In one sentence: Phil argues that the way CPI is now calculated (substitution, owners' guesses of rent) understates inflation, points to ShadowStats for the older method, and says that matters because the Fed may under-react and people may stop trusting the number.
Key ideas
- Where the series is going. Phil opens by saying the real aim is to learn what Buffett did in the 1970s to turn $60 million into $600 million, but they first finish inflation. [00:00–01:00]
- The old method. Price a fixed basket of goods and services (chicken, ribeye, rent, gasoline, a computer) in two periods and compare. [01:00–02:30]
- Substitution. In the newer method, if beef rises sharply and people can move to chicken, the measure counts less inflation. Phil's analogy: if a heavy truck rises $10,000 but you can buy a smaller one, "no inflation". It is a simplified account of the Bureau of Labor Statistics' approach. [02:30–05:00]
- Why he thinks officials like a low number. Social Security and federal pensions are indexed to CPI, as are TIPS. Phil calls this a motive; it is his opinion, not proof. [04:00–09:00]
- Many versions of "the" rate. Danielle notes FRED lists about ten inflation series; the headline most people use is CPI-U, which includes food and energy. Knowing which one a speaker means matters. [05:00–06:30]
- ShadowStats. The site recomputes inflation with older methods. Phil says it implies about 15%, up from a steady 10% for twenty years, and that academics find alternatives are double or more the official number. Caveat: ShadowStats is contested, and many economists dispute its method; treat it as one alternative view, not a measured fact. [06:00–08:30; 12:00–13:00]
- Rent. Phil says the Bureau of Labor Statistics asks about 50,000 homeowners what their home would rent for, while large landlords' data showed about 17% rent growth against the 3.2% in CPI; he says that gap alone moves headline inflation from 7% to 10%. Danielle suggests institutional landlords may own homes in hot areas, and Phil replies they bought whole neighborhoods in places like Las Vegas after 2009. These figures are his and were not checked here. [09:00–12:30]
- Why it matters. If official numbers are too low, the Fed may under-react and people lose trust and behave as if inflation is high (a self-fulfilling pattern). Phil defers consequences to the next episode. [08:00–09:30; 12:30–13:00]
How it maps to RuleOne
- Nothing on the site uses inflation directly; the connection is to the discount rate and growth assumptions you use when valuing a company (valuation pages).
- When a company's stated "real" growth is small, check whether it beats the inflation you personally experience.
Buffett, Munger and Graham links
- Buffett's 1981 Berkshire letter (and his 1977 Fortune piece) treat inflation as a tax on capital and argue that businesses needing little new capital to grow are the best protected.
- Graham, The Intelligent Investor, ch. 2, discusses inflation and why stocks are not an automatic hedge.
Words to know
- CPI-U: the consumer price index for all urban consumers, the headline US inflation measure.
- Substitution: adjusting the basket for consumers switching to cheaper goods.
- TIPS: Treasury Inflation-Protected Securities, whose payouts are indexed to CPI.
Try this
On FRED, compare three inflation series (for example CPI-U, CPI excluding food and energy, and the shelter component) over five years. Note which differ most. Then see whether a company you follow on /stocks/ raised prices by more or less than each one.
Check yourself
- What is "substitution" in the CPI debate?
Answer
Counting less inflation when consumers switch to cheaper alternatives as prices rise. - Why does Phil say officials may prefer a lower number?
Answer
Social Security, government pensions and TIPS are indexed to CPI, so a lower number costs the government less. - How should you treat the ShadowStats figure?
Answer
As one contested alternative estimate to compare with official data, not as settled fact.
Short quotes
"Knowing exactly which one they're talking about makes a really big difference." (Danielle, ~05:45, auto-transcribed)