the ones most likely to take my advice are the ones who don’t want to be bothered doing it themselves, far more than any other single trait.
Of the people who would look up and carry out the financial advice themselves an the ones who come to an adviser to carry it out, there may be some imbalance in terms of wanting to bother. Both groups are still highly selected compared to the general population: as you say, “Fewer than you’d think bother to seek advice.” The results are entirely explicable by the default of self-selection, and that’s much more plausible than advisors matter all that much. (Consider the example of SAT coaching...)
Give me a practical model for a randomized study, please. Until you have that, let’s work with the evidence we have available. And that evidence seems pretty consistent with my beliefs(that I’ve had since before I started this job) that advisors don’t meaningfully improve investment returns per se, but they mildly improve investor tax planning, and they massively improve investor behaviour.
Self-selection is the default explanation; the onus is on financial planners to show that they are helpful.
Give me a practical model for a randomized study, please.
You could… I don’t know, select some people, offer half of them $1k to go to a financial planner and the others $1k in exchange for reporting on financial health, then see if the experimental group is better off a year later? This is not harder than doing things like deworming studies in Africa.
Of the people who would look up and carry out the financial advice themselves an the ones who come to an adviser to carry it out, there may be some imbalance in terms of wanting to bother. Both groups are still highly selected compared to the general population: as you say, “Fewer than you’d think bother to seek advice.” The results are entirely explicable by the default of self-selection, and that’s much more plausible than advisors matter all that much. (Consider the example of SAT coaching...)
Where’s the randomized beef?
Give me a practical model for a randomized study, please. Until you have that, let’s work with the evidence we have available. And that evidence seems pretty consistent with my beliefs(that I’ve had since before I started this job) that advisors don’t meaningfully improve investment returns per se, but they mildly improve investor tax planning, and they massively improve investor behaviour.
Self-selection is the default explanation; the onus is on financial planners to show that they are helpful.
You could… I don’t know, select some people, offer half of them $1k to go to a financial planner and the others $1k in exchange for reporting on financial health, then see if the experimental group is better off a year later? This is not harder than doing things like deworming studies in Africa.