and strongly related to the size of the subsequent output losses.

Wednesday, August 24, 2016


BBC News -- Federal Reserve 'close to meeting targets' for US economy:
The Federal Reserve is close to hitting its targets for US employment and 2% inflation, according to the central bank's vice chairman, Stanley Fischer.

In a speech in Colorado, the Fed's number two policymaker was upbeat about the economy's recovery and prospects.

"We are close to our targets," he said on Sunday, adding that jobs growth had been "remarkably resilient".

He did not mention interest rates, but the remarks are likely to fuel debate about when they may rise.

He did not mention interest rates, but I will: They needed interest rates low to get economic growth so they can raise interest rates and undermine the growth. This is their plan for the economy.

We have to stop thinking always in terms of interest rates. We have to find a better way to prevent inflation.

On second thought, no. We don't have to find a better way. We already have a better way to prevent inflation: Paying down debt destroys money. Paying down debt takes money out of circulation. Paying down debt is a way to prevent inflation.

Paying down debt takes money that is in the economy, and takes it out of the economy. It's money that's in the economy that causes prices to go up.

New borrowing is generally for growth: for a new car or a bigger home or for business expansion, stuff that gets counted in GDP. Growth.

But new borrowing also puts money into the economy. And after that first use, the money's in the economy and it circulates. It is "extra" money in the economy. It may be used to purchase more output, or it may be used to bid up prices. The latter is demand-pull inflation.

Most of the "extra" money eventually works its way out of circulation and into savings. Then it has no demand-pull effect. But the debt that was created when the money was created -- the debt remains in the economy. And the cost of servicing it continues to be imposed on the economy; this creates cost-push inflation. Financial cost push.

Anna Schwartz describes demand-pull inflation:
An increase in the supply of money works both through lowering interest rates, which spurs investment, and through putting more money in the hands of consumers, making them feel wealthier, and thus stimulating spending. Business firms respond to increased sales by ordering more raw materials and increasing production. The spread of business activity increases the demand for labor and raises the demand for capital goods. In a buoyant economy, stock market prices rise and firms issue equity and debt. If the money supply continues to expand, prices begin to rise, especially if output growth reaches capacity limits.

Do consumers feel wealthier since the start of quantitative easing? Has our spending been stimulated? Are businesses increasing production? Is the economy "buoyant"?


Is our inflation, what little inflation we have, is it demand-pull inflation?


Does printing money make prices go up, if nobody's spending the money?


If new borrowing leads to a buoyant economy and demand-pull inflation, is raising interest rates an appropriate policy response?

Yes. Perhaps not the best of policy responses, but it does address the problem of new borrowing. But if the economy is not buoyant and inflation is financial cost push, then raising interest rates is bad policy.

Graph #1

Paying down debt is a way to fight inflation.

Tuesday, August 23, 2016

BIS on the connection between Debt and Labor Productivity

At BIS, from When the financial becomes real:

Financial booms typically go hand in hand with significant resource misallocations. In particular, labour is diverted to booming sectors with relatively low future productivity growth...

The impact of these misallocations became even larger in subsequent years, once the boom turned to bust... Thus, the fallout from credit booms may well have exacerbated the trend decline in productivity growth in advanced economies. By the same token, lower productivity growth in recent years need not be permanent.

... need not be permanent. In other words, it ain't secular stagnation, it's crisis-related. And maybe, the bank is saying, maybe we shouldn't be surprised if we see an improvement in productivity in the next couple three years.

Where have you heard that before?

Also this, remarkable from a bank:

Credit expansions may still boost output growth through higher demand and investment, but not productivity growth. To gain a sense of the economic significance, consider the US experience. Between 2004 and 2007, labour productivity grew by 1.2% per year, but labour reallocations made a negative 0.3 percentage point contribution. Over the same period, private credit to GDP grew by 4.5% per year. Taking the estimates at face value, if credit to GDP had grown by only 1.5%, the drag on productivity growth would have been eliminated.

They're saying we used too much credit. BIS -- one bank to rule them all -- BIS says private credit use was excessive. BIS says private debt was excessive. BIS.

This is big.

Monday, August 22, 2016

Debt Service and Labor Productivity Projections

Here's the graph I couldn't get right for yesterday's post. (X-axis was messed up.)

The blue line shows household debt service, 1980 thru 2015. The black line is a debt service trend line showing the future debt service that I anticipate; same as yesterday.

The brown line shows labor productivity, 1980 thru Q2 2016. The bright red line shows my general expectation for future productivity, Q2 2016 thru Q4 2020. Also bright red, the productivity  data for Q4 1993 thru Q2 1998 -- the data I copied and showed again as future productivity.

Graph #1: Debt Service and Labor Productivity Projections
By the time we get to 2020, the economy will be great. The person we elect President in 2016 will be easily re-elected in 2020. And we'll all be happy again for a while -- until debt goes too high once again and we have another financial disaster.

Anything else?

// The Excel file

Sunday, August 21, 2016

Potential Productivity

"Since 2007, the rate of productivity growth has been disappointing", John Cassidy writes in The New Yorker. "Since 2010, it has been extremely disappointing."

Here's a close-up of productivity growth since the start of 2011:

Graph #1: Productivity Growth 2011 Q1 thru 2016 Q2
The linear trend line is very flat at 0.5 percent. Productivity is low and not improving. What's more, the last few readings show a downward trend. Extremely disappointing, as John Cassidy says.

But we do not know where the red line will go next. It was lower in mid-2013 than it is now, and it went low more quickly. Then it turned around and went up even faster. So you never know.

Here's a close-up of productivity growth in the quiet time before the Goldilocks years of the 1990s:

Graph #2: Productivity Growth 1993 Q1 thru 1995 Q1
The linear trend line is very flat at 0.5 percent. Productivity, low and not improving. The last few readings show a downward trend. Productivity growth was disappointing. Just like the first graph.

Here's a close-up that picks up where Graph #2 leaves off:

Graph #3: Productivity Growth 1995 Q1 thru 1997 Q1
The linear trend rises from 0.5 percent to nearly 3% in two years.

You never know.

To my way of thinking, Graph #1 (our time) is very much the same as Graph #2 (the early 1990s). But I think we are at the end of Graph #2 and ready to start Graph #3. That makes all the difference.

Oh come on, you are saying. Graph #2 shows only a couple years. Graph #1 shows almost six years. There's no comparison. The slump is endless this time.

It's not endless. That's the point. The quiet time before the vigor is longer this time -- about 2½ times as long, my guess. I'm telling you we are at the end of the quiet period. Soon we will see productivity start to climb, just as on Graph #3.

On Friday I showed productivity growth with bright red circles around the two quiet times -- the recent years, and the early 1990s. Our low productivity has lasted much longer than that of the 1990s:

Graph #4: Productivity and Debt Service
Yes, and it shows that the "bottom" in the dull red "debt service" curve has also lasted much longer this time. But you can see from the shape of that curve that it wants to go up.

And you can see from the early 1990s that when the debt service curve goes up it goes up quickly, and productivity goes up with it. And the economy becomes vigorous.

I expect that debt service will soon rise sharply. Productivity will improve, just as happened in the mid-1990s. And the economy will again be vigorous.

Graph #5: The Recent Years
If we want to understand the productivity problem, we have to look at it in context. What context? I suggest we use household debt service for context.

The recent years of debt service show a remarkable drop from the late-2007 peak. Debt service fell rapidly, to a sharp down-spike at the end of 2012.

After the bounce-back from that down-spike, the path of debt service was different. The rapid decline had ended. It had reached a bottom. It was running flat.

If you look, you can see that debt service since the beginning of 2013 shows a fairly smooth curve. The curve bottoms out just at the end of 2014, then starts to rise.

To my eye, debt service drifted downward less quickly in 2014 than 2013, then turned and started drifting upward in 2015. Started drifting upward last year.

Curse the luck, the debt service data ends with 2015. We don't have first- and second-quarter 2016 numbers that let us see what's really happening. But I think 2016 will drift upward a little more than 2015, continuing the pattern that began at the start of 2013.

And I think that after 2016 debt service will rise even faster. That's when we'll start to feel the vigor. That's when we'll see the improved productivity. Here: I mirrored the curve on this next graph to show the kind of future I expect:

Graph #6: Our Near Future?
Just to give you an idea.

I should say, though, that if debt service rises as far and as fast as this graph suggests, then the recession bar after 2025 will be even wider than what the graph shows, and the recession more severe than the last one.

But you take my point: If the debt service curve is going down, then going down more slowly, then going up instead of down, then going up more quickly, this is what the graph must look like. And if debt service goes up and up, productivity will improve and the economy will show vigor -- for a while at least.

I took the eight data values for 2014 and 2015 -- the slowly-down-and-slowly-up drift in the curve -- and used those eight values to create a trend line in Excel. The trend line runs into the past and future, to show where debt service will be (and where it would have been) if those eight values determined the path of debt service:

Graph #7: Household Debt Service and the 2014-2015 Trend
The original eight values are shown in red at the end of the blue line, just at the bottom of the U-shaped trend line. Going forward, the U-shaped trend provides an estimate of the future path of the blue line, the future path of household debt service.

Going backward, the U-shaped line is not an exact match to the blue line during its rapid 2007-2013 fall. But the lines are close enough to make you stop and think. And that means the right side of the U-curve, the side that imagines the future, is likely also a pretty good estimate.

Time will tell. In the meanwhile I have to look. I have to see if I understand what's going on. I have to see if I understand the economy.

I made a screen capture of household debt service in the 1990s so we can look at the "dip, bottom, and recovery" (DBR) of that time. It is shown here in place (in the 1990s) as a dotted green line in a black box, overlaid on the original blue line:

Graph #8: Capturing the "Dip, Bottom, and Recovery" of the 1990s
I took that image of the 1990s, with the green dots there, took that image and scaled it up by a factor of 2.5: that is, 2.5 times as tall and 2.5 times as wide. Then I moved it so that the "dip" of the 1990s lines up with the 2007-2013 dip on the blue line:

Graph #9: Looking at Current Conditions as 2½ times the size of the 1990s Conditions
The green dots -- the 1990s data, scaled up by a factor of 2.5 -- the green dots make a very good match to the 2007-2013 downtrend of the blue line. A very good match.

The green dots also suggest that the debt service "recovery" will happen sooner than Excel's U-curve says. This makes me stop and think. Excel's U-curve is based on the most recent two years of debt service data. Only two years of it, but the most recent two years.

Perhaps the debt service recovery will not come as quickly this time as it did in the 1990s. That seems a reasonable conjecture. Excel's U-curve seems to me a better bet than my scaled-up green dots from the 1990s. So I grabbed the image of the 1990s again and stretched it, still 2.5 times as tall but this time 3 times as wide. Three times the duration.

I fitted the green overlay to the blue line as before:

Graph #10: Looking at Current Conditions as 3 times the Duration of the 1990s Conditions
The scaled-up green dots from the 1990s align with the 2007-2013 fall of the blue line as before. And this time, the green dots of the 1990s recovery align quite nicely with Excel's U-curve.

What does it mean, really? Really, it means nothing: It's a prediction. Still, if the prediction turns out to have been correct, it means maybe I do have a pretty good handle on the economy. Time will tell.

It all comes down to prediction on my part.

You don't expect the economy to pick up. You don't expect vigor. You don't expect productivity to rise. I look at debt service in the '90s and say "That pattern is being repeated right now."

"No," you say. "The economy is different since the crisis." And you are right: Different it is. Maybe things won't pick up this time. Maybe the economy will stay as flat as my recent Blogger stats.

But everything I read tells me that people are tired of this washed-out recovery, which is less a recovery than a continuing depression. People are tired of it. People are ready for recovery. Nobody believes me when I say "vigor", but everybody's ready for it.

It will happen. It won't happen because of my charm and wit, but it will happen.

It will happen because people are ready for it. Nobody was ready yet, in 2010. Everybody wanted it, but everybody knew vigor was unrealistic. Today, people don't say vigor is realistic, but everybody wants it. We're not saying vigor is possible, but we're hungry for it.

You know what that is? That's expectations.

Expectations have turned. Can vigor be far behind?


The Excel file for the first three graphs makes use of the FRED series OPHNFB_PC1: Nonfarm Business Sector: Real Output Per Hour of All Persons, Percent Change from Year Ago, Quarterly, Seasonally Adjusted.

Friday, August 19, 2016

Productivity and Debt Service

This graph shows productivity. I have circled the low productivity of the past few years, and also the low productivity of the 1990s in the years before "Goldilocks".

Graph #1: Productivity
When I take that graph and add household debt service to it, you can see that the circled years of low productivity are quite obviously related to the lows in debt service:

Graph #2: Productivity and Debt Service
Need I say more?

Thursday, August 18, 2016

Low Debt Service plus Growing Debt equals Productivity

The red line is productivity, shown as percent change from year ago.

Graph #1: Household Debt & Debt Service (blue) and Productivity (red)
The blue line combines the household debt service ratio and the growth of household debt.

The two lines don't always match, mostly because of the big spike in productivity that we typically get after recessions. But set those spikes aside, and the lines do at least hint at similarity.

The red oval mid-graph highlights productivity and debt conditions during the "Goldilocks years".

The red oval on the right shows debt conditions even lower now than during 1993-1995, so there is plenty of room for things to go up.

Productivity has been low since 2011, as the graph shows. But productivity was also low after debt conditions bottomed out in the early 1990s, just before the Goldilocks years.

Productivity is going to pick up soon.

Wednesday, August 17, 2016

Illicit use of the Hodrick-Prescott?

The Hodrick and Prescott (1980, 1997) filter (hereafter, the HP filter) has become a standard method for removing trend movements in the business cycle literature.

I'm reading Notes on Adjusting the Hodrick-Prescott Filter for the Frequency of Observations, a short PDF by Morten O. Ravn and Harald Uhlig. The paper is © 2002 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology, so they're probably right.

Most applications of this filter have been to quarterly data, but data is often available only at the annual frequency, whereas in other cases monthly data might be published. This raises the question of how one can adjust the HP filter to the frequency of the observations so that the main properties of the results are conserved across alternative sampling frequencies.

Their topic is tweaking the HP calculation to allow for different data frequencies.

I use Kurt Annen's VBA code in Excel to do the HP calculation. The code creates a function named HP( ) that does all the work. All you have to give it is the range of source data and a constant.

The constant is how you allow for different data frequencies. Changing the constant changes the amount of "smoothing" you get when you graph the results. The trick is to use the right value for the constant.

I've always used the values 100 for annual data, 1600 for quarterly data, and 14400 for monthly data. Something I picked up from the EViews User Forum a while back. I have some old notes on it.

I shouldn't say I always use those values. I always start with those values. Sometimes I change them. For examples of how changing the value of the smoothing constant affects the results, see this old post.

In this recent post I show why, for one graph of monthly data, I abandoned my usual monthly constant of 14400 in favor of my default annual value 100. The larger value smoothed all the information out of the result -- like Kruger smoothing the head off a statue on Seinfeld.

For quarterly data, Ravn and Uhlig say, the value 1600 is commonly used. (It seems 1600 is the value Mr. Hodrick and Mr. Prescott used in the article that introduced the HP filter.) But for annual data Ravn and Uhlig note four different values: 100, 400, 10, and their own personal favorite, 6.25:
We then show that our recommendations work extremely well on U.S. GDP data: using a value of the smoothing parameter of 6.25 for annual data and 1600 for quarterly data produces almost exactly the same trend. This leads us to reconsider the business cycle “facts” reported in earlier studies. As an example, we cast doubt on a finding by Backus and Kehoe (1992) ...

Using a constant of 6.25 rather than 100 for annual data, the authors produce "almost exactly the same trend" that they get for quarterly data with a constant of 1600. Sounds good. But let me ask: Do you really want to get the same HP trend for the two series on this graph?

Graph #1: Quarterly (blue) and Annual (red) RGDP
I don't think so. The quarterly trend should show more detail.

Maybe if I'm looking at 50 years of data I'd want to think of "the" trend for the data, and it should be the same for both annual and quarterly. But if I'm focused on only a few years of data, it's probably because I'm looking for more subtle differences and I'd want to see more wiggle in the trend for the more wiggly line. I'm not looking at the 50-year trend. I'm looking at what's happened since the crisis.

Sometimes there's good reason to want the same trend from data with different frequencies. But sometimes there's good reason to want different trends.

Here are the two graphs from my "recent" post linked above, and my thoughts at the time:
Today I want to look at the monthly GDP data from Macroeconomic Advisers. I have their data thru May now:

Graph #2: Monthly RGDP since Jan 2009 with an Unresponsive H-P Constant
The blue line shows RGDP growth from 12 months prior. The red line is the Hodrick-Prescott using the constant I'd normally use for monthly data. I think this constant makes the red line a little too unresponsive, there being only about seven years of data.

Here is the same graph with a more responsive Hodrick-Prescott:

Graph #3: Monthly RGDP since Jan 2009 using a More Responsive H-P Constant
Now the red line follows the blue more closely. It helps us see the up-and-down pattern in the jiggy blue data. We are at a low spot now, and evidently RGDP growth has been trending down since the end of 2014.

Graph #2 shows a trend that is essentially flat. Graph #3 shows that RGDP growth has been trending down for a year and a half. Graph #3 is much more informative, and its trend line clearly does a better job of showing the path of RGDP than does Graph #2. But the thing of it is, Graph #3 uses the "wrong" smoothing constant. It uses my default annual value on monthly data!

I chose the annual value on purpose, because I wanted about as much smoothing on the monthly data as I normally get on annual data. It worked.

PS: I went out of my way to find monthly data for Graphs #2 and #3. Monthly, because it shows more variation. It would make no sense to smooth the monthly trend down till it showed as little variation as the annual!

Tuesday, August 16, 2016

The Relation Between Debt Service and RGDP Growth

Showing Hodrick-Prescott curves for Household Debt Service and Real GDP Growth:

1980-83: A low level of debt service (blue) allows rapid increase in RGDP growth.
1984-86: Rising debt service hinders growth. The red line peaks and turns downward.
1986-89: Reduced growth and reduced borrowing allow debt service to fall.
1990-93: Falling debt service encourages growth. The red line bottoms and turns upward.
1994-97: Rising growth leads to increasing debt service.
1997-99: Rising debt service hinders growth. The red line peaks and turns downward.
2000-06: Reduced growth and increased borrowing keep debt service rising.
2006-08: Reduced growth and reduced borrowing allow debt service to fall.
2008-10: Falling debt service encourages growth. The red line bottoms and turns upward.
2010-13: Falling debt service encourages growth. The red line rises.
2013-16: The fall in debt service slows. The increase in growth also slows.

Easier to follow if you print out the page and mark it up.

// The Excel file (contains VBA code)

Monday, August 15, 2016

Debt service and economic growth

Some generalizations:

Debt service is a financial cost imposed on the non-financial sector -- on production and consumption. To the extent that debt service moves money from the non-financial to the financial sector, it hinders production and undermines consumption.

The data on debt service is for households. So I will confine my remarks to household expenditure, to consumption.

When debt service is low, consumers have more money left over for other things. When debt service is high, consumers have less money left over for other things.

When debt service is rising, it is an indication that consumer debt is high or that consumer borrowing is on the rise. If borrowing is on the rise, GDP is probably on the rise. If debt is high, maybe not.

When debt service is falling, it is an indication that debt has been falling or has been rising more slowly than usual. If the latter, GDP may be on the rise. If the former, maybe not.

Sunday, August 14, 2016

Reverse Engineering the Household Debt Service Ratio

Looking at household debt service as a percent of disposable personal income. Wondering how income affects the data, how debt affects it, and how saving affects it.

At FRED, the debt service number is given as "percent of disposable personal income". I figured saving the same way, and put the two together on a graph:

Graph #1: Debt Service (blue) and Household Saving (red) relative to DPI
My first impression was that the two lines tend to move in opposite directions: away from each other before the 1982 recession, but toward each other after it... away from each other after the 1991 recession and then toward each other again... and then away from each other from the late 1990s to the crisis, and then toward and past each other.

So then I took and put a minus sign in front of the formula for the red line, to turn it upside down:

Graph #2: Debt Service (blue) and Saving with a Minus Sign (red)
With the red line other-side-up, the red and blue do show signs of matching. I got low and then high from 1985 to 1990... low and then high again... then higher from 2000 to the crisis, and then low. The lines don't match well, but do show signs of matching. Makes me wonder what I can do to make them match better.

I thought about adding consumer debt (relative to disposable personal income) to the calculation of the red line. And then I remembered seeing those formulas that add two terms together after assigning a weight to each. And I said Yeah I can do that.

Ended up using "change in household debt" relative to DPI. And I just guessed some weights until I got something that looked somewhat like the debt service line. And I added a constant to push the red line up closer to the blue. Here is the result:

Graph #3: Using Savings and Household Debt to Reverse Engineer the Debt Service Ratio
Oh, was I pleased with myself!

The red line is very jiggy compared to the blue. But I expect I can smooth it out by using a Hodrick-Prescott calculation on it. Now I'm picturing a red line that looks even more like the blue, except everything happens about two years early. So I can lag the red line about two years to  make the lines more similar.

What that means is I'll get a peek into the future. My most recent data, from Q1 2016, will show up as a prediction of debt service for 2018! Now it's getting interesting.

I have to bring the data from FRED to Excel to do the lagging and the Hodrick-Prescott. I got all the data, which will let me calculate a number back to 1952 instead of 1980. I love this stuff.

Long story short, I figured the Hodrick-Prescott, tweaked the "weight" values a bit (by eye), and lagged my calculated numbers 8 quarters. Here's the result, back to 1980:

Graph #4: Not Perfect, but Not Bad!
The two-year lag is just about perfect for the high area (from 2000 to 2009). For the early years the red line should be lagged a little less.

But look at that big drop after 2008 Q1. At the peak the red line is a perfect match to the blue. At the bottom, by 2012, the lines have crossed. The economy slowed down during the big drop. I would need a longer lag there at the bottom, to push the red line more to the right.

But the most interesting thing on this graph, I think, is that it makes a prediction about the path the debt service ratio will take. It's going to go up. The blue line, like the red, is going to go up.

// The Excel file

Saturday, August 13, 2016

What I said eleven months back

Here it is. I finally found my old post from September 2015. Almost a year ago.

I'm gonna chop off the introductory stuff, revise what's left to eliminate references to the introductory stuff, and present it again:
2½ Times Longer and 2½ Times Deeper

Suppose we take the "Household debt service" graph and stretch the X axis a few years into the future. (FRED lets me do that, but when I capture and display the link, it reverts back to 2015.) Anyway, here's a screen shot:

Image #1

Next I change the blue line to dashed red, turn off the recession bars, stretch the graph out to the year 2030 again, and take another screen shot. Then I load it into Paint.NET, use the "magic wand" (at 20% tolerance) to erase the white background, and save the file as a GIF:

Image #2

Then I crop it so it shows just the 1990s, and put a black border on it. Actual size:

Image #3

Then I open up Image #1 so I have two files to work with in Paint.NET. I create a new layer in Image #1.

I switch back to the cropped GIF file, make it twice as big, no, 2½ times as big, and copy it.

Then I switch back to Image #1, paste the enlarged cropped GIF into place, "flatten" and save the file, and show you what I got:

Image #4

See how the first few years of the dotted red overlay follow very much the same path as the blue line after the 2007 peak? This doesn't guarantee that anything will happen tomorrow, of course. But the striking similarity of the two downtrends makes me think that the uptrends might also turn out to be similar.

So we may soon get a few years when the economy seems pretty good again, while debt races upward even faster than it did before the crisis -- if we have the stomach for that. When those good years come, you know whatever political party is in power will claim all the credit.

Right now, though, nobody expects a few good years.

Funny thing. I had to cut off the last sentence. This was the original ending:
Right now, though, nobody expects a few good years.

I don't either.

I didn't expect the good years back in September of 2015, but I do now. So that last sentence had to go.

Friday, August 12, 2016

Bill McBride: "I don't expect a downturn for employment any time soon"

Bill McBride:
This graph [at Calculated Risk] shows the job losses from the start of the employment recession, in percentage terms, compared to previous post WWII recessions. Since exceeding the pre-recession peak in May 2014, employment is now 4.3% above the previous peak.

Note: I ended the lines for most previous recessions when employment reached a new peak, although I continued the 2001 recession too on this graph. The downturn at the end of the 2001 recession is the beginning of the 2007 recession. I don't expect a downturn for employment any time soon (unlike in 2007 when I was forecasting a recession).

Bill McBride's graph says no downturn. Sounds good.

Thursday, August 11, 2016

The secret

Base money, 1918-2016, on a log scale:

Graph #1: Base Money (AMBSL) on a Log Scale
I marked it up by eye so you can see what I'm looking at:

Graph #2: General Trends in Base Money Growth
There's a flat in the 1920s and a flat in the 1950s and a flat again now. Lots of people like the flats, because no inflation. But that is over-simplified. There is something more to be said.

When base money runs flat for a while, as it did in the 1920s, conditions arise which require massive increase, as in the 1930s and '40s. After that increase, base money can run flat again for a while, as in the 1950s.

Another example: Money took a turn around 1960 and started to increase. It increased for a good long time. Then around the year 2000 somebody took it into their head to gradually flatten out base money again. They created a nice, gradual curve to achieve the flattening. Then, all of a sudden, conditions arose which required massive increase. Just like in the 1930s.

(For the record: Yes, it was the flattening that created the crisis.)

After you get massive increase, base money can run flat for a while. That's what happened after World War Two. It is happening again now. We know it worked after World War Two. We don't know if it will work now. I think it will work, and predict vigor. Everyone else seems to think otherwise.

That's fine. I don't know what will happen. I'm reading my graphs and saying as loudly as I can what I think the graphs are telling me. But I could be wrong. Maybe the massive increase since 2008 wasn't massive enough. Maybe going flat now is bringing the economy down again. Today's graphs offer no hint. Yesterday's graphs tell me the massive increase was adequate.

After you get an adequate massive increase, base money can run flat for a while. But after a while, the lack of increase in base money seems to create big problems that require another massive increase in the money. This is important. This is something we need to look at.

After you get the massive increase, you don't have "too little" base money any more. You get economic vigor -- that word, again -- and economic growth. And then, after a while if you flatten the money (or just leave it flat) you get the big problems and you need the massive increase again.

What's missing from this picture is private debt. The growth of base money and the growth of private debt must match. If this does not happen, if private debt grows faster than base, eventually times get hard. And after that you have a crisis. The solution, which is massive increase in base money, solves the problem by reducing the private-debt-to-base-money ratio.

The solution appears to work reliably. But some people don't like it. They don't want the quantity of money inflated.

Hey, if you want base to run flat, then you have to make private debt run flat. If you want private debt to increase, you have to make base money increase. The secret is that the two must stay in proportion.

The two must stay in proportion. Knowing this, you would want to seek the debt-to-base ratio that gives the best economic growth. And you would then want to change base and debt together, so they grow at a rate that keeps prices stable. That way, you cover all your bases: growth, and price stability.