Hacker News Reader

Comment by Bostonian | Hacker News Reader

The chart is misleading because it adjusts for inflation using M3, rather than the Consumer Price Index (CPI). People assess inflation by the prices they pay, not some monetary aggregate. Using CPI, the S&P 500 has recovered to its 2000 peak. Just as important, the owner of an S&P 500 index fund would have received dividends, so her total return would have exceeded inflation.

The caption at the bottom suggests why this misleading analysis is shown. The author is pushing crypto. "Invest wisely and you can maintain or increase your standard of life. Invest poorly and the road to serfdom is real. Invest in crypto"

View on HN

Isn't CPI politicized and doesn't accurately account for real cost of living changes?

Why do you think CPI is a better measure?

> Isn't CPI politicized and doesn't accurately account for real cost of living changes?

CPI ain't perfect. Anyone can argue about the compensation of the basket and make points that it might understate some portions of household spending. And then there's confounds in the opposite direction: it doesn't take into account hedonics at all really. But it's not terrible, either: it's the best measure we have.

> Why do you think CPI is a better measure?

See sibling comments. If CPI doesn't quite accurately report purchasing power... M3 is nowhere close as a metric.

Hedonics are one of the best arguments against CPI - https://wolfstreet.com/2019/12/05/what-worries-me-about-hedo...

This M3 nonsense is ridiculous though, I agree.

Hedonics (which I had mentioned) are an inescapable confound. We can't go back and live in 1990 with 1990s goods.

The cost of an equivalent quality of life (by our best reckoning) is measured in the various CPI. But people expecting more in life (safer products, more computing, etc) isn't exactly "inflation". And if we treat increases in quality of life as an increase of cost of living, we're really measuring something else.

> Isn't CPI politicized and doesn't accurately account for real cost of living changes?

There isn't a single CPI. Headline CPI is widely reported because it's approximately right for most people.

If you're doing inflation-sensitive planning, you should be using a more fine-grained metric. Predicting your personal cost of living changes using national core CPI is like aiming a gun with a compass. (That said, most people--when discussing inflation--aren't doing anything practical with it. For conversational purposes, headline is fine.)

CPI is misleading, M3 is more accurate. CPI is heavily influenced by technology trends and offshoring of labor for widgets.

House and asset prices tell different story.

> CPI is misleading, M3 is more accurate

This assumes a constant velocity of money. That's empirically false. Money supply can remain stable while prices soar or collapse solely on the back of velocity changes.

> CPI is heavily influenced by technology trends and offshoring of labor for widgets.

As opposed to M3, which is influenced by economic growth exceeding inflation, population growth, just overall changes in velocity of money, etc...

Correct. Some would argue that CPI is specifically engineered to mislead.

Why do banks and asset managers trade billions in CPI derivatives if it is a useless, engineered metric?

CPI includes housing and M3 is effectively never a good measure of inflation. People have this misconception that the supply of money is directly inversely proportional to the value of one unit of that currency. That is not how that works at all. Inflation is demand driven. When demand falls off a cliff you can have periods of deflation during which money supply increased significantly.

CPI covers Food energy housing etc because people aren’t just buying any one thing. We could argue about the correct mix to use, but M3 alone ignores population and economic growth etc making it a vastly worse measure of inflation.

This becomes really obvious if you look at M3 vs prices over long periods. If you have say 2x as many people and M3 is 2x the size then a 1$ candy bar that’s still 1$ somehow demonstrates inflation and increases in productivity. Except productivity per worker was unchanged only total economic output which is irrelevant to individuals.

>CPI is misleading, M3 is more accurate.

I wonder what happens if we apply that to other metrics...

https://i.imgur.com/2GkpBLZ.png

Uh oh. If we take raw gdp (ie. not adjusted for inflation/cpi), and adjust it by M3 instead, we find that the economy has shrunk by nearly 50% since the mid 90s. That can't be right, can it?

Chart was made in excel from these data sources:

https://fred.stlouisfed.org/series/MABMM301USM189S

https://fred.stlouisfed.org/series/GDPC1

https://fred.stlouisfed.org/series/GDP

edit: fixed axis format.

Reply on news.ycombinator.com