Can intuition be measured? Measuring Traders’ intuitive skills on a seemingly unpredictable guessing game - Draft

Nicolas Hamelin, Archana Jois, Andrew Youssef.

Abstract

Recent research has pointed to gut feeling as one of the factor contributing to traders’ success. To date there was little scientific measure of intuitive skills or predictive anticipatory activity. In this research 27 traders from an investment fund in Singapore were asked to participate in a guessing game while their autonomic body reaction were measured using a high accuracy ad hoc Galvanic Skin Response (GSR) detector. They were also asked to fill a self-report survey on their trading habits and demographics. On this guessing game running on mobile phones, traders were given 10 seconds to decide which image out of 4 images would be randomly selected by the app. After 10 seconds one of the 4 images was randomly displayed on the screen for 5 seconds. The guessing game was repeated for 8 rounds. Galvanic Skin Response levels were recorded using Intuition Pro device and app during both the guessing period and the period in which the randomly selected image was displayed.  Once the result were displayed GSR values for right or wrong answers were found to have statistically different values, corresponding to different emotional reactions to a correct or incorrect guess. Surprisingly statistically significant differences were also recorded in the guessing period, that is within the 10 seconds preceding the display of the result. This research shows that higher level of intuition are significantly correlated to higher capital return. The research also shows that traders relying on newsfeed have a lower level of intuition and that holistic practices increase intuitive skills. Most importantly these findings confirm the existence of a predictive anticipatory activity and the importance of interoceptive skills for traders. Finally this research postulates that enhancing intuitive skills could be a game changer for traders.

Methodology

Then results from the quizzes are averaged and a t-test calculates if the difference between right and wrong averages are significantly statistically different:

Then results from the quizzes are averaged and a t-test calculates if the difference between right and wrong averages are significantly statistically different:

The experiment:

We use 2 intuitions pro devices and app (10 microS accuracy) and tested 27 traders from RV Capital Management Private Ltd based in Singapore. They also had to fill a self-report survey (appendix 1)  about their trading habit, trading skills and demographics. Each trader was requested to take a minimum of 3 quizzes of 8 round each.

 

 

Findings: Descriptive statistics

27 traders were involved in the experiment. Traders were assumed to have equal access to information and market liquidity. Out of these traders 93% were men, and 7% women. Most traders came from India 48% followed by Singaporean 22%, 18% were Chinese (including 7% Hong Kong Chinese), 4% were British, 4% Taiwanese and 4% were American. Respondents were highly educated with 85% claiming to have a master degree and the remaining 15% had a Bachelor degree. Regarding age, the majority 74% were in the 30 to 42 year old age bracket, 22% of the respondents were above 42 year old and the remaining 4% between 25 and 30 year old. Most respondent 41% declared to have between 8 to 15 years of trading experience, 30% had more than 15 years’ experience, 19%  had been trading between 4 to 8 years, 7% declared to have between 1 to 3 years trading experience and only 4% had less than 1 year trading experience. Regarding trading habits, 56% declared to be medium risk takers, 33% low risk takers and 11% affirmed to be high risk takers.

 

Out of 104 quizzes, 72 had valid results (t-value <0.05), and 32 had invalid value (no statistical difference reported between right or wrong levels). The following results are for quizzes with valid results, in other words, quizzes for which the differences between average right and average wrong is statistically significant.

 

Is there a statistically significant difference between groups of traders ?:

Prob F =0.0197 < 0.05 with F = 3.15; The difference between experience group is statistically different with inexperienced and highly experienced trader having a lower Intuitive Score than their peers. F(4,67) = 3.15, p = 0.0197.

News Feed

Prob F =0.0004 < 0.05 with F = 6.82 The difference statistically significant, the more news feed are preferred for trading decision, the lesser the intuition. F(3,68) = 6.82, p = 0.0004

Prob F =0.0020< 0.05 with F = 4.74, The difference statistically significant, the more a trader is into holistic practices, the higher his or her intuition. F(4,67) = 64.74, p = 0.0020

Robust Linear regression testing correlation between measured intuition and ordinal independent variables. Greyed rows are statistically significant variable (p < 0.05). Number of observations = 72, p value = 0.0051. R-squared = 0.4428, Adj R-squared = 0.2674, Root MSE = 0.30881, F( 17, 54)  = .52

Copyright © 2017 IPT, Pty Ltd. All rights reserved.

  • Grey Facebook Icon
  • Grey Instagram Icon
  • Grey Twitter Icon
  • Grey YouTube Icon

Enquiries (AUS): 1300 940 239