Thank you for a wonderful thought provoking talk! In your model of restarting dynamics, have you done any analysis with regards to US historical data estimating the Tau parameter (reallocation of wealth over time) which was discussed in Colm Connaughton’s presentation? It would be very interesting to understand how these dynamics could change/be affected by during different periods of positive vs negative societal reallocation. Thanks again!
Also, a more generalized investing question (if there’s time!). Is there a threshold where “passive” investment in the whole market causes a breakdown in price discovery to the extent that the market can no longer function as a mechanism of capital allocation? Thank you for your time and expertise!
Much to discuss here. I’d like to highlight the connection you make on slide 1 to the EE philosophy of “process first.” I alluded to it in my presentation where I contrast growth optimisation and utility theory.
The dominant way of thinking in orthodox economics, from what I can tell, is utility theory, and that is fundamentally an individualist type of theory: people act the way they do and find themselves in the positions they’re in because of trait-like properties of those people.
Assuming such trait-like properties (preferences etc) can often trivially reproduce observations (especially from panel data, i.e. observations at a snapshot in time).
But our work shows that for many observations we’d like to explain — for instance the “persistent stratification” Jakob mentioned in his talk — these assumptions are not necessary. They emerge from the relevant model — geometric Brownian motion or reallocating GBM, for instance.
Our first approach is not individualist but structural in this sense, and I think that deep down, it is that order of things — understand what the process alone can do without assuming differences among people — that gives ergodicity economics its power to engender such a different worldview.
One more: great use of the word “optimisation.” In some circles optimisation is a dirty word these days, and I don’t think that’s good. It comes from optimising the wrong thing (in EE we often point at irrelevant expected values being optimised, whereas the story changes completely when you optimise over time). So it’s not “don’t optimise” but “optimise the right thing.” You propose reliability over efficiency, and I guess that’s typically a reflection of EE’s mantra of long-term over parallel worlds.
I think your model is tweaked GBM, where you lose half your wealth if you ever go as low as $1000? This increases the stickiness of low wealth (makes it harder to recover from low values); it may be possible to model that continuously through greater concavity of the ergodicity transformation (more concave than log), in our framework.
Can you comment on the link you mention between “asset hoarding” by the older generation (saving for retirement) and reduced income for the younger generation?
Also, I’ll share some simulations with you — it’s really interesting to ask what happens when we replace an annuity/social insurance-type model with defined contributions of individual investment accounts. In aggregate that may look great, but it’s very questionable whether that justifies what happens to the wealth distribution in old age.
Thank you for a wonderful thought provoking talk! In your model of restarting dynamics, have you done any analysis with regards to US historical data estimating the Tau parameter (reallocation of wealth over time) which was discussed in Colm Connaughton’s presentation? It would be very interesting to understand how these dynamics could change/be affected by during different periods of positive vs negative societal reallocation. Thanks again!
Also, a more generalized investing question (if there’s time!). Is there a threshold where “passive” investment in the whole market causes a breakdown in price discovery to the extent that the market can no longer function as a mechanism of capital allocation? Thank you for your time and expertise!
Much to discuss here. I’d like to highlight the connection you make on slide 1 to the EE philosophy of “process first.” I alluded to it in my presentation where I contrast growth optimisation and utility theory.
The dominant way of thinking in orthodox economics, from what I can tell, is utility theory, and that is fundamentally an individualist type of theory: people act the way they do and find themselves in the positions they’re in because of trait-like properties of those people.
Assuming such trait-like properties (preferences etc) can often trivially reproduce observations (especially from panel data, i.e. observations at a snapshot in time).
But our work shows that for many observations we’d like to explain — for instance the “persistent stratification” Jakob mentioned in his talk — these assumptions are not necessary. They emerge from the relevant model — geometric Brownian motion or reallocating GBM, for instance.
Our first approach is not individualist but structural in this sense, and I think that deep down, it is that order of things — understand what the process alone can do without assuming differences among people — that gives ergodicity economics its power to engender such a different worldview.
One more: great use of the word “optimisation.” In some circles optimisation is a dirty word these days, and I don’t think that’s good. It comes from optimising the wrong thing (in EE we often point at irrelevant expected values being optimised, whereas the story changes completely when you optimise over time). So it’s not “don’t optimise” but “optimise the right thing.” You propose reliability over efficiency, and I guess that’s typically a reflection of EE’s mantra of long-term over parallel worlds.
I think your model is tweaked GBM, where you lose half your wealth if you ever go as low as $1000? This increases the stickiness of low wealth (makes it harder to recover from low values); it may be possible to model that continuously through greater concavity of the ergodicity transformation (more concave than log), in our framework.
Can you comment on the link you mention between “asset hoarding” by the older generation (saving for retirement) and reduced income for the younger generation?
Also, I’ll share some simulations with you — it’s really interesting to ask what happens when we replace an annuity/social insurance-type model with defined contributions of individual investment accounts. In aggregate that may look great, but it’s very questionable whether that justifies what happens to the wealth distribution in old age.