2023-04-04 07:35:00 ET
Summary
- You can find evidence in the inflation numbers for almost any proposition you choose and there is always uncertainty in the analytical effort.
- We tend to look at monthly and annual inflation. This is unfortunate – a month is really too short a time period; a year is probably too long.
- If we look at quarterly inflation, we get a much better picture of trends and it now looks like we are getting down around 4%.
- Past inflationary episodes almost always involved much higher energy prices. These devastated the U.S. economy because we were net importers of energy and the higher prices sucked money out of the U.S. economy.
- This time, we are a NET EXPORTER of energy, and this may make an important difference on the overall macroeconomic impact of the higher prices and the consequent interest rate increases.
Co-produced with Philip Mause.
I worked for years in a large law firm, and we poured over each month's financial report and then spent a lot of time on the annual report. The problem was that a month was such a short period of time that the results could be distorted by a big collection or expense that happened to land on a certain day. On the other hand, by the time we pulled together the annual numbers, it was already 14 or 15 months after some of the events we were reviewing. One day it occurred to me that we might learn something from corporate America and take a look at quarterly numbers instead.
Monthly Inflation Numbers
"It is the best of times; it is the worst of times...." Inflation statistics seem to offer evidence of a wide variety of prognostications. We can look at different measures of inflation, different subsets of the data, and different trends and conclude almost anything we like at a time like this. Every month there are two releases - the CPI and the PCE. It is said that the Federal Reserve gives more credence to the PCE, but the media seems to trumpet the CPI numbers a bit more loudly. Within each of these indices, we also have "headline" and "core" (excluding food and energy) inflation. We are told that the Fed focuses more on the core data because it is more stable. We also have "trimmed mean" data which excludes extreme results. There still is something to be said for the oft-derided "headline' numbers. If you come to think about it - at a time like this when higher wages seem to be a key issue of concern and a potential inflation driver, it may be that headline inflation is more important. After all, employees have to buy food and gasoline and their total monthly expenses (better captured by "headline" numbers) may play a key role in wage demands.
A vector of analysis that is too often neglected is "period selection" - or the choice of the time period over which to measure inflation. We seem to be limited to monthly and year-over-year analysis and this can create some problems. A month is probably too short a time period to generate meaningful data. The table below provides the monthly change in headline CPI data and also provides an annualization of that number (the annual increase that would occur if that monthly number persisted for 12 months).
Month | Monthly HL CPI | Annualized |
Jan 2022 | 0.6 | 7.2 |
Feb 2022 | 0.8 | 9.6 |
Mar 2022 | 1.2 | 14.4 |
Apr 2022 | 0.3 | 3.6 |
May 2022 | 1.0 | 12.0 |
Jun 2022 | 1.3 | 15.6 |
Jul 2022 | 0.0 | 0.0 |
Aug 2022 | 0.2 | 2.4 |
Sep 2022 | 0.4 | 4.8 |
Oct 2022 | 0.5 | 6.0 |
Nov 2022 | 0.2 | 2.4 |
Dec 2022 | 0.1 | 1.2 |
Jan 2023 | 0.5 | 6.0 |
Feb 2023 | 0.4 | 4.8 |
These numbers are striking because they seem to jump around quite a bit from month to month. Now, no one seriously believes that inflation dropped from an annual rate of 15.6 in June to an annual rate of 0 in July – but that's what the monthly numbers suggest. So it is obvious that an overemphasis on monthly data can be very misleading.
What is a bit less obvious from the table but is still a problem is that the annual numbers also are suspect. The annual numbers involve comparing prices a year ago with prices today. But, as the table indicates, we had 3 very big monthly increases in March, May, and June last year (this time April wasn't "the cruelest month"). As long as those numbers are baked into the year-over-year total, annual inflation will basically be 3.5% plus whatever happened in the other 9 months. Even if inflation in the other 9 months was only 0.3% per month, each month – the annual total will still be 6.2%.
Conversely, as time passes and as those 3 very bad inflation months drop out of the annual calculation, there may be a misleading "decline" in year-over-year inflation. This pattern is soon to begin. For example, even if inflation for March 2023 runs at 0.8% – the year-over-year number reported in April (including data from April 2022 through March 2023) will be 0.5% lower than the year-over-year number reported in March (which will include the very bad March 2022 number).
There is no perfect solution to this problem but it is worth trying to get a better measure of the important inflation phenomenon.
Measuring Inflation Quarterly
Corporations report their data on a quarterly basis. All sorts of issues have been noted in this connection but it may provide a third time period over which to analyze the numbers and try to discern trends. A bad month that occurred 11 months ago has no impact on quarterly numbers but quarters are three times longer than months and should be less subject to wild swings.
A) Quarterly CPI Inflation
The first table below looks at annualized CPI data (both headline and core) using the 4 quarters of 2022. The second table provides a rolling quarterly (showing the month in which the rolling quarter ended) analysis with the most recent quarter ending with data from February 2023.
Quarter | Headline CPI | Core CPI |
Q1 2022 | 10.4 | 5.6 |
Q2 2022 | 10.4 | 7.6 |
Q3 2022 | 2.4 | 6.0 |
Q4 2022 | 3.2 | 4.0 |
Here is the rolling quarter table.
Quarter ending | Headline CPI | Core CPI |
5/2022 | 10.0 | 8.4 |
8/2022 | 6.0 | 6.0 |
11/2022 | 4.4 | 4.8 |
2/2023 | 4.0 | 5.2 |
While there is some bouncing around, the data is not nearly as volatile as the monthly data. It is also the case that all four sets of data show a clear downward trend and suggest a current level of inflation of 4% or, perhaps, a little lower.
B) Quarterly PCE Inflation
The next table shows quarterly PCE inflation (both headline and core) for 2022.
Quarter | Headline PCE | Core PCE |
Q1/2022 | 8.4 | 5.2 |
Q2/2022 | 8.0 | 5.2 |
Q3/2022 | 2.0 | 4.8 |
Q4/2022 | 2.4 | 3.2 |
And now rolling quarter PCE inflation.
Quarter ending | Headline PCE | Core PCE |
5/2022 | 6.8 | 3.6 |
8/2022 | 4.8 | 4.8 |
11/2022 | 3.6 | 4.0 |
2/2023 | 4.0 | 4.8 |
The quarterly number for both sets of data show headline inflation very high in the first half of 2022 - this was impacted by oil and also food prices and a general run-up in commodity pricing. In the second half of the year, we had the emergence of a strong dollar and decreasing energy and commodity prices. In both CPI and PCE, core inflation seems to be stickier but also trending down – although the one troubling data set is core PCE for rolling quarter time periods. If you take the most recent quarterly numbers for core inflation from the 4 tables above, they average close to 4 percent.
The trend looks pretty favorable and the numbers do not seem out of control. On the other hand, there is nothing here that suggests that the Fed's work is completely done. What is encouraging is that interest rates have gotten up to the point that - if we take 4% as the current inflation number - they are beginning to be restrictive in the sense that "real interest rates" are beginning to be positive.
Hope for a Soft Landing - the Oil Price Connection
There has been surprise among economists about the strength of the economy and the resilience of the job market. This implies a comparison with previous periods of inflation followed by aggressive Federal Reserve tightening. It is interesting to observe that in virtually every inflationary episode there has been a big run-up in the world oil price. This can, of course, suggest four things – 1. mere coincidence, 2. inflation causes the world oil price to rise, 3. one or more other factors cause both inflation and the world oil price to rise, or 4. the world oil price increase causes inflation.
I am not enough of an econometrician to fully ventilate this topic but I have always believed that there is a strong argument that increases in the world oil price (often caused by political or military factors which do not have independent economic causes) are a substantial factor leading to higher inflation. While retail energy prices can be excluded from inflation by limiting our analysis to core inflation, it is also true that core inflation includes something called "transportation services" and it is reasonably obvious that higher energy prices bleed quickly into higher transportation costs. In addition, the shipment of virtually any product consumes energy and this makes the impact of higher energy prices pervasive in terms of the pricing of goods. As to services, landlords sometimes master-meter and in any case have to pay the costs for heating and lighting common areas so energy costs probably contribute to higher rents. And - as discussed above, employees faced with higher costs of living and much higher commuting expenses are likely to demand higher wages or - in some extreme cases - drop out of the job market entirely.
It thus becomes plausible that the 1973 Oil Boycott, the 1979 Iranian Revolution and oil production interruption, the 1990 Iraq War, the big run-up in oil prices in 2007-08, and now, the 2022 oil price surge due to the Ukraine War all played a big part in leading to inflationary episodes.
In each of the past instances, the oil price explosion occurred at a time when the United States was a massive oil importer. The impact was therefore at least twofold – 1. massive inflation, and 2. huge amounts of money being drained from the U.S. economy to pay higher prices for imported oil. The proceeds went to wealthy oil exporters and it was often remarked that it was difficult to foresee the funds being quickly recycled into the U.S. economy.
This is where a major change has occurred. The U.S. is now an ENERGY EXPORTER. We are a net exporter of the combined category of petroleum and refined products (we are a net exporter of refined products and a net importer of crude oil and the net result is that we are a net exporter of the combined category), we are a growing net exporter of natural gas (of late, LNG going to Europe to replace Russian gas), and we are a significant net exporter of coal (both met coal and thermal coal).
This may make a difference. Instead of having the higher gasoline prices paid by U.S. consumers going overseas, they now go to domestic royalty holders, oil field workers, oil company shareholders, and state governments. This may add resilience to the U.S. economy by giving consumers, businesses, and government more spending power. It will probably continue to reinforce the strength of the dollar which may make it a bit less painful to conquer inflation. And these combined effects may be just enough to make the difference between a moderate recession and a shallow recession or even no recession at all.
Conclusion
We are not completely out of the woods, but the data does not suggest the need for an enormous amount of additional tightening. The fact that we are now an energy exporter may, to a degree, shield the economy from some of the macroeconomic downsides associated with the historic high energy prices / high interest rates cycles. Not all current news is bad.
For further details see:
Quarterly Inflation At 4%, Soft Landing Still Possible