If you are an employer hiring untested potential traders, I suppose you may want to have a test. As an employer, I would not allow one. However, if you are an individual Trader-To-be, screw the test. Tenacity always wins...everytime
well... to me there would be at least one important test, the test of 'character', notably the ability to objectively analyse and accept one's mistakes, and move on... not a test that can be taken on paper though... now as regards this pilot study, hounddogone is correct that it's flawed in a number of ways of course, however to me it seems ok to use a population of noobs and try & identify which traits seem to be related to better relative performance... but yeah, they didn't have the pressure of real money, let alone their own money! but then again, someone applying strict money management rules such as never risking any more than a 2% loss per trade (i.e. NOT me ) wouldn't be under that much pressure either... but yeah perhaps 38 pages was a bit overkill ;-) hope nobody made a print out!
I thought these tests went out years ago. The problem has always been the same person will answer the same questions differently depending on their given circumstances at the time they take the test. In other words raise the temperature in the room by ten degrees and you will get a different set of responses. These things actually come in and out of vogue every decade or so, and are usually linked to the Human Resources Industry needing a different idea to market, sort of like the stupid Corporate Leadership Survival weekends of a few years ago. Come on fools garb the rope ladder and then join the drum circle.
You wanna talk about correlations between personality and successful trading? Let's assume for the sake of argument that the class of all ET members who post during RMH day includes some successful traders (I can personally attest that there is at least one such). Now I think we can reasonably assume given the general tendency to potty-mouth, and the amount of time spent surfing and posting on ET, that they are not logging on from the workplace, or they would have been fired (indeed, perhaps they were at some past time). So they are probably trading from home. Jobless. Desperately struggling to trade successfully. But not managing to apply their full attention to trading, or they wouldn't be posting. I could continue this line of reasoning, but you know yourself where is is going because I am talking about YOU. A social loser. Bathes infrequently. An auto-erotic. A loner. With delusions of grandeur. Self-deluded, seing systems that aren't there. Hasn't checked the brokerage account balance lately. But, hey, doing great! Likes to brag on ET, but never backs it up with verifiable substance. But SOME such misfits MUST be profitable. So what does that tell you? About as much as this stupid fucking thread (Nononsense, as usual, excluded from the indictment. Quit playing with yourself, NN!).
a couple of things, beyond all the applicable caveats... take a look at this part in particular (starting p21), there are a few insights... could be BS of course, but could also be relevant... just food for thought between two infrequent baths ;-) "So, four of our six personality traits indeed make a difference, with locus of control being the only exception. The face value of the three personality profiles, interpreting the score differences as to the four personality traits that discriminate across our three clusters, is satisfactory. Looking at the below and above-average scores per cluster, we have the following interpretation. 1. Profile 1 (n = 7) is characterized by individuals that try to avoid regret, have a dislike for sensation seeking, and reveal type-B behavior. Hence, Profile 1 reflects relaxed, risk-averse persons. 2. Profile 2 (n = 19) captures individuals high on self-monitoring and sensation seeking. So, Profile 2 implies self-conscious, risk-loving persons. 3. Profile 3 (n = 6) includes individuals who are associated with clear type-A behavior. Therefore, Profile 3 relates to impatient, highly competitive persons. Table 2 provides means, standard deviations and correlations. [INSERT TABLE 2 ABOUT HERE] Multicollinearity is not an issue. A number of interesting significant correlations emerge. For instance, nicely in line with the interpretation of our personality profiles, sensation seeking and self-monitoring are positively correlated. This provides further evidence for the face validity of our personality profile classification. We also find gender to be negatively correlated with locus of control â in line with previous findings that, on average, women score lower on this trait â that is, are more likely to be externally minded. Below, therefore, we run preliminary multivariate analyses with this set of three personality profiles as our key independent variables, with Profile 1 and Profile 3 included, and Profile 2 left out as the reference category. 5. PRELIMINARY EVIDENCE Due to the small sample, we were quite limited in the extent to which we could use more complex statistical methods or larger batteries of variables for investigating our primary research question â that is, the impact of individual participantsâ demographic and personality features on their trading performance. Among the demographic characteristics, we mainly focused on Age and Gender as two standard demographic variables. Gender, in particular, has been shown to potentially affect (trading) performance (Barber and Odean, 2001). Running an OLS regression with only these two independent demograhic variables (and a constant) gives the results listed under Model 1 in Table 3. [INSERT TABLE 3 ABOUT HERE] Neither the overall model nor of the individual coefficients are significant. This is not surprising given the small sample size, relative homogeneity of the sample, especially in terms of Age, and the dominating influence of the Percentage of lucky draws (see below). Before turning to the full Model 3, which includes all control and independent variables (i.e., demographic features, personality profiles and Percentage of lucky draws), we discuss Model 2, which contains both demographic variables and the three presonality profiles in the form of two dummies for Profile 1 and Profile 3. We find that adding the personality perspective enhances the overall modelâs ability to explain variation in payments considerably. More importantly, the third personality profile, which reflects type-A individuals with a tendency toward impatient, highly competitive behavior, has a significant negative effect on trading performance, compared to the default Profile 2. Note that the participants with this third profile were spread over four of our five sessions. Profile 1 has a positive, though insignificant effect on trading performance. In Model 3, finally, we added the Percentage of lucky draws control variable. This variable accounted for the fact that differences in payment arose not only from differences in participantsâ trading performance, but also from variation in the number of realized dividend payments (following from the random lottery draw) across sessions. Unfortuantely, in our pilot, this percentage varied widely, ranging from 16 to more than 50 per cent of all lottery draws per session. As a result, this variable has a very strongly positive and highly significant impact on our performance measure, which obscures the comparatively more subtle influences of the personality profiles. Nevertheless, the significance level of the dummy variable for Profile 3, though not satisfactory by the standard rules of thumb, does not deteriorate very drastically, even in a sample as small as ours. 6. APPRAISAL In this paper, we suggested to integrate insights from personality psychology into behavioral finance. To date, behavioral finance has been dominated by cross-pollination with cognitive psychology. Following a long-standing tradition in behavioral management, we argued that, next to mechanisms suggested by cognitive psychology, personality traits may offer complementary explanations for trader behavior and trading performance. For the sake of the argument, after presenting our general framework, we focused on six examples of such potentially relevant personality traits: locus of control, optimization preference, regret attitude, self-monitoring, sensation seeking, and type-A/B behavior. Given our low number of observations, we reduced this set of six variable to a three-category personality profile classification, each of which represents a well-interpretable combination of four of our six personality traits. Using this profile classification in preliminary multivariate analyses indeed revealed an interesting finding. That is, Profile 3 individuals perform significantly worse than their Profile 1 and Profile 2 counterparts. The interpretation of this significant result may run as follows. Profile 3 individuals stand out for their type-A behavior. Apparently, in our noisy setting, such impatient and highly competitive traders are outperformed by their more patient and less competitive counterparts. If this interpretation is correct, one would expect that Profile 3 persons trade more, given their âcompetitive impatienceâ, than Profile 1 and Profile 2 individuals. Indeed, in our pilot, our Profile 3 participants bought x and sold y units, on average, vis-Ã -vis x sales and and y buys for their Profile 1 and Profile 2 counterparts, as a combined group. Additionally, we believe that the insignificance of Locus of control and Optimization preference, which dropped out from the cluster analysis, is interesting as well. After all, we are in the early stages of behavioral finance â personality psychology research. Part of the job, therefore, is to find out, step by step, which personality traits do indeed matter, and which do not, perhaps conditional on the settingâs microstructure. In our experimental setting, at least, locus of control and optimization preference are irrelevant. Of course, whether or not this is a robust finding cannot be judged on the basis of a preliminary pilot study as ours. Future work is needed to further explore this issue. Of course, we were not able to run full-blown multivariate regression analyses, due to the low number number (32) of obsevations in our pilot study. Another reason for our inability to produce better results is that the experimental setup we used is associated with too much noise. That is, the random variation in the potential returns to trading was so large that the relative impact of other potential derminants of performance (as reflected in dominant effect of the Percentage of lucky draws control variable), such as our personality profiles, became too low. In the near future, we plan to replicate the current esxperiment with a much larger number of participants and with a more noise-free experimental setup. Then, we can not only include our full set of personality trait variables, but also a series of other measures that have emerged as relevant and powerful in behavioral management and personality psychology research. Two examples are educational background (Frank, Gilovich and Regan, 1993; Boone and van Witteloostuijn, 1999) and need-for-closure (Ford and Kruglanski, 1995; Kruglanski, 1996). Take the example of need for closure, by way of illustration. Need-for-closure relates to an individualâs desire to come quickly to a closure in decisions and judgments. It has been characterized as an individualâs need to settle for any answer, rather than remain in a state of ambiguity. Need-for-closure has been shown to be triggered by situational context (e.g., time pressure), or dullness of a cognitive task. At the other extreme, individuals with very low need-for-closure exhibit a tendency to postpone decisions. Both conditions carry costs, and need-for-closure especially, has been associated with judgmental mistakes (van Hiel and Mervielde, 2002). Apart from this replication and extension type of work, we would like to conclude with two further suggestions for future work. "