Camera bug

My wife took this picture at my request. That's the second woolly bear I saw in September, and both of 'em had the long black band in front, and a short black band at the rear.

What does it mean? Maybe that the winter starts out rough, but spring comes early?

Saw a third one just the other day, and he had the same pattern: long in front. Except, he didn't move at all, so I'm not sure it was the front.

A couple years ago all of 'em had tiny short black ends, both ends, and that was a very mild winter around here. It's funny how the color pattern seems to vary uniformly, different patterns in different years but great similarity in any one year. (If you can find "great" similarity among only three bugs.)

Kevin Myatt's Weather Journal recently had a picture of a woolly bear that shows the same long black slash short black pattern. From a comment by TQ on that post:

How to Forecast with Banded Woolly Bear Caterpillars
Long-range winter forecasters who use the banded wooly bear caterpillar look at the width of the black stripes on the worm's front and back and the ratio of black-to-orange.

With that information in hand, the following forecasting rules apply:
- If the black stripes are narrow – defined as less than half the worm's length – then the winter will be mild.
- If the black stripes are thick – defined as more than half the worm's length, then the winter will be cold or harsh with harsh being undefined.

Some banded woolly bear caterpillar forecasters can tease out additional information by looking out the difference between the front and back black bands:
- If the front band is larger than the back band, then the first part winter will be colder or harsher than the last half.
- If the front band is merely dipped in black, that portends a mild first half, whereas if the back band is merely dipped in black, that portends an early spring.
- If the caterpillar is mostly black, then winter will get cold early and there will be a lot of precipitation.
- If the caterpillar is all black, then the winter's snowfall will be light.
- If small brown spots are present, then the winter's dominant precipitation type will be drizzle

—
The 2013-2014 forecast!
Overall mild
Start of winter will be more harsh than the end of winter.
Early spring.

I know my dad kept an eye on the woolly bears.

I wonder if there was some cost that suddenly started eating up corporate profits, that caused a great moderation in the growth of real investment

Graph #1: Corporate Profits (blue) and Corporate Interest Paid (red)

(Pssst... This is a follow-up to yesterday's post.)

Another oddity in the annals of capacity utilization

Graph #1: Growth Rates of RGDP (blue) and Capacity Utilization (red)
(Click the graph for a bigger picture in a longer time period)

The 1982 recession was followed by an uncharacteristically large increase in RGDP growth (blue) and an unusually small increase in Capacity Utilization (red).

Hm. If capacity increased fast enough, RGDP might increase faster than we could put the new capacity to use. I wonder: Did capacity grow particularly fast after the 1982 recession?

By the standard of later years at least, it did:

Graph #2: Growth Rate of Investment

So maybe that's it: A vigorous bloom of new capacity allowed economic growth to increase faster than capacity utilization. But the vigorous bloom did not last. There followed almost immediately a great moderation of investment spending.

It looks to me as if the brief bout of vigor was brought on by great expectations, and the end of the vigor by expectations disappointed. And perhaps we know the source of that disappointment -- disappointing profits:

Graph #3: Corporate Profits relative to GDP

There is no more certain extinguisher of expectations, than reality.

I didn't notice until now

Graph #3 from mine of 18 Feb 2011:

Graph #1: The Declining Peaks of Capacity Utilization

In that old post I pointed out that since the 1960s, the high points of capacity utilization "have come at progressively lower levels."

I also pointed out the "two exceptionally low peaks in the 1980s that I did not mark with red lines, as they do not fit the down-stepping trend. These two severe lows are associated with the Volcker squeeze."

But I did not point out the three peaks that share the next-to-last red line. I noticed it, but I didn't know the reason it happened.

First of all, as I now think, it is not really three peaks. It is two peaks. The first of the two -- the late 1980s peak -- is comparable to those that came before. The second is different: It is wider than the others; in other words, longer-lasting. And it is higher than the downward trend would lead us to expect. (You can move your mouse on and off the graph to compare the actual level with the downstepping pattern of decline.)

Something made this one peak longer and higher than the other peaks would lead us to expect. What could have caused the difference?

The timing of the extended peak is a significant clue. It first reaches a maximum in late 1994. It tops out again in late 1997. This extended high in capacity utilization matches the early years of the so-called "macroeconomic miracle". And the macroeconomic miracle happened just after a unique decline in the growth of private debt and an unusually rapid growth in the quantity of money prepared the way.

Capacity utilization, then, is one of the data series that show improved behavior in response to the early-1990s decline in the debt-per-dollar ratio.

Popularity Contest

The graph summarizes the first ten items that came up on 28 September when I searched FRED for debt and got a list sorted by popularity.

http://research.stlouisfed.org/fred2/tags/series/?t=debt

Marcus responds

Marcus Nunes of Historinhas, who does awesome graphs, took the time to reply to my Having it both ways. He made such a nice presentation that it seemed a waste to keep it hidden away in my email. So with Marcus's permission, I'm posting his response below. Even though he tells me that I "got it all wrong!"

I've numbered his graphs (because it helps me think) and deleted his opening greeting. Everything else is unedited and strictly his.

Give me a few days to get my thoughts together for a reply.

Here's Marcus:

Art

Basically, you make the mistake you accused me of making: “chopping”. You “eyeballed-in a trend line” from data covering 1975-1979 and said “Marcus assures us that we´re looking at is not an employment gap but the very beginning of a long boom”. In the process you “solved my problem” (I´m worried about nothing)!

Your Total Non Farm chart with trend is misleading! The chart below shows the level and trend of employment (NFP) from 1955 (that´s the usual start date to ‘take out’ post Great Depression+WWII and Korea War ‘adjustment).

Graph #1

The next chart shows the labor force participation rate both total and women´s participation rate.

Graph #2

Note the impact of the ‘feminist movement’ of the 1960s and 1970s in increasing the LFPR of women (men´s LFPR declines throughout). It peaks in the late 1990´s and overall LFPR falls pulled down by the’ secular’ decline in men´s LFPR. The more recent drop is mostly attributable (I think) to the “Great Recession”.

I still believe it´s true that “Ed Lambert graduated at the ‘right’ time, just as the Volcker adjustment took effect, ‘eradicating’ inflation from the system and stabilizing the economy, paving the way to the Great Moderation.

The next chart depicts RGDP and trend from 1955 onwards. Do you note the “Great Recession” (“Lesser Depression”)?

Graph #3

In growth rate form:

Graph #4

In what ‘world’ do you prefer to live: In “boom-bust” or in “great moderation”? (Remember the mean growth in both periods is the same).

The next chart shows Nominal GDP growth:

Graph #5

A rising growth trend (“Great Inflation”) is indicative of a non-stationary series (means and variances are not ‘identified’). Note that during the “Great Moderation” NGDP growth becomes stationary (the mean growth is 5.5%). Note how ‘off the charts’ was the fall in nominal spending!

The following chart shows the level of NGDP and its trend from 1987 onwards (the period of the “Great Moderation”). This cannot be done for the previous period because the data are not stationary.

Graph #6

Since in my post I showed only a portion of the period (from 2004) you thought I had only done the trend calculation from that point on! So you thought you were ‘replicating’ the exercise when you chose the 1975-79 period to establish the employment ‘trend’ in your chart.

Again, you got it all wrong!

“There must be some way out of here,” said the joker to the thief

Pictures of debt, and the questions that are raised

After I wrote this morning's post I came across Krugman's Debt History from 2011:

As part of that long-term project, I’ve been looking at the long-term debt history of the United States, using both the Fed’s flow of funds data (which start in 1952) and the earlier, not quite comparable, data from Millennial Historical Statistics. Here’s what I think is a key chart; it shows nonfinancial private-sector debt as as percentage of GDP:

Krugman points out that the 1929-1933 spike is the result of collapsing GDP, not growing debt; "call it Irving Fisher’s revenge," he writes. He doesn't point out that the same thing could yet happen to us.

And he points out "the U-shaped trajectory over time." Kind of a key point.


The most recent comment below Krugman's post is a complaint about the discrepancy between the two series:

So, nobody wonders why the new series states indebtedness considerably lower than the old series (about 40 % of GDP at the end of the old series)? Looks a lot like making the figures appear better than they are.

A completely wrong focus. First of all, it's off the topic of Krugman's post. Second, the commenter has probably never run across the data discrepancy before, or not many times, for no one would express that complaint who is familiar with the data. To me this suggests that the commenter has spent little time looking at pictures of debt.

Maybe I find the comment irritating because I don't like the discrepancy, either. But jumping to the conclusion that they wanted to make the figures look better than they are is less the style of Evel Knievel than Wile E. Coyote.

Handling a discrepancy

You often see me show debt in two mismatched time series, as in this recent graph:

Graph #1: Accumulated Debt relative to the Quantity of Base Money

Here's how other people handle the mismatch:

Looks like they made it match, huh.

Well yeah that's exactly what they did: They made it match. Read the fine print below the graph. (You can click the graph to see a bigger version.) What the fine print says is, "the Historical Statistics of the United States series is scaled (down) to match the Flow of Funds data."

Their method gives a graph that doesn't raise questions the way mine does. But I think it is their scaling of given data to obtain a convenient similarity is the thing that ought to be questioned.

Two different ways of measuring debt? That doesn't bother me at all.

(Off topic, but why they always show debt relative to GDP is beyond me.)