US corporate profitability has indeed been falling

This note follows up more deliberately on a point I made while rushed after the National Accounts profits data for Q1 were released last Friday.  The apparent decline of US corporate profitability is not an illusion created by losses at the Fed or by the distorting influence of the Capital Consumption Adjustment (CCA).

The Fed losses are indeed a source of noise, but they were largely offset by odd strength in the profits of truly private financial businesses. As a result, overall financial profits have not changed much recently, which has meant that we see qualitatively the same pattern in nonfinancial profits (on domestic operations) that we see in the headline all-sector global profits data.  

Separately, there is a very good reason to include the CCA in the depreciation and therefore profits data. Removing its influence would introduce, rather than correct, an error. Some analysts prefer to look at cash flow, rather than profits.  And not being a strategist, I do not have a view on that. But there is no point in torturing profits so that they look more like cash flow. Deprecation detracts from profitability for obvious reasons. We might as well try to measure that fairly. 

In the remainder of this note, I will focus on how the CCA has affected profitability in the domestic operations of nonfinancial corporation, where we have a nice clean data set.  So, I exclude financials, which are not central to the argument I will make here. And I exclude the overseas operations of all US corporations, because of data constraints.  This leaves me looking at 73% of total global pre-tax profits as of the first quarter.

Before I turn to the data, which I will rip through quickly, let me mention two ways I adjust the data before even getting into technical issues central to this note.  First, I adjust the headline profit data to net out the distortion caused by the swing in net indirect taxes occasioned by subsidies related to the Covid shock. To do that, I run a counterfactual in which the net indirect taxes are held constant as a ratio to gross value added for all periods after 2019 Q4. That has little effect on the measured level of profits as of Q1 because the net indirect tax rate has almost full renormalized, as you can see from the chart above. But it affects the path, shaving off the peak and making the cyclical profit decline a more recent development. I call this adjusted series “headline” because I want to focus on other adjustments and don’t want to refer to adjusted. 

Second, when calculating various margins, I smooth out the volatility of nominal gross and net value added during 2020 by assuming a steady growth rate there between the fourth quarter of 2019 and the first quarter of 2020. This has no effect on the profit figures, but it makes the denominator in the various margins less noisy and more meaningful. 

With that in mind, pretax profits are down 6 ½% (not annualized) during the past two quarters, but they are down only 2% if we exclude the influence of the CCA. So, this is obviously an important issue quantitatively.

But why would we exclude the CCA, unless we are really just interested in cash flow, which is a separate discussion? Nominal value added in the corporate sector has been lifted by the steep rise of the general price level (i.e., inflation) during the past few years, so why would we not want to correct the (historical cost) nominal depreciation charges for that same inflation, as the CCA does, although admittedly imperfectly?  Note the bottom left panel of the chart above, that including the CCA in the profits calculation has led to a smaller overshoot at recent business cycle peaks and that the CCA exclusive measure of margins has collapsed to the less volatile CCA inclusive measure on each of the most recent couple occasions. Why would we want to miss that? If anything, we would want to port this issue over the company level data. Are reported and operating earnings controlling for this issue properly or are they overstating earnings?

Or just look at deprecation charges, measured as a “margin” to gross value added with and without the CCA, as in the chart above. The CCA inclusive measure is less noisy and more intuitively sensible because it is, well, the more accurate measure. Depreciation charges have an upward trend, because of shortening capital lives related to tech advance, and they tend to rise above trend late in the cycle.  The correct data reflect that, while the wrong data do not. 

Finally, if we are going to nitpick the profit data, let’s do it properly!  The oddity in the National Accounts data is not the CCA, but the bizarre and seemingly unrelenting decline of the net interest margin, despite rapidly rising corporate debt (relative to value added) and the recent turn higher in interest rates. Let’s conservatively net that effect out by running a counterfactual in which the ratio of net interest to value added is held at its end 2018 level. That brings the profit decline over the past two quarters from 6 ½% to 9%. If anything, this likely understates the issue, because I am being conservative to make the point directionally reliable.  Barring a recession, which would be its own issue for profits, the net interest margin is fated to rise steeply. I am confused why it is not yet in the data, to be fair. 

Profits have been falling, which is hardly surprising, given the cyclical influences on them just now.  In fairness, stripping out the CCA is not as kooky as relating the discount rate in a t-bill with ten days to run to “default risk.” That bit of journalism sticks with ya!