I’m not sure we have much evidence in whether actual prediction markets reliably benefit from an influx of new participants. I suspect it’s as complicated as other endeavors: it’ll depend on the selection and expectations of those new people, and how much training and/or accommodation is needed for them.
In my company, we often talk about “maximum team onboarding rate” in terms of how quickly we can bring new team members up to productivity and retain our team goals and culture. We do pretty reliably grow in scope, but not unboundedly and not without quite a bit of care in terms of selection and grooming of new members.
I don’t know if this counts as evidence, per se, but DeLong, Schleiffer, Summers and Waldman had a fairly seminal paper on this in 1987: The Economic Consequences of Noise Traders. In it, they explain how the addition of “noise traders” (i.e. traders who trade randomly) can make financial markets less efficient. Conventional economic theory, at the time, held that the presence of noise traders didn’t reduce the efficiency of the market, because rational investors would be able to profit off the noise traders and prices would still converge to their true value.
In the paper, DeLong, et. al. demonstrate that it’s possible for noise traders to earn higher returns than rational investors, and, in the process significantly affect asset prices. Key to their insight is that, in the real world, investors have limited amounts of capital, and the presence of noise traders significantly raises the amount of risk that rational investors have to take on in order to invest in markets that have large numbers of noise traders. This risk causes potentially wide, but not permanent divergences between asset prices and fundamental values, which can serve to drive rational investors from the market.
I don’t see any reason to believe that prediction markets would behave differently from the stock markets that DeLong et. al’s paper targeted. My hypothesis would be that prediction markets have shown increasing accuracy with increasing participation so far, but that relationship will break down once the relatively limited pool of people who are willing to think before they trade is exhausted and further increases in prediction market participation draw from a pool of noise traders.
I’m not sure we have much evidence in whether actual prediction markets reliably benefit from an influx of new participants. I suspect it’s as complicated as other endeavors: it’ll depend on the selection and expectations of those new people, and how much training and/or accommodation is needed for them.
In my company, we often talk about “maximum team onboarding rate” in terms of how quickly we can bring new team members up to productivity and retain our team goals and culture. We do pretty reliably grow in scope, but not unboundedly and not without quite a bit of care in terms of selection and grooming of new members.
I don’t know if this counts as evidence, per se, but DeLong, Schleiffer, Summers and Waldman had a fairly seminal paper on this in 1987: The Economic Consequences of Noise Traders. In it, they explain how the addition of “noise traders” (i.e. traders who trade randomly) can make financial markets less efficient. Conventional economic theory, at the time, held that the presence of noise traders didn’t reduce the efficiency of the market, because rational investors would be able to profit off the noise traders and prices would still converge to their true value.
In the paper, DeLong, et. al. demonstrate that it’s possible for noise traders to earn higher returns than rational investors, and, in the process significantly affect asset prices. Key to their insight is that, in the real world, investors have limited amounts of capital, and the presence of noise traders significantly raises the amount of risk that rational investors have to take on in order to invest in markets that have large numbers of noise traders. This risk causes potentially wide, but not permanent divergences between asset prices and fundamental values, which can serve to drive rational investors from the market.
I don’t see any reason to believe that prediction markets would behave differently from the stock markets that DeLong et. al’s paper targeted. My hypothesis would be that prediction markets have shown increasing accuracy with increasing participation so far, but that relationship will break down once the relatively limited pool of people who are willing to think before they trade is exhausted and further increases in prediction market participation draw from a pool of noise traders.