A recent paper in PLoS argues that variation in genes that regulate dopamine (5-HTTLPR) and serotonin neurotransmission (DRD4) influences financial risk taking: at the 5-HTTLPR gene, one homozygous form (two identical copies or alleles) took on 28 percent less risk, while those who had the ‘7-repeat’ form of the DRD4 gene took 25 percent more risk. The Mad Biologist has issues regarding this study.
To me, the total amount of money the subjects could win (or not win) could influence the results. Someone who is risk adverse at $23 might be risk accepting at $2,300: if the participant were to gain what he (the study involved only men) considered to be a large absolute amount even if he adopts a risk strategy, risk taking could increase, and possibly even wipe out the genetic effect (feel free to invent your own ‘just-so’ story).
Biologists have addressed this idea through what is called the norm of reaction, which is simply a description of the pattern of phenotypic expression of a single genotype across a range of environments. In English, this is a characterization of traits in different environments. So people who have two copies of the ‘less risky’ allele for 5-HTTLPR might be very cautious with one level of reward when compared to other people, but with a different reward amount (or regime, etc.), they might not show any difference at all (or perhaps engage in more risky behavior than the other genotypes).
My flippant summary of all of this is that people are lousy study organisms. More seriously, I think scientists have to be very careful about not overinterpreting their results.