Review of U.S. Persistent Inflation

　　Compared with the global crisis in 2008, after the “COVID-19” pandemic, the monetary policies of major Western economies have been looser, more consistent in pace, and faster in action. The balance sheets of the Federal Reserve, the Bank of England and the European Central Bank are all at record highs. The effect of monetary policy depends on the characteristics of exogenous shocks and the coordination with fiscal and income policies. The impact of the epidemic is mainly negative on the supply side, and the counter-cyclical policy has obvious characteristics of “monetization of fiscal deficit”, which widens the gap between supply and demand and increases inflationary pressure. The extremely loose monetary policy stifled a “perfect storm” in its cradle, but the “K-shaped” recovery of the economy, superimposed on the soaring prices of bulk commodities and risk assets made the monetary policy a dilemma. Japan at the beginning of this century and the United States after 2008 experienced the coexistence of QE (quantitative easing) and deflation cannot be extrapolated linearly.
　　Whether the Fed sets policy rates or market participants decide on asset allocation, it is crucial to clarify whether inflation is temporary or persistent. Generally speaking, the Fed will not change its policy stance because of temporary inflation, and the market will not reprice. The impact of the epidemic and the policy response have brought great disturbances to prices, increasing the noise in inflation data and the difficulty of identifying trends and cyclical fluctuations. In the second and third quarters of 2021, there will be a phased inflection point in inflation, which will enhance the credibility of the “temporary inflation hypothesis” and make the Fed misjudge the persistence of inflation. But this inflection point is not firm. The formation of persistent inflation is not achieved overnight, but the “relay” of multiple temporary factors.
Fluctuations, Trends and Structures of Inflation

　　The indicators commonly used in the United States to measure the price increase of final goods and services include the Consumer Price Index (CPI), the Personal Consumption Expenditures (PCE) price index, and the GDP deflator. They are all obtained by weighted average, and the difference in value is mainly due to the difference in coverage, weight and calculation method. There is an increasing coverage from the CPI to the GDP (gross domestic product) deflator, which captures the prices of all consumers, businesses, and governments within the United States, as well as all final goods and services purchased abroad. CPI and PCE are more common. The former is compiled by the Bureau of Labor Statistics (BLS) and the latter by the Bureau of Economic Analysis (BEA), but the underlying data still comes from the BLS. Both differ in coverage and weight. The Fed is more concerned with the latter.
　　Overall, the PCE inflation rate is basically in line with the CPI inflation rate, with the former being lower in level and volatility. This can be explained from the compilation and calculation methods of the two. On the one hand, the PCE price index has broader coverage, as total personal spending accounts for nearly 70% of US GDP (gross domestic product). It measures the price change of all consumption items, not just items purchased by consumers out of pocket; on the other hand, differences in coverage directly lead to differences in weights, and PCE weights are adjusted more frequently. For example, health services are weighted 22 percent in the PCE and 9 percent in the CPI because the former includes insurance premiums, deductibles, and co-pays paid out of pocket by consumers, as well as insurance, health insurance, and medical care provided by employers. Subsidy expenditures, which cover only direct consumer spending. Another example is housing, whose weight in the PCE is only 23%, while the weight in the CPI is 42%. This means that the increase in the price of health care services has a much greater impact on the PCE price index than on the CPI index, while housing has a greater impact on the CPI. The PCE price index includes the substitution effect caused by changes in commodity prices, which can more accurately describe changes in consumers’ real living costs.

　　Monetary policy is an aggregate policy that affects aggregate demand. The Fed is unlikely to change its policy stance due to price volatility and sharp increases in prices for a single industry (or commodity). So, the Fed is focusing more on the core price index, which excludes the more volatile food and energy from the headline price index. The volatility of the core price index is significantly lower than that of the headline price index (Chart 1). However, the core price index is still indirectly affected by the price of a single industry good or service other than food and energy. How to better distinguish between price fluctuations and trends, structural increases and general increases has become a topic of concern to monetary policy authorities.
Figure 1: Comparison of Commonly Used Inflation Rate Measures in the U.S. (YoY)

Source: Federal Reserve, CEIC, Orient Securities Wealth Research (as of July 2022). Drawing: Zhang Ling

　　To this end, the Cleveland Fed compiled the median inflation rate and the 16% trimmed-mean inflation rate for CPI and PCE. The Dallas Fed also compiled a truncated average inflation indicator, excluding the lowest 24% and highest 31% subjects in the PCE price index (55% censored average inflation), and performing a weighted average among the remaining 45% subjects . The median and trimmed average inflation rates are less volatile than the core price index, the lowest of which is the Dallas Fed’s trimmed average PCE inflation rate, which is used instead of the inflation rate in the Taylor rule. The obtained modified Taylor rule can better fit the trend of policy interest rate. In addition, the proportion of goods with different price increases can also be calculated to examine the structure of inflation (Figure 2).
Figure 2: Comparison of Commonly Used Inflation Rate Measures in the US (YoY)

Source: Federal Reserve, CEIC, Orient Securities Wealth Research (as of July 2022)

　　In the early stages of the impact of the pandemic, both overall and core price increases have declined to varying degrees. Headline price gains have fallen more sharply, with headline CPI and PCE inflation reaching lows of 0.41% and 0.1% in April and May, respectively — readings of 2.5% and 1.9% in January 2020, largely due to crude oil prices rising from 2020 At the beginning of the year, it fell from US$65 per barrel to US$20 at the end of April, and then stabilized due to the agreement between OPEC (Organization of Petroleum Exporting Countries) and Russia to reduce production (about 10% of global production). Until the end of 2020, the inflation rate was significantly lower than before the epidemic. However, starting from the first quarter of 2021, under the combined effect of multiple factors such as the resumption of work and production supported by a low base, loose policies, and supply constraints (supply chain and labor) under the impact of the epidemic, US prices began to accelerate upward.
　　As of the end of June 2021, the year-on-year growth rate of the price index ranked in descending order: CPI (5.4%), core CPI (4.5%), PCE (4%), core PCE (3.5%), censored average of 16% CPI (2.9%), CPI median (2.2%), PCE median (2.2%), 55% censored average PCE (2.0%). The sharp rise in overall prices and the relative stability of trend indicators indicate that inflation does not yet have a broad basis. From a structural point of view, the number of subjects with price increases greater than 10%, 5% and 3% accounted for 11.6%, 32.6% and 45.0% respectively. During the two oil crises, the values ​​in December 1973 were 12.6%, 48.0% and 74.8%, and in November 1978 were 11.2%, 71.7% and 92.1%. The proportions of subjects with price increases of more than 10% at the three time points are very close, but the proportion of subjects with price increases of more than 3% and 5% in June 2021 is obviously low. The structural indicators agree with the conclusions of the trend indicators.

　　In the statement of the regular meeting in April 2021, Powell put forward the “temporary inflation hypothesis” for the first time, saying that the rise in inflation was caused by transitory factors. In the “Monetary Policy Report” in July 2021, the Federal Reserve described the situation of inflation in detail, explained the reasons, and responded to the market’s two views on inflation: (1) Without changing monetary policy, as the economy Returning to normal, inflationary pressures will subside on their own; (2) Inflationary pressures are persistent and will force the Fed to change its policy stance. Inflationary pressures at that time should have been expected by the Fed. In the column, the Federal Reserve demonstrated the “temporary inflation hypothesis” mainly from the perspective of inflation expectations. Because indicators of medium and long-term inflation expectations in multiple dimensions show that the 2% inflation target is still anchored. The FOMC (Federal Open Market Committee) will not abandon this hypothesis until the November 2021 regular meeting (five regular meetings during the period) decides to start Taper.
　　From the rearview mirror, at least in the second and third quarters of 2021, the “temporary inflation hypothesis” has certain credibility, but after a short correction in the third quarter, the trend and structure of inflation deteriorated further, causing the Fed to have to Accelerate the pace of normalization. As of July 2022, although the core PCE and core CPI reached a staged peak in March 2022, the median CPI and the two censored average inflation indicators are still at new highs. The “temporary inflation hypothesis” has become a “stain” for the Fed.
Inflation Relay: Temporary or Permanent?

　　Only after a clear understanding of which goods or services are increasing in price and why can we more accurately judge future price trends. A common analysis idea is to open the “black box” of the price index, split the contribution of subdivided goods or services in the overall price index changes, and focus on analyzing the reasons for the price changes of items with higher weights or larger contributions. Inflation may be temporary if the reason for the price increase is seasonal or driven by an exogenous shock. Since persistent inflation is often a “relay” of temporary inflation, potential price increases should not be ignored.
　　In the CPI basket (Figure 3), the weights of goods and services are about 39% and 61% (by the end of 2021). Compared with the early 1990s, goods have dropped by 5 percentage points and services have increased 5 percentage points. Among the eight industry categories, the top three by weight are: housing (42.1%), transportation (15.7%) and food and beverage (14.8%), accounting for about three quarters of the total. In the PCE basket, housing (22.6%) and health care (22.3%) are the two single items with the largest weight, the former is 19.5 percentage points lower than the CPI, and the latter is 13.5 percentage points higher than the CPI.
Figure 3: Weights of US PCE and CPI Price Indices

Note: The outer layer is the weight of PCE, and the inner layer is the weight of CPI. Source: CEIC, Orient Securities Wealth Research

　　We can first grasp the overall impact of the new crown epidemic on prices from the distribution of price increases. Take, for example, the 121 subdivisions in the core PCE basket. From March 2020 to June 2021, the distribution of price increases experienced a change from normal to left-skewed, and then back to normal. The kurtosis at the end of the period is lower than that at the beginning of the period, and the left and right sides show fat tail characteristics, indicating that there are subjects with large fluctuations, and as time goes by, the fat tail characteristics on the right side are more obvious. There are certain differences in the distribution of goods and services, and the peak and left-skewing characteristics of services are more obvious. This fits the peculiarities of the pandemic shock. The impact is also more pronounced because a large number of service industries require close contact. Among commodities, there are also differences between durable goods and non-durable goods. The former has turned from negative growth (average -1.9%) for 25 years before the epidemic to positive growth, with a year-on-year growth rate of 7.2% in June 2021. This is mutually verifiable with the ever-expanding consumption of durable goods in the early stage of the recovery—consumption of durable goods forms a substitute for consumption of services.
　　The first wave of increases in headline PCE inflation came largely from commodities. The base effect is an important explanation, because the commodities with large year-on-year price increases are similar to those that fell from March to May 2020, mainly energy and transportation (such as new and used cars). In May 2021, the core commodity PCE rose by 8.58% year-on-year, and energy and transportation prices rose by 28% and 20% year-on-year, contributing about one-third of the current inflation—from March 2020 to the end of the year , the increases in the energy and transportation price indices were both negative. The high points of year-on-year growth in commodity PCE and core commodity PCE occurred in January and February 2022, with readings of 11.5 and 15.3%, respectively. By July 2022, they have dropped to 5.6% and 10.8% respectively. In comparison, the upward trend of service industry PCE is relatively slow, and it dominates the trend of core PCE.
　　Logically, if the rise and fall of prices is mainly caused by the impact of the epidemic, it will also return to normal as the epidemic subsides and the resumption of work and production advances. Shapiro, an economist at the Federal Reserve Bank of San Francisco, distinguishes between COVID-sensitive and non-COVID-sensitive subjects from the two dimensions of quantity and price: if the price or quantity of a commodity or service changes from If there are very significant changes from February to April 2020, it is COVID-19-sensitive, and vice versa (Shapiro, 2020a; 2020b). As shown in Figure 4 (below), the subjects with a large decline in quantity basically belong to the service industry. Among them, entertainment, travel, hotels, casinos, catering and air transportation have a decline of 80% to 100%. Air transport and hotel prices also fell sharply, reaching 23% and 13% respectively. In terms of commodities, the items with a decrease in quantity of 40%-60% include: automobiles, jewelry, watches, clothing and footwear. Among them, the price of second-hand cars has dropped by 13%, and the prices of other items have fallen within 10%.

　　As of July 2022 (see Figure 4, above), the prices of most goods and services have returned to before the epidemic. Compared with February 2020, there is still a small negative price gap for: information processing equipment ( -6.68%), video and audio equipment (-4.2%), motorcycles (-3.4%), telecommunications (-6.3%) and insurance (-1.2%), etc. With the exception of a few items such as automobiles (new and used), consumption of goods has largely exceeded pre-pandemic levels. However, there are still a large number of service consumption gaps, such as photography (-49.4%), entertainment tickets (-32.1%), camping (-31.1%), etc. Service consumption is still on the way to recovery.
Figure 4: The impact of the new crown epidemic on consumer behavior – based on the decomposition of the volume and price of PCE items

Note: Bubble size is based on the average of actual spending in July 2022 (above) and April 2020 (below). Source: BEA, CEIC, Orient Securities Wealth Research

　　In core PCE, COVID-sensitive subjects account for two-thirds of the weight and are the main source of price change during the pandemic. Before the epidemic, new crown-sensitive subjects contributed about 0.8-1 percentage point to core PCE inflation, which plummeted to 0.3% in April 2020. Due to the base period effect, the contribution in the second quarter of 2021 increased sharply to 2.6%, equivalent to three-quarters of PCE inflation in the same period. Among the COVID-sensitive subjects, health services and used vehicles (cars and trucks) contributed the most to inflation, but for different reasons.
　　In the five years before the pandemic, the contribution of health services to core PCE inflation was equal to about 0.2 percentage points. The epidemic expanded the demand for health care services, and superimposed the epidemic-related changes in health insurance payment legislation (Shapiro, 2020a), which increased its contribution to 0.6 percentage points in the first quarter of 2021. However, the emergency relief measures are temporary and will act as a drag on price increases when lifted. In fact, the price growth rate of PCE health care services will enter a downward channel in March 2021.
　　Social distancing measures related to the epidemic have reduced demand for public transport and boosted demand for used cars. And because the shortage of chips restricts supply, the widening gap between the supply and demand of new cars has increased the contribution of used cars to inflation from -0.1% before the epidemic to 0.5% in early 2021. According to IHS manufacturer survey data, the shortage of chips will ease after the third quarter of 2021 (but may still be in short supply until early 2022).
　　In the third quarter of 2021, driven by the continued decline in commodity price growth, inflationary pressures in the United States have slowed down. Is the temporary inflation in the post-epidemic era a prelude to sustained inflation in the medium and long term, and will inflation spread from the United States to other countries (or regions) in the future? These issues were unknown at the time. Among the world’s major economies, only the US’s core CPI inflation has exceeded 4%, followed by the UK (3.1%), China’s core CPI inflation and the euro zone’s coordinated inflation are still below 2%, and Japan is still struggling with deflation the edge of.
　　The Fed remains confident in emphasizing the temporary nature of inflation. In a review article at that time, the author also agreed with the credibility of the “temporary inflation hypothesis”, but emphasized that this “temporary” was based on historical attribution and partial analysis, and did not take into account potential inflation in the future. price increase factor. Because in the second half of 2021, with the vaccination and the restart of the service industry, the service industry that once dragged down inflation will become the driving force of inflation, such as aviation and accommodation. The virus is still mutating, which will definitely prolong the recovery cycle of the service industry. The high point of core CPI inflation will appear in February 2022, mainly because core commodity inflation will fall sharply after the inflection point (12.4%). However, affected by the conflict between Russia and Ukraine and the sanctions imposed on Russia by Western countries, energy prices will continue to rise in 2022, and the inflection point will not appear until June. Food and services (core and non-core) inflation did not turn around until August. The former is mainly exogenous, mainly affected by the natural environment, seasonality and the conflict between Russia and Ukraine, while the latter is mainly determined by the increase in wages in the United States. In the short term, the U.S. labor market will remain in short supply, and wage increases will remain rigid, which will support service prices. The interpretation of the “wage-price spiral” is the key to future inflation trends.
　　The formation of persistent inflation does not happen overnight, and different stages often have different driving factors. Currency often plays the role of “threading the needle” in it. The extremely loose monetary and fiscal policies implemented by the United States during the epidemic have greatly increased the disposable income and savings rate of residents, which is the driving force for the continuous rise in prices and an important reason why the inflation pressure in the United States is greater than that of other Western economies.
　　Throughout the period from 2022 to 2023, in terms of policy stance and strength, the Fed needs to strike a balance between insufficient tightening and excessive tightening, and the corresponding policy goal is to suppress inflation at the cost of minimal employment loss. The primary and secondary objectives of the policy switch over time.
　　In the first half of 2022, the endogenous momentum of the economy is strong, the job market is stable, and the supply of energy and food brought about by the conflict between Russia and Ukraine is the only goal of the Fed. The lack of tightening is the main aspect of the contradiction. In the second half of the year, the Fed began to need to strike a balance between its dual missions, but this did not change the contradictory relationship between primary and secondary. In the process of the “Phillips Curve” changing from flat to steep, the relationship between the dual missions has also changed from unity to opposition. This means that the path of inflation convergence to the policy target depends on labor market conditions. The Federal Reserve’s more-than-expected “hawkish” stance is to ease inflationary pressures before the contradiction between inflation and employment completely intensifies. The end of the fourth quarter of this year may be an important inflection point. The index of labor market conditions began to decline, the momentum indicator turned negative, and the unemployment rate began to bottom out. However, core inflation remained stubborn and contradictions began to intensify. In the labor market, the United States will still face the problem of insufficient labor supply in the short term. Wage rises combined with energy and food supply shocks and the resulting inflationary spiral are still the main factors that determine the stance of monetary policy. Medium and long-term inflation expectations are the “last bastion” of the Fed. This time, the Fed’s interest rate cut cycle may significantly lag behind the recession cycle