6 ms·

It didn't. It contracted by 9.5% in Q2, which, if repeated for four consecutive quarters, would become 32.9%. Annualized figures might make sense in ordinary t

It didn't. It contracted by 9.5% in Q2, which, if repeated for four consecutive quarters, would become 32.9%. Annualized figures might make sense in ordinary times, but these are not ordinary times.

kgwgk 6y ago
[this is terribly wrong: “It did contract by 32.9% in the second quarter (not in the four quarters ending in June 2020)”] Edit: I assumed it was compared to the second quarter 2019, not sequentially to the first quarter 2020, I didn’t stop to think that 33% was just too much. 9% is bad enough: from $21.34 trillion to $19.41 trillion (Q2 2019 to Q2 2020).
- rayiner 6y ago
  No. “The U.S. economy contracted at a record 32.9% annual rate last quarter.”
- MrPowers 6y ago
  The title of the article says "U.S. Economy Contracts at Record 32.9% Annual Rate". A HN mod should change the title here on this site cause it's misleading.
- FabHK 6y ago
  No. Goes to show how misleading the headline is (particularly the HN headline - was there no space for the important clarification "annual rate"?) @Dang? add "annualised" or "annual rate"?
  - nickles 6y ago
    The article headline changed since this was posted. I've updated it to reflect the new title.
- function_seven 6y ago
  No it didn’t. Here’s NYT’s headline: U.S. economic output fell 9.5 percent in the second quarter, the biggest drop on record. That translates to a 32.9 percent annual rate of decline.
RivieraKid 6y ago
Exactly... it makes no sense to report an annualized number on this case.
pwaivers 6y ago
Here's a pretty good explanation of how the figures are reported (and why it can be misleading): https://www.nytimes.com/2020/07/29/business/economy/us-gdp-report.html https://www.nytimes.com/2020/07/29/business/economy/us-gdp-r....
draw_down 6y ago
Fortunately this pandemic will only last a single quarter!
Zigurd 6y ago
The headline is wrong. The number isn't. The headline should have said "Contracted at an annual rate of.." To make snapshots comparable, they are stated at an annual rate, even if the snapshot is monthly or quarterly. Actually it looks like just the HN headline is wrong. The article states it correctly.
- nickles 6y ago
  The headline changed since this was posted. It's been updated to reflect the new title.
koheripbal 6y ago
To be clear, in finance, we _always_ quote annualized numbers. It would be super unconventional and misleading to everyone in finance to list an absolute number as you suggest. This is more like the general populace not understanding how finance works rather than the finance press exaggerating the number.
- vladd 6y ago
  In finance, you would compare 2020-q2 with 2019-q2. You would still get a similar 8% - 9% drop. That figure (32.9%) is based on future projections. Stating the predicted future as something that took place already would get someone fired in finance. >> This is more like the general populace not understanding how finance works It looks to me more closely as someone applying mathematical formulas either without understanding what you are doing de-facto behind them, or being ill intended in the first place to manipulate the market. Either way, it's bad. Unless you're explicitly writing "in Q2, the GDP dropped at an annualized rate of -33%", you're committing the capital offense when it comes to writing good journalism: not knowing your average audience, not adapting your content to the audience and probably most significantly, not understanding as a journalist what you're writing about.
- em500 6y ago
  Note that national accounting growth rate reporting conventions are local. In the US, official GDP comes from bea.gov, which annualizes the quarter-on-quarter growth by default. In the UK it's done by ons.gov.uk, which reports (seasonal-adjusted) quarter-on-quarter without annualizing. In China, www.stats.gov.cn used to report quarter-on-quarter in the past, but they changed to year-over-year rates (which is different from annualized quarter-on-quarter) a few years ago.
castratikron 6y ago
Let's say they reported 9.5% . Would it still not be the biggest drop in history? Isn't that the important bit here?
qertoip 6y ago
9.5% x 4 = 38% How exactly annualization works?
- bhupy 6y ago
  It presupposes that the quarterly trend will continue at the same rate. In finance, annualizing is a good way to try and extrapolate from the current trend, but in this unprecedented pandemic, the current trend is extraordinarily volatile.
  - pieter_mj 6y ago
    well yes, but as long as there's no vaccine, it remains fairly predictable it seems.
- Fishkins 6y ago
  1-.095 = .905 .905^4 = .671 1-.671 = .329 = 32.9%
  - qertoip 6y ago
    Of course. This was probably my dumbest question in 2020.
    - pieter_mj 6y ago
      Still a long way to go, you might top it. Thanks for the clarification in grandparent comment, i wasn't aware.
  - dkn775 6y ago
    What is the purpose is using the 0.095 figure?
    - hungryhobo 6y ago
      the economy contracted 9.5% -> 0.095
JeremyNT 6y ago
Was the title of this post changed? I can't really understand why this comment is at the top, since "it didn't" and "contracted by 9.5%" seems to refer to some specific complaint about the thesis of the article. It's typical for us in the HN crowd to have a superficial knowledge of a domain and get lost in quibbling about the minutae of how the data is represented, but the thesis of this article is the record decline, and focusing on the use of the annualized metric feels like a tangent.

It didn't. It contracted by 9.5% in Q2, which, if repeated for four consecutive quarters, would become 32.9%. Annualized figures might make sense in ordinary times, but these are not ordinary times.

kgwgk 6y ago

[this is terribly wrong: “It did contract by 32.9% in the second quarter (not in the four quarters ending in June 2020)”] Edit: I assumed it was compared to the second quarter 2019, not sequentially to the first quarter 2020, I didn’t stop to think that 33% was just too much. 9% is bad enough: from $21.34 trillion to $19.41 trillion (Q2 2019 to Q2 2020).

rayiner 6y ago
No. “The U.S. economy contracted at a record 32.9% annual rate last quarter.”
MrPowers 6y ago
The title of the article says "U.S. Economy Contracts at Record 32.9% Annual Rate". A HN mod should change the title here on this site cause it's misleading.
FabHK 6y ago
No. Goes to show how misleading the headline is (particularly the HN headline - was there no space for the important clarification "annual rate"?) @Dang? add "annualised" or "annual rate"?
- nickles 6y ago
  The article headline changed since this was posted. I've updated it to reflect the new title.
function_seven 6y ago
No it didn’t. Here’s NYT’s headline: U.S. economic output fell 9.5 percent in the second quarter, the biggest drop on record. That translates to a 32.9 percent annual rate of decline.

RivieraKid 6y ago

Exactly... it makes no sense to report an annualized number on this case.

pwaivers 6y ago

Here's a pretty good explanation of how the figures are reported (and why it can be misleading): https://www.nytimes.com/2020/07/29/business/economy/us-gdp-report.html https://www.nytimes.com/2020/07/29/business/economy/us-gdp-r....

draw_down 6y ago

Fortunately this pandemic will only last a single quarter!

Zigurd 6y ago

The headline is wrong. The number isn't. The headline should have said "Contracted at an annual rate of.." To make snapshots comparable, they are stated at an annual rate, even if the snapshot is monthly or quarterly. Actually it looks like just the HN headline is wrong. The article states it correctly.

nickles 6y ago
The headline changed since this was posted. It's been updated to reflect the new title.

koheripbal 6y ago

To be clear, in finance, we _always_ quote annualized numbers. It would be super unconventional and misleading to everyone in finance to list an absolute number as you suggest. This is more like the general populace not understanding how finance works rather than the finance press exaggerating the number.

vladd 6y ago
In finance, you would compare 2020-q2 with 2019-q2. You would still get a similar 8% - 9% drop. That figure (32.9%) is based on future projections. Stating the predicted future as something that took place already would get someone fired in finance. >> This is more like the general populace not understanding how finance works It looks to me more closely as someone applying mathematical formulas either without understanding what you are doing de-facto behind them, or being ill intended in the first place to manipulate the market. Either way, it's bad. Unless you're explicitly writing "in Q2, the GDP dropped at an annualized rate of -33%", you're committing the capital offense when it comes to writing good journalism: not knowing your average audience, not adapting your content to the audience and probably most significantly, not understanding as a journalist what you're writing about.
em500 6y ago
Note that national accounting growth rate reporting conventions are local. In the US, official GDP comes from bea.gov, which annualizes the quarter-on-quarter growth by default. In the UK it's done by ons.gov.uk, which reports (seasonal-adjusted) quarter-on-quarter without annualizing. In China, www.stats.gov.cn used to report quarter-on-quarter in the past, but they changed to year-over-year rates (which is different from annualized quarter-on-quarter) a few years ago.

castratikron 6y ago

Let's say they reported 9.5% . Would it still not be the biggest drop in history? Isn't that the important bit here?

qertoip 6y ago

9.5% x 4 = 38% How exactly annualization works?

bhupy 6y ago
It presupposes that the quarterly trend will continue at the same rate. In finance, annualizing is a good way to try and extrapolate from the current trend, but in this unprecedented pandemic, the current trend is extraordinarily volatile.
- pieter_mj 6y ago
  well yes, but as long as there's no vaccine, it remains fairly predictable it seems.
Fishkins 6y ago
1-.095 = .905 .905^4 = .671 1-.671 = .329 = 32.9%
- qertoip 6y ago
  Of course. This was probably my dumbest question in 2020.
  - pieter_mj 6y ago
    Still a long way to go, you might top it. Thanks for the clarification in grandparent comment, i wasn't aware.
- dkn775 6y ago
  What is the purpose is using the 0.095 figure?
  - hungryhobo 6y ago
    the economy contracted 9.5% -> 0.095

JeremyNT 6y ago

Was the title of this post changed? I can't really understand why this comment is at the top, since "it didn't" and "contracted by 9.5%" seems to refer to some specific complaint about the thesis of the article. It's typical for us in the HN crowd to have a superficial knowledge of a domain and get lost in quibbling about the minutae of how the data is represented, but the thesis of this article is the record decline, and focusing on the use of the annualized metric feels like a tangent.