How to Read a CPI Report

Unwanted news

On February 13, the Bureau of Labor Statistics published his report for the January Consumer Price Index. In response, the US stock market immediately plunged, with the Morningstar US Market Index losing 1.49% that day. You can't always tell why stocks trade the way they do, since stocks don't answer the questions, but this time the cause was obvious: investors didn't like the CPI announcement.

Let's try to understand why. Applying its standard format, the survey began: “The Consumer Price Index for all urban consumers (CPI-U) increased 0.3% in January on a seasonally adjusted basis, after increasing 0.2% in December, The US Bureau of Labor Statistics reported today. Over the past 12 months, the all-items index increased 3.1% before seasonal adjustment.

The annual outlook

Articles about CPI releases often highlight the 12-month change. Thus, Barron The headline covering that report read: "Inflation rose 3.1% more than expected in January." Two months earlier, CNBC had written: "Inflation slowed to an annual rate of 3.1% in November." It's understandable that journalists view data that way. After all, the Bureau of Labor Statistics itself opened a discussion of the December report writing: "Consumer prices for all items increased by 3.4% between December 2022 and December 2023."

Understandable, but useless. The annual figures have two major drawbacks. First, those measures are underway. Because what is removed from ongoing calculations is as important as what was added, changes cannot be interpreted simply by comparing one month's reading with the previous month's reading. Second, while using “all items” is indeed the correct starting point, as it represents the burden faced by consumers, doing so obscures key information.

As a result, the 12-month figure for all articles is analytically irrelevant. The following chart shows that statistic for the last three years.

Nothing in that chart suggests that the January CPI report was unfavorable. If anything, their news appears to have been positive, reversing the previous month's increase. However, not only was the January report mocked, but on the day the December 2023 CPI figures were released, stocks met. Twelve-month figures are illusory. Use them to understand what was, but not to consider what may be, which, in the case of investment prices, is all that matters.

Monthly numbers

Sometimes monthly changes appeared instead (in the case of January, 0.3%). Let's test if they are more useful. The chart below shows seasonally adjusted monthly percentages for all items over the past three years.

Better. According to monthly results, January inflation was the highest in four months. This would seem to justify investors' disappointment. What's more, the last time the monthly CPI increase exceeded 0.3% in January, during the months of August and September 2023, stock prices fell on the days those results were announced. The market largely ignores 12-month results, but pays close attention to month-over-month changes.

Core inflation

Another way to evaluate monthly activity is by excluding food and energy prices. Not only are they volatile, potentially distorting “everything” data with sharp increases one month and then declines the next, but their effects are often outside the Federal Reserve's control. After all, for all its powers, the Fed can't pump more oil or hatch more eggs. Therefore, economists tend to emphasize a statistic they call core CPI, which eliminates those two factors.

In recent months, economists' preferred measure has been less correlated with same-day market reaction than the purported crudest statistic of all. Specific case: although the publication of the August 2023 CPI was widely announced how to enter hottest that expected, causing the Nasdaq composite to lose 1% of its value on the day of the announcement, the core CPI hit its lowest level in two years. Basically, investors were worried about rising gas prices, while economists said: "What worries me?"

Score one for the economists. Since then, the US stock market has gained 14%, which is an attractive annualized rate. The short version of a very long story is that economists generally ignore changes in food and energy prices, consumers never do, and the stock market falls somewhere in between. My sympathies tend towards the former, but I must confess that this mentality is not suitable for a secular rise in commodity prices, as in the 1970s.

(In addition to the basic CPI, economists apply various other adjustments to the calculation of all items, resulting in such exotic items as the fixed price CPIFixed price CPI minus food and energyand CPI with fixed prices, less food, energy and housing.)

The great unknown: housing costs

At least for the moment, the debate over food and energy inflation has been decided in favor of economists. The average price of food has increased by modest 2.6% during the past year (in this case, it is appropriate to use the 12-month figure). Meanwhile, oil prices are 20% below their February 2022 levels. However, economists also said that house prices would fall, albeit with a delay, because rents only adjust periodically. Unfortunately, that has not been the case.

Oh! It's one thing to ignore housing inflation in the summer of 2022, when the Federal Reserve had only just begun raising interest rates. It's quite another to do it 18 months later, when rental prices that were supposed to go down don't appear in the data. At 0.6%, the increase in CPI housing costs in January 2024 equaled those in May, June and July 2022. Where is the progress?

Thus, the most relevant data in current CPI reports is neither the annual calculation, nor the monthly variation, nor even underlying inflation and its variants. It's about whether housing costs have really peaked. If so, then the Fed can actually begin lowering interest rates with some confidence that its previous actions had full effect. However, if not, US interest rates will remain in limbo.

Nowadays, it's all about shelter.