When a member of the New York Fed staff releases a paper on the topic of Anxiety, Overconfidence, and Excessive Risk Taking [7], and in which there is a section on "Self-Manipulation with Alcohol and Drugs", which explains that "pathological gambling is more common among people with alcohol use disorders," adds that "there is evidence that drugs are used strategically to induce performance changes, and particularly so for individuals with greater degrees of horizon-dependent risk aversion", observes that 'Anecdotal evidence on the “widespread use of [...] cocaine by professional traders” is consistent both with strategic self-manipulation and with our observations about cross-sectional overconfidence across environments", and finally recommends to all the risk-averse BTFDers and BTFATHers, that "the performance of anxiety-prone individuals should then improve with moderate levels of drug-induced overconfidence" one should probably listen and trade - since the Fed's only remaining wealth effect and "monetary transmission channel" is to BTFATH - while both drunk and high.
From the NY Fed white paper [7]:
Self-Manipulation with Alcohol and Drugs
A second interpretation of how the belief manipulation of our model may be implemented in practice is through the use of alcohol and other drugs. This section gives a brief review of psychological evidence on the effect of alcohol and other drugs on (i) risky behavior, (ii) forgetting and confidence, and (iii) performance changes. In addition, we discuss evidence on the strategic use of alcohol and other drugs by anxiety-prone individuals to induce effects (i)–(iii).
The finding that alcohol is associated with more risky behavior is robust across domains. In the field, alcohol consumption has been shown to lead to risky sexual behavior (Halpern-Felsher et al., 1996; Cooper, 2002), accident-related injuries (Cherpitel et al., 1995), and dangerous driving patterns (Donovan et al., 1983). Pathological gambling is more common among people with alcohol use disorders, and vice versa (Grant et al., 2002; Petry et al., 2005). In the lab, Lane et al. (2004) establish causality from alcohol consumption to risky behavior.
Riskier behavior can be driven either by reductions in risk aversion, or by a decreased perception of risk. Cohen et al. (1958) show that the more risky driving behavior caused by alcohol consumption is associated with a higher degree of overconfidence. Supporting the channel suggested by our model, alcohol has also been shown to lead to forgetting, especially of negative signals (Nelson et al., 1986; Maylor and Rabbitt, 1987).
In the context of our model, the performance of anxiety-prone individuals should then improve with moderate levels of drug-induced overconfidence. James et al. (1977) as well as Brantigan et al. (1982) show that the use of beta-blockers improves the erformance of musicians with stage fright.
Lastly, there is evidence that drugs are used strategically to induce performance changes, and particularly so for individuals with greater degrees of horizon- dependent risk aversion. Steptoe and Fidler (1987) find that 17 percent of professional musicians with high performance anxiety reported taking sedatives as a method of coping, compared with 4 percent with medium and none of the respondents with low performance anxiety. Anecdotal evidence on the “widespread use of [...] cocaine by professional traders” (Bossaerts, 2009) is consistent both with strategic self-manipulation and with our observations about cross-sectional overconfidence across environments.
Needless to say, when the NY Fed discusses "self-manipulation" using blow and booze, listen! They are completely clueless about everything else, but on this topic, they are consummate experts...
