From the article:
The study used experimental stock markets with some 140 male participants (called traders) taking part. Cohorts of traders were given testosterone or placebo gel prior to their trading sessions. In total, the experiment consisted of 17 sessions that allowed traders to buy, sell, bid, and offer money for shares of stock, similar to a simplified professional trading platform. Each session had three rounds of 12 trading periods where traders competed to make money against each other.
The results showed the causal effects of testosterone on financial asset mispricing. By administering testosterone to traders before they trade financial assets for real money, testosterone directly increases the size and persistence of stock market bubbles. Testosterone drove these changes in market dynamics by increasing bidding, selling prices, and volume and changed traders' perception of a stock’s current value even though true values were known during trading.