A note to President Trump from Barry Habib on Inflation Calculations

Dear President Trump,

It’s always an honor to see you at Mar-a-Lago. Recently, Roger Stone suggested I share my thoughts with you on ways to make housing more affordable by using more updated and indicative inflation metrics.

I understand how important low rates are to you and our country, and I believe that the Fed can be influenced to cut rates aggressively if they understood how flawed their methodology is.

The Fed believes that inflation is too high. Their favorite measure of inflation, Core PCE (Personal Consumption Expenditures) sits at 2.8%, much higher than their 2% target. Core CPI (Consumer Price Index) is another important measure that sits at 3.25%, also much higher than their 2% target.

Their constant jawboning about needing to get inflation under control has caused concerns in the bond market, which has caused longer-term interest rates to remain stubbornly high, thus hurting home affordability.

The problem is that the Fed’s methodology is flawed and antiquated. The largest part of the Core inflation calculation is Shelter. As you know, this measures actual rents, as well as the imputed guesswork of what an owner would charge rent for their home, not home prices or mortgage payments. While this type of data gathering can be improved, the major problem is the lag in the data compared to what is happening in real-time.

Shelter is the largest component in both readings, and accounts for 18% of Core PCE and a whopping 46% of Core CPI. Therefore, it is critical that these readings are reflective of what is actually happening because they will influence both the Fed and bond market.

To give you an idea of how important and impactful this is, the Shelter components in PCE and CPI are as follows:

·       PCE Shelter: +4.8% year-over-year as of November (most recent data)

·       CPI Shelter: +4.6% year-over-year as of December

Both of these readings account for the lion’s share of the perceived inflation problem and are also in major contrast to real-time data. There are multiple readily available, reliable sources like CoreLogic, showing that rents on a blended basis between new and renewals are actually up just 1.5% year-over-year.  Other sources like Apartment List and Realtor.com, which measure new rents, can be blended with sources like Zillow, which measures renewal rents, to show an identical result of a 1.5% annual increase. This is good data that tells us that both PCE and CPI are dramatically overstating inflation.

Eventually, these numbers will catch up, but it will take the better part of a year to do so. I will offer an explanation below on the current flawed methodology, and how it has kept interest rates both higher and lower than they should’ve been, causing boom and bust cycles. But first, let’s take a look at how the Fed, through this miscalculation, was complicit in causing the inflation problem that we have experienced.

Back in 2022, anyone with a bit of common sense, or who had first-hand grocery shopping experience, would have known that inflation was rising quickly. But the Fed kept fueling the flames of inflation with QE and zero interest rates. Their rationale was due to low numbers on both Core PCE and Core CPI. However, these numbers were being held artificially low due to the enormous weighting in Shelter, which was reporting low numbers due to the aforementioned lag. Had they used a real-time look at Shelter, they would have seen that inflation was already becoming problematic and would have moved more quickly to preempt the inflation that was about to come.

Here is a look at how the BLS and BEA calculations on Shelter create a lag:

Their calculations look at the country in six segments … they call them Panels: A, B, C, D, E, and F. Each Panel takes a turn in being used for the government’s inflation calculation. Apparently, the BLS is not equipped to handle looking at all six segments each month. Years ago, this type of real-time and reliable data from sources outside of the BLS may not have been available. Today, we should consider using this data since the BLS may not have the ability to expeditiously calculate this information. Let’s look at an example of how this can generate a significant lag during periods of inflection.

In January, the government will use Panel A … only Panel A, so the other five regions are ignored. Then each following month, one of the other regions is calculated. The obvious problem is that the next time we have a look at Panel A is in July. That means that any change in rents, like the significant decreases we’ve been seeing of late, are not factored in for six months. To add insult to injury, the calculation also fails to contemplate that when Panel A is surveyed for the amount of rent being paid, it may be reflective of a lease that was signed several months earlier.

We estimate, and back-testing data supports this, that the lag is approximately one year. This also explains why the government is looking at 4.6% to 4.8% Shelter readings, when the real number is approximately 1.5%. Now let’s take a look at the impact this has …

Core PCE currently                      2.8%

Shelter component weighting   18.0%

Government data Shelter           4.8%

Actual market data Shelter         1.5%

Overstatement of Shelter           3.3%

If the actual Shelter reading of 1.5% were used, the overall Core PCE inflation rate would be 2.2%.

Using the same calculation on Core CPI, the inflation reading would be 1.8%.

If both the Fed and markets were looking at these readings, it would paint a completely different and more realistic picture on inflation. As a result, I believe the Fed would be looking to cut rates aggressively, and the bond market would also improve significantly with mortgage rates closer to 6%.

A simple change to a more real-time methodology would solve the current problem and may assist in preventing future miscalculations on Fed action.

Your loyal friend,

Barry Habib

‍