Not long ago in a locale near me, I paid $12 for a hot dog at Billy’s Gourmet Hot Dogs. I did not believe this $12 dog would be as fantastic as Billy believed it was. It was my first time at Billy’s and I had no reference point of what he believed his hot dog was worth historically. However, in the following days, I noticed that McDonalds believed their burgers were worth more than they once were – charging $9.18 instead of ~$7 for a Big Mac meal. You can see where this is going.

If you’re more worried about my eating habits than you are the general rise in prices I’ll take comfort in that, but you shouldn’t be. The price rises we’re witnessing aren’t just rapid – they are historic.

The US CPI reading this week headlined:

“The index for all items less food and energy rose 0.9 percent in April, its largest monthly increase since April 1982.”

Other indices showed similar increases and it is no question that anecdotal and statistical evidence is all around us. Inflation is here.

Yet, there is much debate. Where is it coming from? Where is it going? How long will it last?

Indeed, inflation is a measurement, but prices are not. Prices are information to be observed and the observers respond with changes in their behavior. One price means different things to different people – an increase tells creators to make more while consumers to consume less. Further, prices are connected. An increase in one price precedes increases in others. A material change in a price ripples across everything it is nearest, like a stone skipping across a lake. Once the first skip occurs you wonder how far until the next skip, how many more skips will come after, and how far those ripples will spread.

In summary, prices are:

• Relative – they mean different things to different people.
• Connected – a change in one affect change in others
• Temporal – prices exist across time. Based on present circumstances, people infer how they will change in the future and just how far into the future that change will occur.

You see, prices are the core of economic activity. They are the information used by economic actors to make decisions. How we measure them is a separate question. Fundamentally, we use money to represent the value of something and the change in that value shows us a change in it’s price. This relationship, of course, is subject to supply and demand. It is also subject to the measuring stick – money. When our money changes, our prices change, irrespective of changes in supply or demand. If a price is an indication of supply and demand, then changes to its representation (money) distort that price.

The salient question: What proportion of these historic price rises are a consequence of changes in supply and demand vs. the changes in the supply of money that represents our prices?

Sadly, this essay will likely provide more questions than answers. Our inflation metrics are poor and manipulated. Further, inflation means something different to many different people. The task of creating a broad measure that applies to the majority of people is doomed in many ways. However, we can try our best to understand:

• What we want – if we could perfectly measure inflation what would that look like?
• What we choose – what have the high priests of finance & economics chosen as measurement?
• What is optimal – what alternatives could the masses turn to?

Let’s begin.

What we want

Inflation is a general rise in prices. This definition can be stratified in a variety of different ways depending on what grouping of prices we believe are most relevant towards a particular group of people. Most often, we’re referring to general group of the population that most often consumes a standard set of goods and services. We call this the Consumer Price Index (CPI).

More specifically, there is a distinction between price inflation and monetary inflation. Price inflation measures observed changes in prices. Monetary inflation measures changes in the money supply. What’s important is that price inflation is affected by monetary inflation but not vice versa. Price inflation incorporates changes in supply and demand as well as changes in the measuring stick (money). Monetary inflation only considers changes in the measuring stick.

If the goal is to determine a general rise in prices for a majority of people – monetary inflation could be our upper bound on that measurement while all the different types of price inflation measurement fall somewhere below. When the money supply is increased, prices inevitably rise. The question is: where do they rise and does this matter for the majority of people?

I would argue, without complete certainty, that monetary inflation appears most immediately in the prices of financial assets. This is a product of the nature of our banking system. When money is created it is:

1. Injected into our banking system.
2. Banks use it as collateral to make loans to companies.
3. Companies invest in capital and financial assets.
4. Assets produce revenue.
5. Revenue pays expenses to people and results in profits.
6. Profits go to people and back into businesses.

Much of the new money gets stuck at step 1: in assets. Asset values inflated massively and immediately at the beginning of the 2020 pandemic stimulus:

It took 9 – 12 months to start seeing material change in consumer and producer prices:

Just taking this simple assessment shows how complex this all becomes: prices rise in different ways for different reasons throughout time. We look at this through a variety of lenses. When the money supply increases it hits somewhere, just not the CPI perhaps.

What we Choose

Broadly, the CPI is our index of choice used to commonly measure inflation. If there is one goal I have from this writing it is convince you, in detail, that the CPI does not measure what it is portrayed to measure.

Here we go.

The money supply has increased ~27% since the beginning of 2020 and the CPI has increased ~3%. The “true” rate of inflation is likely somewhere in between these two points.

The CPI measurement of inflation appears to be strikingly low from my, and others, experiences.

Why is this?

Below is a table I’ve compiled from a variety of sources highlighting the difference between observable price data and what the CPI says the price data was from the period of 1971 – 2019:

(If you’re curious about some of the changes that occurred starting in 1971 check out wtfhappenedin1971.com)

Let’s walk through this table:

1. CPI increased 6.3x since 1971 – this is our baseline for measurement.
2. Income increased 7.1x – this tells me people are gaining income above the rate of inflation so things should be getting more affordable.
3. All in, a group of the most major expenses in peoples lives have gone from 34% to 93% of income
4. And this group of expenses increased 19x the amount in 1971 while the CPI increased only 6x
   1. New home 10x: CPI 7.7x
   1. New Car 11.0x: CPI 2.7x
   1. Apartment Rent 13.6x: CPI 6.8x
   1. Healthcare 29.4x: CPI 15.5x
5. However, M2 money supply increased 22x, a number much closer to the increase in expenses

The CPI materially understated some of the most significant expenses to the average American. College tuition costs appear to be consistent with primary data sources. However, college tuition costs increased from 16% of median household income to 47% as of 2019. The CPI weights this cost in the index as 2.9%. Meaning the BLS believes the average consumer spends 2.9% of their income on college expenses.

There is a striking disconnect between the reality of consumer spending and the CPI measurement of it.

Of course, this isn’t a perfect comparison, but it shows a strong disconnect between the CPI and the reality that many are experiencing today.

Why is the CPI so far off?

The CPI is published by the Bureau of Labor statistics (BLS) with the stated objective to be a price index of market goods compiled to express the costs paid by the average consumer. This basket includes a wide range of goods and services deemed common to the average consumer which is updated over time. It excludes savings and investment items such as stocks bonds, real estate, and life insurance. The original methodology was to observe the prices of a basket of goods and multiply them by some weighting deemed appropriate (price*%weight).

This index began as a product of WWI price measurement. Curiously, the index underwent a variety of changes starting in the 1980s, soon after inflation peaked at 15%. The following changes occurred:

• Composition: The general theory at inception of this index was to observe change in prices from constant-standard-of-living. Meaning that the basket of goods was fixed, and this particular standard of living should reflect changes in that cost by remaining static. So, the components are the same over time and the change in price that occurs is how much incomes must also rise to maintain that standard of living. In fact, the CPI used to be known as the Cost-of-Living Index. This changed in 1990s when the government argued that the CPI was overstated because it didn’t account for consumer substitution. Meaning, if a good (like steaks) gets too expensive consumers will substitute their consumption with something less expensive (like burgers). Thus, the composition of the index should be changed to reflect the actual consumption habits of consumers.
  • Criticism: In effect, the ruler of measurement changed significantly from this. If you are going to measure period to period changes in something, then that something needs to be same in each period. By changing the composition of measurement, the length of the ruler was now subject to government discretion. Of course, the basket changes over time as consumer spending habits change but this is different. This is saying that specifically goods in the basket whose prices increase will be removed. This methodology change puts downward bias on prices as goods that rise in price are systematically substituted with cheaper alternatives. Do consumers actually do this? Sometimes yes, sometimes no. To say it should be methodology implies certainty that this behavior is broad and persistent.
• Weightings: under a constant standard of living the weight applied to each good was chosen to reflect this standard. However, the BLS implemented geometric weighting – a mathematical calculation that increases weights of goods falling in price and decreases weights of goods rising in price.
  • Criticism: In effect, this changed the internal composition of the basket to apply a downward bias on price measurement. The BLS also changed re-weightings from every 10 years to every 2 years making the period-to-period composition even less consistent.
• Quality Adjustments: the quality of a good is reflected in its price. There is an argument that the quality of a good can fundamentally change over time and thus be a functionally different good than it was before. For example, as iPhones have risen significantly in price the technology has also granted much greater utility to the consumer. Thus, adjusting the price downward to reflect this fundamental change in the product is appropriate to keep the unit of measurement consistent. This was what the BLS did when implementing “hedonic quality adjustments” into the CPI.
  • Criticism: this logic is contradictory to the rational for changing composition and weightings. Further, I would consider this a natural “change” in the consumer bask. Thus eliminating the need to change the composition arbitrarily. What matters, no matter how much iPhones change, is that more and more people own them. If their quality changes, then the consumer basket is changing – the primary purpose of the assumed changes in composition.

The goal of the composition and weighting changes was to reflect how consumers are spending in the current period while the goal of quality adjustments was to reflect how they used to spend. These goals are in direct contradiction.

What is consistent between all of the changes is that they were ultimately used to adjust prices downward. The estimates used to adjust quality are based on computer statistical modeling which are far removed from the actual expense’s consumers were paying. Maybe the new iPhone has a better camera with features a consumer will never use but none the less still has to pay for to get the new iPhone.

The point is that sifting through theoretical and statistical minutia is not real, while the actual prices being paid by consumers is real. There is a natural change of consumption being reflected in prices through quality improvements. None the less, this was another layer of assumptions that the BLS could implement and interestingly they adjusted prices downward with them. Technology costs were adjusted lower while the decreasing quality and higher prices (upward adjustment) of air travel was ignored. All the while, consumers do not care and pay what they must.

In summary, you have a list of goods in the index, that are applied specific weights, and have observed prices in the market. Throughout the 80s and 90s the BLS changed all three of these variables and their calculation methodology. They referred to them as “improvements” but really just granted them more influence over the calculation to be what they wanted it to be.

The aggregate impact of these changes has been to reduce the reported level of inflation in the CPI as admitted in the 1999 paper CPI research series using current methods, 1978-98 by Stewart and Reed of the BLS. In this paper they apply a retroactive application of the methodological changes to the CPI-U (CPI-U is the primary version of the CPI used for policy) and call it the CPI-U-RS. So not only were changes being made over time to the CPI, but they are not applied retroactively to the numbers prior to implementation. The changes occurred from the period of 1978 to 1998 and the below chart shows the cumulative effect that these changes had.

Over the 20 year period, due to changes implemented by the BLS, the CPI was stated by 12% which is roughly 0.6% per year. Simply, had these changes been implemented retroactively inflation would have been lower. This impact, while material, wasn’t devastating.

What gives us a better view is looking forward. Had the original CPI-U methodology been used without these changes, what would inflation be today? Admittedly this exercise is challenging and the BLS does not report statistics for this comparison. The process requires reverse-engineering the CPI-U-RS and applying it to future periods. Meaning, estimating the amount of inflation lost from the BLS methodological changes and adding it back into the official CPI-U inflation after 1999. Shadowstats.com has attempted this exercise and achieved the following results:

As of April 2021, official CPI inflation is 4.2% while inflation estimated without changes to methodology is ~13.0%.

This is a significant difference. While this statistic is not gospel and itself has assumptions, there is no argument that official inflation statistics are not being materially understated. This is because the BLS has changed the CPI from being a fixed point of measurement to a constantly changing group of assumptions. You cannot periodically measure a change in something if the length of your ruler is constantly changing as well.

Why would they want to understate inflation?

1. Inflation is a constraint to Federal Reserve policy: expansion of the money supply has the cost of increasing inflation. Manipulating official statistics of inflation can make their policy appear more productive than the reality.
2. Social security payments and government contracts: are tied to the CPI-U. Lower inflation statistics means they pay lower expenses. It allows members of congress to lower social security benefits without having to vote for it.

Had CPI methodological changes not been implemented, social security benefits would cost roughly twice as much as they are today. (shadowstats.com)

• Tax bracket adjustments: occur annually and are tied to inflation. Under our progressive tax system higher percentages are tied to higher tax brackets. Let’s assume over a 2-year period everybody’s income remained the same and the CPI was 5%. If the bottom tax bracket was $10,000 then it would increase to $10,500. Since incomes remained the same – more income now falls into the lower tax brackets and thus is taxed at a lower rate. So, by keeping the CPI lower the government benefits by having more income fall into higher tax brackets, earning more tax revenue.

So, you see, I do not believe the CPI is a good measure of inflation. The question is: what is?

What is optimal

Originally inflation meant an increase in the supply of money but over time has been defined as a general rise in prices. The two are linked – an increase in the money supply will eventually lead to rising prices. Milton Friedman stated in Studies in the Quantity Theory of Money:

“There is perhaps no other empirical relation in economics that has been observed to recur so uniformly under so wide a variety of circumstances as the relation between substantial changes over short periods in the stock of money and in prices; the one is invariably linked with the other and is in the same direction; this uniformity is, I suspect, of the same order as many of the uniformities that form the basis of the physical sciences.” (Friedman, Studies in the Quantity theory of money p20-21)

The fact is that measuring aggregate price rises is complex and challenging. A price of something is reflective of all knowledge that exists among market participants. Conceptually, a price is information that allows buyers and sellers of a good to adjust their behavior for changes in supply and demand. For any given good trading in a market there are thousands or millions or billions of buyers and sellers who all have different motivations and circumstances and resources. The market price of a good encapsulates all those things at a single point in time through a process of complexity beyond our comprehension. A price is information for people to act upon, not something to be controlled. Trying to replicate such a process with artificial prices, weights, and methodologies is laughable. What can be measured far more accurately is the supply of money.

Here are the current options:

Monetary Inflation:

• M1 – includes currency held by the public and checkable deposits. Intended to define the amount of money people hold for the purpose of making transactions
• M2 – M1 plus non-transaction accounts (e.g., money market accounts and certificates of deposit). This is the preferred measure by policymakers because it has the most stable relationship with interest rates.
• MZM (money zero maturity) – basically is M2 less small-time deposits plus wholesale money market mutual funds available only to large institutions. The purpose of this metric is to represent instantly accessible forms of money – a concept defined by Murray Rothbard 30 years prior.
• TMS – All components are money of zero maturity that are instantly accessible by their owners

MZM was discontinued by the Fed and TMS I haven’t found updated in a while. I think the true consumer inflation rate is less than the money supply increases. If only there was a government agency that collected data on a typical basket of goods and did not manipulate it. We’ve all seen what has happened with public company adjusted EBITDA numbers – they run persistently high for “one-time” changes. However, I think following the shawdowstats.com readjusted indicator is reasonable measure to use and to be considered with money supply increases.

Sorry, I don’t have a clear answer for this.

Will inflation be transitory?

Inflation, whatever it is, has consequences. The Fed believes our current spike is the result of an imbalance between supply and demand. The idea is that we are witnessing supply shortages as prices have not caught up yet with demand. Once they do, supply will increase, and the market will find its new equilibrium.

The Fed’s rational and my response:

• Inflation expectations: surveys show consumers expect inflation to 3.4% over the next year and then drop.
  • Response: this is supporting evidence, but we all know how accurate surveys are. Anecdotal evidence from my life says otherwise.
• Labor slack: without more people working and spending like they did pre-pandemic it will be hard to increase prices.
  • Response: they are still getting money from the government that they are still spending.
• Sticky Prices: rising commodity prices are the result of supply and demand imbalances, not persistent inflation. They support this point with this graph that strips out any outliers. I think that it is precisely the outliers that matter in this environment…

• • Response: their index is stupid, but I do believe some of the price rises can be attributed to supply and demand while others to monetary inflation.
• Disruptive Technologies: technological advancements leads to disinflation and this will continue.
  • Response: this has remained static and overall impact only matters if you’re downward adjusting the CPI for it.
•  Base Effects: Twelve-month inflation readings have big swings when there are large changes in year-earlier months, creating “base effects” which will be particularly pronounced because prices fell so steeply this time last year.
  • Response: I agree but I’m not really looking at the CPI anyways.

I don’t think inflation is transitory. I think we’ve massively expanded the money supply and the rippling effects are still spreading. I think that the world is waiting for the rock to land from it’s next skip and the expectations are historic – just like everything else over the past year. LOL.

Buy Bitcoin.