US CPI: 2023 inflation forecast update
Major disinflation likely lies ahead in May and June, which is likely to ramp up pressure on the Fed to cut interest rates during 2H23.
In terms of the inflation debate, the next two months are likely to be of critical importance, with MAJOR disinflation likely on the way — I currently forecast a decline in the CPI and CPI ex-shelter, to 3.1% and 0.9% respectively, in June.
In-turn, this is likely to have major medium-term ramifications for the Fed’s policy outlook, with increasing political and broader public pressure likely to be applied on the Fed, to loosen its aggressive tightening.
For these reasons, and with several months having now past since publishing my initial 2023 inflation forecasts (which you can find here), I have decided that now is an important time to provide an update to my medium-term inflation forecasts.
Before delving into my medium-term forecasts, remember that inflation is highly unpredictable, incredibly difficult to forecast, and is subject to a wide array of changing variables. While this forecast is applicable at the time of publishing, my US CPI forecasts are therefore subject to ongoing revisions. Furthermore, this economics research report is not investment advice. It represents my own analysis, opinions and views, is general in nature, and does not constitute personal advice to any person — please see the end of this report for the full disclaimer.
Taking stock of underlying price foundations
There’s something else that we need to do before delving into my medium-term inflation forecast, which is to take stock of two key factors: 1) the M2 money supply; and 2) the price cycle. This is vital, as these factors will help dictate the manner in which longer-term price forecasts will be made.
For while shorter-term forecasts (i.e. 1-3 months), will be dictated largely by recent price trends, over a longer time frame (i.e. 4-12 months), one will need to incorporate a bias that reflects what the M2 money supply and price cycle are signalling.
For instance, rising M2 growth indicates growing medium-term inflation pressures, while moderating/falling M2 growth indicates moderating inflationary/deflationary pressures. If the price cycle indicates that disinflation has already occurred to a significant extent, any future disinflation is likely to be more moderate. The price cycle can also show what the general price movement bias is for a certain key category (i.e. whether prices are broadly inflating or deflating for durables, nondurables or services).
In terms of inflation, the M2 money supply is generally the one key underlying driver that matters far more than anything else.
The medium-term pressure on prices that the current change in M2 is exerting, is clear: downward pressure. This comes as M2 has fallen 4.0% over the past year.
Meanwhile, the current annual average change is -2.0%, which represents the largest decline since the Great Depression.
Looking at how prices have evolved since COVID, one can clearly see the leading nature of the money supply, which was the first to rise and fall.
Next came durables prices, followed by nondurables prices. Services prices, which were particularly impacted by COVID restrictions, and were thus the last area to see demand increase from the previous money printing, were the last to see their prices rise, and are now the last to see price growth moderate.
Having taken stock of the current rate of change in the M2 money supply, and where the US currently is in its disinflation cycle, we can now begin to unpack my medium-term inflation forecast.
Durables: another month of stronger growth expected in May, with MoM growth expected to largely normalise thereafter
Since COVID, used car prices have proven to be one of the most volatile components of the CPI, seeing huge cumulative growth across ~2H20-1H22, only to then see prices plunge from August 2022-February 2023.
More recently, wholesale prices, as measured by the Manheim Used Vehicle Value Index (Manheim Index), saw another surge higher in Q1 2023.
Historically, wholesale prices lead the retail prices that are measured by the CPI’s used cars & trucks index by ~2 months, which suggests another month of strong used car & truck price growth in May’s CPI report.
Though with wholesale price growth falling away in April, CPI used car & truck price growth would be expected to moderate significantly in June. Furthermore, wholesale prices have recorded a decline of 0.9% for the first half of May, providing a further indication that the Q1 price spike won’t be enduring. This also points to a decline in CPI used car & truck prices in July.
While new car prices have shown some above average MoM strength in March and April, just as Q1’s strong wholesale used car price growth appears to have stopped, relative strength in new car price growth would also be expected to fade away over the course of 2023, for the same reasons that strong used car price growth wouldn’t be expected to be enduring.
What are those reasons? Remembering the price cycle, durables prices have already broadly disinflated, meaning that they have already seen their prices adjust to incorporate the prior surge in the M2 money supply. With no subsequent surge in M2, it’s therefore difficult to envisage any major and sustained resurgence in durables demand, and therefore prices, over the remainder of 2023.
Instead, I expect prices to end the year broadly flat on an annual basis.
Nondurables: energy deflation expected to peak in June, food disinflation expected to continue throughout 2023
Turning to the second key phase of the disinflation cycle, being nondurables prices, the best approach is to break it down into its two key components: energy commodities and food prices.
Energy commodities
Energy commodities have been a key driver of the disinflation that has been seen in nondurables prices, and the CPI as a whole, with YoY growth falling from an enormous 60.6% in June 2022, to -12.6% in April 2023.
With very high MoM growth in May and June 2022 set to fall out of the YoY equation over the next two months, the annual rate of decline in the CPI energy commodities index is likely to both accelerate and peak over the next two months.
During the first two weeks of May, regular all formulations gasoline prices, as measured by the U.S. EIA, are down 1.9% versus April’s average. Assuming that this holds for the rest of May, and given the strong historical correlation between gasoline prices and the CPI’s energy commodities index, that it flows through in a proportionate manner, the YoY decline would be expected to increase beyond 20.0% in May.
Given the negative impact that a falling US M2 money supply has on nominal economic growth, I believe that the most likely medium-term outlook for oil prices continues to be one of moderating prices. As this would place downward pressure on gasoline prices, I am thus choosing to currently incorporate modest MoM declines in the CPI’s energy commodities index across the rest of 2023.
Doing so, shows that the YoY decline will peak in June at greater than -25%, before moderating across 2H23 as the relatively larger price declines from 2H22 fall out of the YoY calculation. By December, this results in a 5.8% YoY increase in the CPI energy commodities index, as the large 12.4% monthly decline in December 2022 rolls out of the YoY calculation.
Food
Moving now to food prices, it’s important to distinguish between CPI food at home prices, and CPI food away from home prices.
The former, which is more inherently linked to underlying food commodity prices, has begun to materially disinflate. The latter, which includes items such as restaurant dining, and is thus more exposed to lagging services prices, has continued to see price growth that is well above its historical average.
Food at home
First, let’s break down the good news, being the disinflation in monthly food at home price growth.
After seeing an astounding 25 consecutive months of monthly price growth that was above its historical corresponding monthly average (across 2010-19), CPI food at home prices have now recorded two consecutive months of price growth that has been below its historical average. This comes after food at home price growth first began disinflating in August 2022.
This disinflation follows major price declines that have been seen in underlying food commodity prices. One such measure of underlying food commodity prices, is the UN FAO Food Price Index. After seeing YoY growth most recently peak at an enormous 34.0% in March 2022, annual growth has since fallen to -19.7% in April. This huge shift comes as the UN FAO Food Price Index recorded 12 consecutive MoM declines from April 2022 to March 2023.
Historically, CPI food at home prices have been highly correlated to the UN FAO Food Price Index on a 6-month lagged basis. This time round was no different, with annual CPI food at home price growth beginning to moderate 6 months after the YoY peak in the UN FAO Food Price Index.
The historical correlation suggests significant further annual price disinflation lies in wait. I currently forecast CPI food at home price growth to fall below 2% in October 2023.
Food away from home
As touched on earlier, the picture for CPI food away from homes prices is markedly different.
As opposed to seeing monthly price growth disinflate to historical levels, it has continued to generally remain well above corresponding historical average monthly growth rates. As was seen in CPI food at home prices, CPI food away from home prices have now also recorded an astounding 25 consecutive months of monthly price growth that has been above its historical average.
While data for November and December 2022 suggested a moderation in price pressures, data for Q1 2023 showed a reacceleration of monthly price growth.
Most recently, April’s relatively more moderate price increase is again providing a signal that declines in underlying food commodity prices could be beginning to translate to lower food away from home price growth.
Given declining M2 and the move lower in underlying food commodity prices, I do anticipate monthly food away from home price growth to moderate over the course of 2023, but in the absence of stronger evidence pointing to a more material downshift (which we may receive over the next couple of months if April’s moderation continues), I currently forecast that monthly price growth will remain above its historical average across the remainder of 2023.
While such a forecast playing out would mean that March represented the peak in YoY growth, it would continue to be elevated by year’s end, at 5.2%.
Services: CPI data currently points to ongoing broadly elevated price growth throughout 2023
When analysing services prices, a distinction should be made between shelter and other services prices.
Shelter
The reason for this, is that the largest component of the shelter equation, being the rent based components of owners’ equivalent rent (OER) and rent of primary residence (RPR), are mechanically lagging on account of the manner in which rental prices are calculated in the CPI.
Instead of calculating the change in rents based upon newly signed leases, the CPI’s housing sample is instead predominately composed of continuing rents — i.e. rents that were previously signed under a fixed-term contract, of which the most common term is 1-year.
While the BLS employs such a method on account of this reflecting the manner in which lease agreements are entered into by the public, it can send misleading signals when attempting to gain a picture of current price changes.
For when trying to understand and set policy based upon inflation, what matters is not what prices rents were set at 12 months ago (or in some cases, even longer), but what is happening to rents today.
In terms of what is happening to rents today, we can look at indicators such as the Apartment List Rental Index, which shows that annual rental price growth has fallen to just 1.7% in April, down from a peak of 18.3% in November 2021, and 16.5% in April 2022.
Given the lagging nature of rental measurement in the CPI, we can see that the movement in spot market rents, as measured by Apartment List, was not similarly reflected in the CPI, whose rental indexes rose much more modestly on account of the significant amount of time it takes for the BLS’ continuing lease sample to have their leases expire and reset to current market rates.
Not only does this mean that annual price growth was much lower than the peaks recorded by spot market indexes like Apartment List’s, but that it works in reverse on the way down, with the BLS’ rent indexes remaining comparatively more elevated as they continue to catch up to the previous increase in underlying spot market rents.
Given the deceleration/outright declines in MoM growth rates that have been seen in spot market rents since ~September 2022, whilst the BLS’ rent indexes have been recording relatively high MoM growth, the gap between spot market and the BLS’ rental indexes, has now largely converged.
Given that the BLS’ rent indexes have significantly bridged the gap to spot market rents, they have also begun to see their monthly price growth decelerate. A downtrend now clearly appears to be in place for both OER and RPR, which can be seen in the chart below.
Given the moderation in underlying spot market rents, I continue to expect an ongoing deceleration in the monthly growth rate of the BLS’ rental indexes. Though note that for the same reason that the BLS’ rental indexes are lagging, they are also smoothed.
This can be seen in the gradual MoM acceleration over 2021, and more recently, in the gradual deceleration in MoM growth. The abrupt downshift in March is a clear outlier, which as articulated in my April CPI preview, led me to conclude that a bounce back should be anticipated in April:
Following the large moderation in OER and RPR in March, I am conservatively anticipating some bounce back in April.
The reason for this, is that in addition to being lagging, OER and RPR are smoothed measures. This stems from only a small portion of the rental sample consisting of new leases in any given month, with the bulk of the sample instead consisting of continuing lease agreements.
As opposed to an abrupt shift, I therefore forecast a gradual moderation in the MoM growth rate of OER and RPR over the remainder of 2023.
Though given that the monthly deceleration is expected to only be gradual, YoY growth is likely to remain elevated by year’s end, with further time needed for previously high MoM growth to be fully cycled out of the YoY calculation.
For this reason, it is vital that one also looks at the inflation picture on an ex-shelter basis. I will expand upon this point further, later on in this research piece.
Other services
Turning our attention to the rest of the core services index, we can see in the chart below, that prices have been continuing to grow at a pace that is materially above their historical average.
Remembering the price cycle that I outlined at the beginning of this report, core services prices were the most lagging component of the equation, being the last to see prices peak and fall.
Note that in addition to excluding lagging shelter costs, the below chart also excludes health insurance, which is indirectly calculated in the CPI, and has been subject to particularly erratic annual price adjustments over the past few years. Household operations, which have not been consistently measured on a monthly basis since 2021, are also excluded.
Breaking down some of the core services components, a major driver of higher services price inflation has been transportation services. Though recently, a moderation in motor vehicle maintenance price growth has been providing some positive signs on the transportation services price front. I currently expect a continued gradual moderation in monthly motor vehicle maintenance price growth over the course of 2023.
Over time, this would be expected to help feed into lower vehicle insurance price growth, which has been running at very high levels. Though given the very high levels that motor vehicle insurance prices have been increasing at, while some moderation is anticipated, materially higher than average growth is currently assumed for the rest of 2023.
Internet services, an item that has generally seen relatively little inflation, and indeed, recorded negative annual price growth in several years across 2010-19, saw price growth accelerate in Q4 2022. This continued into Q1 2023, but most recently, in April, prices saw a significant monthly decline.
While one month doesn’t make a trend, and I thus expect some bounce back in May, given the downward price pressure from falling M2, and the signs of the wider disinflation process continuing to filter through the economy, I expect internet services price growth to moderate and largely return to historical monthly norms by the end of the year.
One area that didn’t show any moderation in April, and instead showed a major reacceleration, was the CPI other personal services index, which recorded its highest ever MoM growth, in data dating back to 2010.
Given recent trends, I continue to expect that other personal services MoM growth will remain materially elevated versus historical norms across 2023.
The story is broadly similar for the recreation services category, where prices once again returned to growth that was well above their historical monthly average in April, after seeing weaker than average growth in March.
More broadly, the trend in recreation services prices continues to be one of elevated price growth. I thus continue to expect prices to grow at a rate that is significantly above their historical average, but for the extent of this faster growth to moderate over the course of 2023.
The theme here should be clear, being that while many services continue to record monthly price growth that is well above their historical average, declines in the M2 money supply would be expected to result in MoM price moderation over time, particularly given that a holistic view of the price cycle shows that disinflation is spreading.
Though in the absence of more material evidence of a deceleration in monthly price growth for many core services, I view it as prudent to continue forecasting MoM price growth that is often above the applicable historical average, over the course of 2023.
Should clear evidence emerge of a faster moderation in many core services prices over the months ahead, the forecast can then be revised accordingly.
One factor that would drive not only services prices, but prices lower more broadly, is a recession, which historically, often leads to a major reduction in inflation.
Putting it all together — big disinflation likely lies ahead in May and June
Once everything is put together, we can see why the next two months are likely to prove critically important for the overall inflation and Fed policy debate — I currently forecast the headline CPI to see a significant moderation over the next two months, falling to 3.1% in June.
I currently expect annual CPI growth to remain between 3.1%-3.5% from June-December 2023.
This forecast represents a slight increase versus my forecast in January, as energy commodities, other personal services, and recreation services price forecasts are revised upward, whilst food at home price forecasts are revised downward.
Though this only tells part of the broader disinflation story. Remember, shelter is a huge component of the overall CPI (~34.6% of the CPI), and is heavily lagging.
Given that we know that underlying rents have significantly disinflated, in order to gain a better understanding of the current price pressures that are impacting the economy, it’s vitally important to also look at the CPI on an ex-shelter basis.
This reveals a far more significant disinflation picture, with CPI ex-shelter currently expected to plunge over the next two months, from 3.4% in April, to just 0.9% in June. Annual CPI ex-shelter price growth is expected to then remain between 0.9%-2.7% over the remainder of 2023.
Again, this represents an increase versus my January forecast.
The core CPI, which I believe to be an inferior measure of annual inflation, but which the Fed pays significant attention to, is expected to continue to moderate over the course of 2023, but in a much more modest manner, falling below 4% by the end of the year.
Though again, we need to look at the change on an ex-shelter basis, with this lagging component having an even greater weight in the core CPI.
I currently expect core CPI ex-shelter to fall to 3% in June, and remain between 2.5%-3.1% thereafter.
Significant disinflation makes it harder for the Fed to justify not loosening its aggressive tightening
Should inflation play out roughly as I currently forecast, from June onwards, it is going to become significantly more difficult for the Fed to justify maintaining the federal funds rate at 500-525bps.
Doing so would result in a large spread between the CPI, and a very large spread between the currently more appropriate measure of inflation, being the CPI ex-shelter index.
By June, the federal funds rate would also exceed the core CPI, and further significantly exceed the core CPI on an ex-shelter basis.
With many market participants already somewhat aware of the major disinflation that lies ahead, this likely underpins current market expectations for rate cuts in November and December of this year.
An expectation for significant disinflation certainly played a big part in my stated expectation for rates to be cut in 2023, which I outlined in January as part of my 2023 outlook. Given my continued expectation for additional major disinflation, I continue to anticipate that rates will be cut at some point during 2023.
Recent US bank failures, which are a symptom of the Fed’s aggressive tightening, are likely to increase pressure on the Fed to loosen its aggressive tightening, for as long as its policy stance encourages a decline in bank deposits and the M2 money supply, bank stress will grow.
In summary
With major disinflation likely to be seen over the next two months, it’s critical to have a thorough understanding of the medium-term inflation picture.
While subject to ongoing revisions, I currently expect the CPI to fall to 3.1% in June, and the CPI ex-shelter to fall to 0.9% in the same month.
With such forecasts pointing to a federal funds rate that will soon begin to significantly exceed the annual rate of CPI inflation, market, political and broader community commentary, is likely to increasingly apply more pressure on the Fed to loosen its aggressive interest rate tightening.
Should additional bank collapses occur, and the unemployment rate begin to rise, that pressure is likely to very quickly reach fever pitch.
I continue to expect that interest rates will be cut in 2023.
Thank you for reading! I hope you enjoyed my latest research piece.
In order to help support my independent economics research, please consider liking and sharing this post. Given that Economics Uncovered is very much an independent endeavour, your support in sharing the word provides a great help in allowing me to continue to provide you with independent, institutional grade research.
Please feel free to leave any comments below, to which I will endeavour to respond.
If you haven’t already subscribed to Economics Uncovered, subscribe below so that you don’t miss an update.
amazing steven--kudos thx and keep up the great work