Arthurian Profit Study 1.0

Corporate Business Profit is not the same as Corporate Profit. It's not; I doublechecked. I don't know what the differences might be. But on both these linked graphs (quarterly since 1947 and annual since 1929) corporate profits by the end are a little over 2000, and corporate business profits are a little under 1600: Less than 80% of corporate profits are corporate business profits. And that's the way it is, as Cronkite used to say.

While I'm checking things... I have been using "corporate business profits" minus "nonfinancial corporate business profits" as a measure of "financial corporate business profits". But while I was making the graphs linked above, a FRED search turned up Domestic financial sectors; corporate profits before tax excluding IVA and CCAdj, Flow, which should be just what I need: profits of financial corporate business.

The word "flow" on the end of the series name troubled me; but Wikipedia points out that

Some accounting entries are normally always represented as a flow (e.g. profit or income)...

So if I compare my calculated measure of financial profit to this new series I found, it's a flow-to-flow comparison. Good! (If it was stock-to-flow the numbers would certainly differ.) Here's the comparison graph. I'm happy to say the numbers match:

Graph #1: My Calculation (blue) and the FRED Series (red)

My calculated values are displayed as a wide blue line. The FRED series is a narrow, bright red line. The red line is neatly centered on the blue, so the values are identical, or very nearly. To see a bigger version, click the graph. Or click the "Graph #1" text to see it "live" at FRED.

This comparison gives me confidence in my calculation, confidence that FRED's "domestic financial sectors" series would have been the right one to use, and confidence that I'm correctly interpreting the difference between "corporate profits" and "corporate business profits".



I'm so full of confidence now that I'm not even going to use FRED's financial profit series. I'll stick with my calculation instead. Convenient, because I have another half dozen graphs ready to go. First is the financial share of corporate business profits: financial as a percent of total corporate business profits.

Graph #2: Financial Business Share of Corporate Business Profits

There's no straight-line path here, but there is a general upward trend. The financial share runs from less than 10% in the late 1940s to about 25 percent in recent years. This tells me that the nonfinancial share was more than 90% in the late 1940s, and that it has since worked its way down to about 75%.

Be aware, however, that the share of profits going to Nonfinancial Corporate Business (NCB) arises in part from nonfinancial activity and in part from financial assets. Some of the 75 percentage points of profit accruing to NCB are, to my way of thinking, financial (F) profits. As opposed to profits arising from productive activity and the generation of output.

I don't know what the numbers are, but let me give an example. Let's say 50 of the 75 percentage points of NCB profit arose from productive business activity. That means the other 25 points of profit arose from the financial (nonproductive) business activity of those nonfinancial business corporations.

To my way of thinking, these 25 points of profit should be counted with the "about 25 percent" that accrues to financial corporations, for about 50% in total. This would mean that corporate business profits arise half from the productive activity of NCB corporations, and half from the financial activity of corporate business in general. Half and half. Of course, the 25 points we're looking at is not the actual number. It's not even an estimate. It is only an example.

As we saw on Graph #2, financial is gaining on nonfinancial when it comes to shares of Corporate Business (CB) profits. But financial is also gaining on nonfinancial when it comes to the assets of nonfinancial corporate business. What's good for the goose is good for the gander, so I have to assume that for nonfinancial corporate business, their profit from financial operations is gaining on their profit from nonfinancial operations.

If I'm right about this, we can be certain that profits from financial operations are rising faster than we saw on Graph #2. We can be certain that corporate business profits arising from financial operations are increasing faster than those arising from productive business activity.

The partitioning of corporate business into "nonfinancial" and "financial" may be accurate as far as the standardized nomenclature of "business types" is concerned. But that type of partitioning is not an accurate way to measure profits arising from financial vs nonfinancial business activity.

Does it matter? Yes, certainly. If financial profits are gaining on nonfinancial, financial business is the better business to be in. Capital moves out of productive business and into financial business. Finance continues to grow, while productive business continues to tread water and productivity continues to disappoint.


For my next graph, I duplicated Graph #2 in Excel and added a Hodrick-Prescott trend line. The H-P may be out of favor these days, but I was last in line to start using it, and I'll probably be last in line to stop. Anyway, I used it:

Graph #3: Same as Graph #2 but with Hodrivk-Prescott Trend (red) (and in Excel)

I have in mind to figure an average level for regions between the high points of the H-P, and look at those average levels relative to economic performance. But it's all just a foggy notion right now. I did take the H-P trend line and put it on its own graph:

Graph #4: The H-P Trend line from Graph #3

Okay, so that's the "trend" of the financial profits data we were looking at. And next -- no one ever showed me how to do this so I might have it wrong, but I don't think so -- next I took the original data (the blue line on Graph #3) and subtracted the trend line values from it. If I'm doing it right, this gives me the "detrended" data.

Where the blue line is above the red on Graph #3, the blue is above zero on Graph #5. Where blue is below red on #3, it is below zero on #5. All the differences visible on #3 are visible on #5. But instead of being different from the wavy line on Graph #4, they are different from the zero level:

Graph #5: The Trend (graph #4) Subtracted from the data (graph #1)

Graph #6

Graph #7

What these interesting patterns might possibly mean, I have no idea. Interesting patterns often mean nothing. And hey, I haven't spent a lot of time looking at the financial business share of profits. Besides, this is the first time I've looked at a detrended anything. So I have no clue if these patterns mean something. I just thought they were interesting. That's allowed, right?

I have one more variant of the detrended financial business share graph, one more that we need to look at -- a copy of Graph #5 with recession bars:

Graph #8

The blue line has a peak at every recession.

"But," you will point out, "the blue line also has peaks in several places where there is no recession."

Indeed. If we were using this graph to predict recessions, it would be no more reliable than a yield curve chart! But we're not using our graph to predict anything. I had in mind to use it to see what happens to profits over the course of the business cycle. But I'm disappointed: It doesn't look useful for what I have in mind.

I expect to see the Financial share of profits rise before recessions and fall after, while the Nonfinancial share falls before recessions and rises after. Back in January I looked at corporate business profits and said

Early in the expansions that follow recessions, profit of non-financial corporations typically rises (as a share of corporate business profits). Later in the typical expansion, when concern about inflation arises and interest rates go up, rising financial costs drive non-financial profit down while financial profit rises.

But I'm not seeing that pattern in the detrended graph. I think detrending removes the information I need.

//

I changed the H-P smoothing factor from 1600 (for this quarterly data) to 100. The H-P trend changed, so the detrended graph changed:

Graph #9: Like Graph #5,  but with H-P Smoothing Factor=100

The "interesting" features have disappeared: The expanding cone of Graph #6 doesn't seem to be there; and the "bunchiness" of Graph #7 is different now. Those interesting features apparently arose not from the profit data but from the amount of smoothing applied by the H-P calculation.

On the other hand, the peaks of the blue line on #9 are more closely tied to recessions. To my eye, only the 1987 peak now stands out as unmatched to a recession.

Meanwhile, the reduced smoothing produced a possibly more useful H-P trend line:

Graph #10

In the present post I am looking for a way to test and evaluate the behavior I noticed in January, with financial profits falling (and nonfinancial profits rising) during recovery from recession, and with financial profits rising (and nonfinancial profits falling) on approach to the next recession. That pattern is somewhat visible in the blue line on Graph #10. I was thinking that using the red line could provide a less subjective indication of the change from decline to increase and of the timing of that change.

The lows following the 1949 and 1954 recessions are clear, but the red line shows no low between the 1958 and 1961 recessions. And then the low after the 1961 recession is clear, but the low after the 1970 recession doesn't occur until after the 1974 recession!

I still need a better indication of the timing of changes from decline to increase. Maybe I can tweak the smoothing factor more and come up with that better indication. For now, though, ten graphs in, that's about all I can take for one day!

//

See also: A rising cost that can turn expansion into recession