Magic dust (a fable)
In the mid-1990s, the consensus view was that central banks should target inflation at 2%. Then in the late 1990s, a light rain of “magic dust” descended on the islands of Japan. The dust made people passive and fatalistic, and the economists there suddenly forgot how to create inflation, a previously inconceivable development.
At the time, Western economists were stunned. “What’s wrong with the Japanese? Why don’t they just do X, Y and Z?”
A decade later, the same magic dust fell on North America and Europe (fortunately the southern hemisphere was spared.) Now Western economists forgot how to create inflation. They began to claim that it was impossible when interest rates were zero.
A few lonely economists had a genetic mutation that made them immune to the effects of the magic dust. They wondered why economists had changed their minds about the efficacy of monetary policy. After all, the Fed and ECB had also failed to do X, Y and Z.
Another decade went by, and the effects of the magic dust began to wear off. The next time deflation threatened in an environment of zero rates, central banks began doing some of the X, Y, and Z that they had recommended to the Japanese a few decades earlier, And it quickly created lots of inflation.
Tags:
2. September 2024 at 05:55
For what it’s worth, here’s my latest attempt to help formalize market monetarism. This is my replacement for the Gordon Growth Model:
https://img1.wsimg.com/blobby/go/bd8d1cdb-89b7-40d5-9ad0-3361e16c70ee/downloads/NGDP%20Growth%20Derived%20Stock%20Valuation%20Model.pdf?ver=1725284589259
Or, one can just visit my personal website here, to see it:
https://michaelsandifer.godaddysites.com/
It features NGDP growth expectations as the independent variable, rather than relying on interest rates.
For anyone interested there’s also a short article about my behavioral economic model for emotions, mood, and motivation:
https://img1.wsimg.com/blobby/go/bd8d1cdb-89b7-40d5-9ad0-3361e16c70ee/downloads/Bridging%20Psychology%20and%20Economics.pdf?ver=1725284589259
It does not focus on the actual equations, but you can get some of those details by chatting with one of my chatbots, the link to which is also on my website.
2. September 2024 at 14:17
You have change in G in the denominator. But if that variable is very small (close to zero), doesn’t the ratio become extremely large?
2. September 2024 at 17:12
Scott,
Yes, but the expected earnings fall enough to compensate. This is justfied by the derivation.
For example, if expected NGDP growth falls 50%, stock prices fall 50%, and expected earnings fall 75%. Actual earnings are a great deal more volatile than this, of course.
This reflects the fact that PE ratios can rise due to the expectation of catch up economic growth, adding the market monetarist nuance missing in many foggy debates about the significance of high PE ratios. Like interest retes, PE ratios must be taken in context, most directly with the growth path of earnings.
2. September 2024 at 23:16
Sorry, but there was a typo. The denominator should be just “G”, not “change in G”. I corrected this in the paper.
https://img1.wsimg.com/blobby/go/bd8d1cdb-89b7-40d5-9ad0-3361e16c70ee/downloads/NGDP%20Growth%20Derived%20Stock%20Valuation%20Model.pdf?ver=1725347392600
3. September 2024 at 10:08
This is embarassing, but I got that equation wrong the second time too. Here’s the corret version:
https://img1.wsimg.com/blobby/go/bd8d1cdb-89b7-40d5-9ad0-3361e16c70ee/downloads/NGDP%20Growth%20Derived%20Stock%20Valuation%20Model.pdf?ver=1725386773297
Also, to make up for it, here’s the ability to use the formula as inputted into Excel, via Google Sheets:
https://docs.google.com/spreadsheets/d/1lGHnT6OeqYedx5aunJsNv4j4yVQ_kTZmku757ZhXXug/edit?usp=sharing
That lets people play with it who are interest in seeing how it works.
3. September 2024 at 11:26
Sadly, I’m unusually stupid in some ways, such as with sense of direction and a general absent-mindedness that made me such a dangerous driver, that I had to stop driving sometime ago. Not only did I need GPS to show me how to drive to local places I’d been to 20 times, but I also had trouble concentrating on my driving and would routinely miss turns, even when warned by Google Maps.
I have the opposite of ADD/ADHD, which is dopamine sensitivity, which means I have a tendency to focus too much at a given moment, often on my thoughts at the expense of attention to the external environment. I had the LLMs double-check my derivation, but not my equation, which was dumb. I was in a hurry, given my schedule, and nothing good comes of me trying to rush through such tasks, especially while using LaTeX for the first time to format the document.
It’s no excuse, but an explanation. and I apologize again.
3. September 2024 at 14:31
Michael, No problem. So if we move E to the other side, does it say P/E = G plus delta G over G squared?
4. September 2024 at 11:01
Scott,
No, you would invert the right side, so that you have G^2/(G + delta G).
You can think of it this way. If you had P = E/G, if you just move E to the other side, you’d end up with P/E = 1/G.
4. September 2024 at 11:14
This is a very odd equation, if you think about it. Two of the three variables are unobservables, yet because of the proportional relationship between changes in P and changes in G, the second unobservable is inferred. This equation uniquely depends upon the derivation for confidence that it works. I suspect that means some would have philosophical issues with it. If one doesn’t buy the premise, one doesn’t buy the equation.
This is one reason it took me a long time to develop such a simple equation, because I spent so much time trying to develop one with only a single unobservable, which was expected NGDP growth.
4. September 2024 at 13:38
Why would you invert the right side?
4. September 2024 at 13:58
Scott,
Because otherwise, you’re not being consistent with the relationship between variables.
P = E/G is the same as G = E/P, when the multiplier isn’t obscurring the fact.
4. September 2024 at 14:24
But we aren’t talking about E/P, we are discussing P/E.
5. September 2024 at 21:59
Scott,
Yes, I glanced too quickly and misread it. The whole point of the multiplier is to adjust expected earnings, consistent with the 1:1 relationship between changes in P and G.
7. September 2024 at 06:18
Explicitly you have:
P = (E/G)[(G + delta G)/G]
P/E = 1/(G + delta G)
7. September 2024 at 06:19
Thank you for raising these questions, because it’s clear I’ll need to address them in any introduction to the equation.