[idea] Econometrics
gdp = c + i + g + (im - x)
to forecast change in gdp, fill in _change_ of each component
c = consumer spending (roughly 75% of gdp)
i = investment
g = government spending
(im - x) = imports - exports
1. easy mode
change in gdp == change in consumer spending
gets most of the way there
consumer spending == -effects of inflation - effects recession - effects of higher interest rates
consumer spending == -5% - 5% - 5% == -15%
change in gdp == -15%
2. hard(er) mode
change in gdp == change in consumer spending + change in investment + change in government spending
change in gdp == -15% + change in investment + change in goverment spending
change in investment == no change but may move to shorter time horizons
government spending == no change to falling because of inflation pushing up borrowing costs and being inflationary
change in gdp == -15% + 0% + 0% == -15%
3. time horizons
3 mos
6 mos
12 mos
24 mos
48 mos
4. gdp time series
historical sanity check
https://fred.stlouisfed.org/series/GDP
2008 - 2009: -3%
2020 - 2021: - 11%
5. stock market (sp500)
take overlapping gdp change and stock market periods
2008 - 2009: -38%
2020 - 2021: +16% (money printing + government spending)
2020-2021: -26% (3 mos pandemic)
average: -38% - 26% = -64%/2 = -32%
https://www.macrotrends.net/2324/sp-500-historical-chart-data
6. hypothesis
gdp falls 15%
stock market (sp 500) falls 32% to 3,261
7. effects
safety net beneath safety net: these are the local charities that provide aid government does not - already strained, get strained some more
8. Bayes Theorem
new evidence updates prior probabilities
i.e. continuous not discrete
9. disclaimer
mental exercise - not investment advice
No comments.