CPI is undercooking inflation – Interview w/ Truflation

There is arguably no bigger issue facing today’s economy than the rampant inflation we are seeing across the globe.

US inflation hit a 40-year high of 9.1% according to the most recent CPI reading. Across the Atlantic in the UK, the situation is similar with 9.4% the reading for June – indeed, all across Europe we are seeing similar figures.


Are you looking for fast-news, hot-tips and market analysis?

Sign-up for the Invezz newsletter, today.

But with the US number often making the headlines, is the CPI really a reliable metric to go off? If you ask the average citizen on the street what percentage inflation they think they are currently experiencing, it’s likely their answer will be well north of the “official” CPI rate of 9.1%.

This undercooking is, many argue, the by-product of the fact that the government reporting on inflation is the very definition of a conflict of interest. Politicians’ approval ratings depend on keeping the public content, and inflation is universally despised. So, does the administration not have every motive to keep that number low?


This takes me to Truflation, a company which aims to provide independent, censorship-resistant financial and economic data on-chain. That’s a very “2022” description for a company, but essentially it serves to leverage the power of the blockchain to provide absolutely transparent – and hopefully more accurate and independent – readings on a variety of metrics. One of those metrics, and the one we are focusing on here, is the inflation rate.

I definitely believe there is scope for innovation here, and it does frustrate me that so many don’t understand how the CPI is measured. I was therefore quite enthused to interview Stefan Rust, CEO of Truflation, on the shortcomings of the CPI and how Truflation’s approach may help.  Take a read of the full interview below and see what you think about which metric is superior. Personally, I don’t believe there is any debate here…

Invezz (IZ): Could you sum up what you believe are the problems with the CPI metric?

  • Methodology: The statistical methods used to calculate inflation are outdated, based on often biased surveys, using statistical smoothing techniques and other adjustments.
  • Frequency: reports on CPI/CPIH are released monthly with no clear indication of when the snapshot in data was taken (many data sets suffer from delays inherent to the survey collection cycles). 
  • Transparency: It is often non-transparent how countries collect, calculate, and adjust inflation. The underlying data is not auditable or verifiable. 
  • The breadth of data: Depending on a country, CPIs use ~ 80K (US) -180K (UK) prices, which forces them to carefully choose their analyzed basket of goods and services. Our data is far more robust, UK index: +10 data sources across 8 million items, US index: +30 sources, 10 million items.

IZ: Do you believe the government intentionally understates true inflation via the CPI for political reasons, or do you think it is merely an antiquated system that they are reluctant to change because they believe it does give an accurate measurement of inflation?

The CPI evolved in a decades-long process to where it is now, a tool that informs and serves the governments and their policies. It is not designed to accurately inform the public, investors, or companies.

Governments around the globe are incentivized to keep the reported inflation rate low. For example, some of the incentives in the US include: maintaining trust in the US dollar as the world’s reserve currency, higher negotiating power with other countries, preventing stock markets volatility, lower social programs and retirement payouts, preventing panic and the inflation loop.

Truflation’s goal is to help individuals, investors, companies, and institutions make more informed decisions by having access to independent and unbiased economic information.

IZ: CPI is also often criticized due to the fact it is a cost-of-living index rather than actually measuring the difference in prices. In such a way, it may underreport true inflation if customers switch products as a result of rising prices – does Truflation take account of this shortcoming?

CPI Cost-of-Living (COLI) methodology, together with limited surveying ability, results in the need to select items they count in their basket of goods and services. It may lead to substitutions, especially when consumer trends change. 

Truflation solves it to some extent by taking the wide breadth of data (>10m items, with multiple price points per item). By taking all the available data in all categories, we do not need to pick and choose the basket. However, the consumer trends will have some effect as people move to smaller or less expensive items or companies discontinue some stock-keeping units (SKUs). 

IZ: Do you believe the CPI also struggles to adapt to changes in the quality of goods, and will Truflation take account of this?

Truflation naturally corrects for consumer trends through the sheer amount of data. That’s why we call it a developer approach, scrapping all data at the actual point of sales. Truflation does not adjust for quality, either better or worse given that this is already incorporated in our data.

IZ: At the moment each country has their own technique for measuring inflation. Truflation says its indices are easier to scale across regions allowing for improved comparability – can you please elaborate on this?

Truflation has access to global, high-quality household expenditure data. Our categories and their relative weights are based on that data and comparable across multiple countries. 

Each of our CPI indexes has the same 12-category framework and calculation methods 

The calculations we take are very simple and repeat across all main CPIs, except we do not include any adjustments or statistical smoothing. A weighted average of price indices. 

The weights are established by taking multiple data sources, including census and mini census data, BLS/ONS surveys, and multiple other data sets to create a consumer expenditure database that is then validated with 3rd party sources, e.g., food spending with retail tapes, mortgage spending data etc. 

IZ: Do you believe inflation would be better measured on a per-person basis rather than a per-dollar basis? To elaborate on what I mean, high-income individuals spend more, meaning they bear a greater weight on the index. In other words, each dollar casts a vote, instead of each person, and hence the inflation rate is often wide of what poorer (or even middle-of-road) families are seeing.

Everyone knows best what their actual inflation rate is. They experience it daily, and if they keep track of their budgets, they can calculate it. Not everyone does, of course, so we are working on custom inflation options where people can see the impact of inflation against their personal expenditures. 

As for the division per income group, many researchers pointed out that it would be interesting to analyze, and with our consumer expenditure trends, we could do that in the future. By that, I mean, we could count different weights for each income group. For example, we saw that the differences are much more correlated to income groups than geographic locations.

The weights will differ by income group but should not have a dramatic effect on the final inflation rate. Most people do need similar items to live. If the rich purchase more expensive wine, their purchases will still be averaged with the whole country and for example, bargain hunters. Additionally, the richest purchase many things not calculated in the CPI, like homes, stocks, and luxury items. 

IZ: If all this data is objective and independent, could a firm (such as yourselves) not just publish it on a monthly basis rather than putting it on the blockchain? I understand the data is unchangeable on the blockchain, but if the reading is published in a public forum, would this not have the same effect?

Our data is already being published daily and available to the public at app.truflation.com.

Our data is available on-chain through a leading decentralized oracle network, Chainlink, because it makes it easier to use by blockchain developers. Chainlink translates real-world data like Truflation’s indices to the programming languages of different blockchains so data can be used in creating novel economic products and programmable money/contracts, i.e., self-adjusting digital salaries, inflation bonds, TIPS, etc.

Additionally, based on Chainlink concepts of cryptographic truth, Truflation is building an infrastructure that would guarantee cryptographically verifiable data by sourcing information from multiple independent providers and putting them on chain. 

Where to buy right now

To invest simply and easily, users need a low-fee broker with a track record of reliability. The following brokers are highly rated, recognised worldwide, and safe to use:

  1. Etoro, trusted by over 13m users worldwide. Register here >
  2. Capital.com, simple, easy to use and regulated. Register here >

*Cryptoasset investing is unregulated in some EU countries and the UK. No consumer protection. Your capital is at risk.