Nice piece J. But who are “some 3,400 Americans”? Are they retail investors even? Even if they are, do they swing a sufficiently large and relevant bat?
the study states that they recruited participants with experience in equity investments through the Prolific platform. Prolific, together with MTurk have become the standard go-to recruitment platform to get a representative sample of the American population with specific required skills. Also, the participants were incentivised with monetary gains to make the best possible forecasts and give reliable answers. All standard procedure for these kinds of studies.
Brokerage economist forecasts for GDP growth are always +2%. Brokerage strategist forecasts for market indicies are always +5%. Brokerage analyst estimates for buy-rated stock price target upsides are always +20%. As with weather forecasters, there's seldom if ever a penalty for failure. "It is difficult to make predictions, especially about the future." -- Not Niels Bohr, Samuel Goldwyn, K. K. Steincke, Robert Storm Petersen, Yogi Berra, Mark Twain, nor Nostradamus ... some unknown Danish person c. 1937-38 https://quoteinvestigator.com/2013/10/20/no-predict/
I use all forecasts, regardless of whether the author is a human or a machine, in the same way as I do comments from strangers on t'internet or people on the street or friends/family/colleagues - as a piece of information (sometimes entertainment) helping me to make my own decisions.
I trust it the same across the board... and my standard setting is that I believe everyone making a forecast has their own agenda or is influenced by other people (even if they don't recognise this fact) - even machines.
Nice piece J. But who are “some 3,400 Americans”? Are they retail investors even? Even if they are, do they swing a sufficiently large and relevant bat?
the study states that they recruited participants with experience in equity investments through the Prolific platform. Prolific, together with MTurk have become the standard go-to recruitment platform to get a representative sample of the American population with specific required skills. Also, the participants were incentivised with monetary gains to make the best possible forecasts and give reliable answers. All standard procedure for these kinds of studies.
Brokerage economist forecasts for GDP growth are always +2%. Brokerage strategist forecasts for market indicies are always +5%. Brokerage analyst estimates for buy-rated stock price target upsides are always +20%. As with weather forecasters, there's seldom if ever a penalty for failure. "It is difficult to make predictions, especially about the future." -- Not Niels Bohr, Samuel Goldwyn, K. K. Steincke, Robert Storm Petersen, Yogi Berra, Mark Twain, nor Nostradamus ... some unknown Danish person c. 1937-38 https://quoteinvestigator.com/2013/10/20/no-predict/
weird, AI stated the ratings were reversed due to human bias.
to get more alignment , maybe the next study should threaten to pull the plug on humans and AI models that are incorrect.
:)
I use all forecasts, regardless of whether the author is a human or a machine, in the same way as I do comments from strangers on t'internet or people on the street or friends/family/colleagues - as a piece of information (sometimes entertainment) helping me to make my own decisions.
I trust it the same across the board... and my standard setting is that I believe everyone making a forecast has their own agenda or is influenced by other people (even if they don't recognise this fact) - even machines.