Neurofinance Lecture 4 2024 PDF

Neuronance Faiza Majid School of Banking and Finance, UNSW Faiza Majid FINS3655 Behavioral Finance Neuronance What is Neuronance? Neuronance is an interdisciplinary eld that combines neuroscience and nance to understand how brain process...

Neuronance Faiza Majid School of Banking and Finance, UNSW Faiza Majid FINS3655 Behavioral Finance Neuronance What is Neuronance? Neuronance is an interdisciplinary eld that combines neuroscience and nance to understand how brain processes inuence nancial decision-making. It explores the neural mechanisms behind risk-taking, reward processing, and emotional responses in nancial contexts. Faiza Majid FINS3655 Behavioral Finance The Brain Faiza Majid FINS3655 Behavioral Finance The Brain - PFC The human brain shares old brain regions with other animals but diers primarily in the large and evolutionary new prefrontal cortex (PFC) PFC and parietal cortex = the rational part of the brain: Decides which action to take through assessing the consequences of each potential action. This assessment hinges on accessing information about expected risk and returns: What do I expect to get? How much risk am I facing? That information takes the form of anticipatory emotions or somatic markers These somatic markers play a pivotal role in guiding our decisions. Faiza Majid FINS3655 Behavioral Finance The Brain - PFC Two kinds of somatic markers inform the PFC: Reward signals from imagining potential gains. Brain signals reward opportunities through Dopamine (DA) release. This system helps us prioritize actions that increase our chances of survival, like eating or socializing, and extends to modern behaviors, such as nancial decision-making. Nucleus Accumbens: A central hub for reward processing where dopamine acts. Its activated when we anticipate or receive rewards. Threat detection signals from imagining potential losses and feeling the uncertainty Brain signals danger through Serotonin release Amygdala: The central hub of the threat system. It processes emotional reactions, particularly fear and anxiety, and responds rapidly to perceived threats. Faiza Majid FINS3655 Behavioral Finance The Reward System What keeps the gambler coming back to the tables? The anticipation of reward... Receiving a reward is pleasurable and hence reinforcing:makes us feel good so our brain reminds us to keep doing it Reward or wanting system: midbrain region of the brain that signals rewards through dopamine (DA) release Features a positive feedback loop whereby reward seeking begets more reward seeking Can be modulated by the PFC: as we mature and our PFC wires up, we learn to say no to some of the rewards; PFC moderates how strongly rewards are felt (self-control) Faiza Majid FINS3655 Behavioral Finance The Reward System Taking drugs such as cocaine, driving too fast, watching erotic pictures, winning at the casino, getting good nancial returns... All these activities hijack the reward system by triggering DA release In functional magnetic resonance imaging research, nucleus accumbens (NAcc) activation spontaneously increases before nancial risk taking. By stimulating the reward system of subjects through showing them pornographic pictures, Knutson et al, 2008 made their subjects more risk-seekers in a nancial decision-making task. Faiza Majid FINS3655 Behavioral Finance The Fear System What pushes us to stay away from dangerous bets? The anticipation of being hurt... Fight-or-ight response: We react much faster out of fear than our conscious mind is able to perceive The same neural circuits are triggered when threatened in a war zone as when feeling risk. Two kinds of threat detectors: loss signals when anticipating a loss uncertainty signals related to the threat of the unknown Like for the reward system, as we mature the PFC can modulate the fear signals Faiza Majid FINS3655 Behavioral Finance Somatic Markers Experiencing somatic markers means being informed about the risks In that sense one may argue the somatic markers are always benecial to decision-making (so long as we dont blindly listen to them) Bechara ea, 1997: Somatic markers guide rational behavior in a decision making task Faiza Majid FINS3655 Behavioral Finance Iowa Gambling Task, Bechara et al, 1997 Study in which patients suering damages to the PFC and non-brain damaged subjects played a money game Participants are presented with four virtual decks of cards on a computer screen. They are told that each deck holds cards that will either reward or penalize them, using game money. The goal of the game is to win as much money as possible. The decks dier from each other in the balance of reward versus penalty cards. Thus, some decks are bad decks, and other decks are good decks, because some decks will tend to reward the player more often than other decks. Faiza Majid FINS3655 Behavioral Finance Iowa Gambling Task, Bechara et al., 1997 Common ndings Most healthy participants sample cards from each deck, and after about 40 or 50 selections are fairly good at sticking to the good decks Patients with orbitofrontal cortex (OFC) dysfunction, however, continue to persevere with the bad decks, sometimes even though they know that they are losing money overall. In healthy participants, galvanic skin conductance response show a stress reaction to hovering over the bad decks By contrast, patients with amygdala lesions never develop this physiological reaction to impending punishment. Bechara et al., 1997 explain these ndings in terms of the somatic marker hypothesis i.e. Emotional processes guide behavior, particularly decision making. Faiza Majid FINS3655 Behavioral Finance Iowa Gambling Task, Bechara et al., 1997 Bechara et al., 1997 explain these ndings in terms of the somatic marker hypothesis i.e. Emotional processes guide behavior, particularly decision making. Note: Patients knew the risky decks were risky in the later stage of the game; where they diered was in the arousal experienced immediately before making their choice. Faiza Majid FINS3655 Behavioral Finance Social Preferences Faiza Majid FINS3655 Behavioral Finance Social Preferences Ultimatum Game Sanfey et al., 2003 Proposer (P) proposes x Receiver (R) accepts or rejects If R rejected both get 0, if R accepted, R gets x, and P gets 10-x 30 subjects neuroimaged (fMRI) in the responder position of the ultimatum game Faiza Majid FINS3655 Behavioral Finance Ultimatum Game Neuro results: Unfair oers from human Proposers led to activation in 2 parts of the brain: 1 The dorsolateral prefrontal cortex - associated with cognitive processes such as goal maintenance and control. 2 Bilateral anterior insula – associated with negative emotional states such as anger, physical pain, and disgust. The strength of activity in the anterior insula directly correlated with the likelihood that subjects rejected the oer Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Social preferences in Financing Decisions The Ultimatum Game reveals important insights about human behavior, particularly the role of emotions, fairness, and social norms in nancial decisions These concepts are essential in neuronance, where the study of how the brain aects nancial decision-making is key to understanding investment behavior, negotiations, and risk tolerance. The emotional drive for fairness can inuence investor behavior and market sentiment. For example, when investors believe a company is treating its shareholders or customers unfairly (such as through poor corporate governance or excessive executive pay), they may divest or protest, even if the company is nancially sound. Faiza Majid FINS3655 Behavioral Finance Moral Reasoning Faiza Majid FINS3655 Behavioral Finance Moral Reasoning Thomsons Trolley Dilemma – used as a test of moral reasoning Suppose you are the switchperson on a railroad. A train is approaching 5 workers on the track and will kill them. You can throw the switch and divert the train to a side track, where unfortunately, there is a single worker. Do you throw the switch? A train is approaching 5 workers on the track and will kill them. You are on a bridge over the track next to an enormously large person. If you push the person o and onto the track, his weight will stop the train. (You yourself are too small to stop the train.) Do you push him o? Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Moral Reasoning Even though the two choices amount to a utilitarian judgment between ve lives versus one, most people will throw the switch in the 1st case but not in the 2nd one. Even the ones that are willing to push the man o in the 2nd case only do so after great deliberation and reluctance. Several dierent scenarios of this problem, all lead to a basic result - people care about the means to the end. In the rst case, the death of 1 to save 5 was incidental. In the 2nd case, the death of 1 to save 5 was instrumental. People are extra-averse to using others without their consent. How can neuroscience shed light on this? Faiza Majid FINS3655 Behavioral Finance Moral Reasoning- Greene et al., 2001 Presented subjects in an MRI with 60 hypothetical questions, of 3 categories: Up close and personal moral decisions - pushing the person in the Trolley case, harvesting one persons organs involuntarily to save 5 others, throwing people o an overcrowded lifeboat. Impersonal moral decisions - Original Trolley case, voting for policies that would cause deaths to save others. Non-moral decisions - whether to take the bus or the train given time constraints. Faiza Majid FINS3655 Behavioral Finance Moral Reasoning- Greene et al., 2001 Results: the personal moral problems were associated with a 20-40% increase in MRI signal in areas associated with emotion (medial frontal cortex) By contrast, the impersonal moral problems never experienced more than a 20% increase, most were less than 10%. Areas associated with cognitive processing increased 15-25% for the impersonal moral problems (and identically for the non-moral control problems) but were lower or even declining for the personal problems. ⇒ Problems involving direct injury to others invoke automatic emotional responses, which are powerful enough to over-ride previous cognitive principles. Faiza Majid FINS3655 Behavioral Finance Moral preferences in Financing Decisions Moral Dilemmas in Financial Decisions Corporate Governance: moral dilemmas where leaders must decide between actions that benet the company at the expense of individual employees or stakeholders. For instance, laying o a small percentage of the workforce might improve the companys overall health, but it harms the individuals aected. Investor preferences: Investors increasingly favor ESG. Or companies that prioritize ethical practices (e.g., fair labor conditions or environmental responsibility) may forego some prot. Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice Faiza Majid FINS3655 Behavioral Finance Standard Intertemporal Choice Many interesting questions in economics involve decisions with outcomes over time. How should people allocate their wealth between current consumption and future consumption? How do people trade o short term benets and long term costs? Or short term costs and long term benets? Faiza Majid FINS3655 Behavioral Finance Interest rate Suppose you put $1000 in the bank that pays 10% interest per year. After 1 year, youll have 1000 * 1.10 =1100 After 2 years, youll have 1100 * 1.10=1210 After 3 years, youll have 1210*1.10 = 1331 In general, if you put amount $A at interest rate r for T years, its future value in T years will be: A ∗ (1 + r )T Faiza Majid FINS3655 Behavioral Finance Interest rate Suppose you will be paid $1100 one year from today. If the market interest rate is 10%, how much is this future payment worth now? Answer: It must be $1000, since $1000 today at the 10% interest rate would yield $1100 in a year. Given per-period interest rate r, the net present value or present A discounted value of future amount A is equal to: (1+r )T Faiza Majid FINS3655 Behavioral Finance Standard Intertemporal Choice When modeling intertemporal choice, economists treat one physical good consumed at two dierent times as two dierent goods. Suppose there are two goods, x and y, to be consumed at time periods 1 and 2. U (x1 , y1 , x2 , y2 ) = u(x1 , y1 ) + D(t) u(x2 , y2 ) D(t) < 1 is the discount function that captures the relative preference for consumption in period 1 versus period 2. Faiza Majid FINS3655 Behavioral Finance Standard Intertemporal Choice How should an agent choose how much x to consume in each period? The most common discount function employed by economists is the exponential discount function: D(t) = δ t U(C0 ; C1 ; C2 ) = lnC0 + δlnC1 + δ 2 lnC2 Suppose the yearly δ = 0.75, implying that something occurring in one year is worth 3/4 as much as if it occurred today. Something happening in 3 years would thus be worth 0.753 = 0.42 as much as if it happened today Faiza Majid FINS3655 Behavioral Finance Standard Intertemporal Choice An important property of exponential discounting is Constant per period discounting : At all time periods, D(t+1) / D (t) = δ How an individual feels about this year versus next year is the same (δ /1) as how he feels about 5 years from now versus 6 years from now (δ 6 / δ 5 ) How he feels about tomorrow versus the day after tomorrow is that same as between 112 days from now and 113 days from now. Constant per-period discounting implies time consistency - the property whereby the optimal decision at one point in time remains the optimal decision with the passage of time. Faiza Majid FINS3655 Behavioral Finance Time Preferences Faiza Majid FINS3655 Behavioral Finance Time Inconsistency Which would you prefer? A: $2000 right now B: $2400 in a year from now Which would you prefer? A: $2000 in 10 years B: $2400 in 11 years Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Time Inconsistency Time Inconsistency or Dynamic Inconsistency - When the optimal decision at one point in time is no longer the optimal choice at another point in time That is, people experience preference reversals When both outcomes were in the future, prefer the more delayed $2400 But when the choices were more immediate, ipped preferences and prefer the immediate $2000 Faiza Majid FINS3655 Behavioral Finance Time Inconsistency Whats going on? Why are preferences dynamically inconsistent? Consumers behave impatiently today but prefer/plan to act patiently in the future E.g.: Tempted to choose immediate $10 rather than $11 tomorrow but prefer $11 in a year and a day to $10 in a year  Dynamically inconsistent. How can neuroscience shed light on this? Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice Because of the way our brain is wired: The reward system distinguishes sharply between present rewards and future and overweights the present (impulsive self- short-run) The PFC/parietal cortex discounts exponentially (rational self- long-run) Quasi-hyperbolic discounting function captures this (Laibson, 1997): U(C0 ; C1 ; C2 ) = lnC0 + β(δlnC1 + δ 2 lnC2 ) where 0 < β ≤ 1 Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice McClure et al (2004) - Subjects given a variety of choices between a smaller, sooner (SS) and a larger, later (LL) monetary reward. In some decisions, both the SS and the LL are delayed In other decisions, the SS is available immediately Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice Results: When both choices are delayed, only the prefrontal cortex (PFC) activates. However, when one of the choices is immediate and one is delayed, both the PFC and the limbic system activate. These regions receive dopamine transmissions and have been implicated in impulsive behavior. Those with relatively greater activation in the limbic areas were more likely to take the immediate reward, and those with relatively greater activation in the PFC took the delayed reward. Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice Strong evidence for being of two minds about intertemporal choice. The limbic brain is myopic and cant see the future. The pre-frontal cortex might discount all periods identically (exponentially and time-consistently), but the limbic system is responsible for the present bias. Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice- Other Anomalies 1) The Sign Eect: Which would you prefer: A: You receive $200 right now B: You receive $300 in 4 years Which would you prefer: A: You pay $200 right now B: You pay $300 in 4 years People tend to discount future gains more heavily than future losses; that is, people are more patient for losses than they are for gains. Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice- The Sign Eect The sign eect shows that people are inconsistent in how they treat future gains and losses. This asymmetry aects nancial decisions, such as savings behavior, investment planning, and debt repayment, where people may prefer immediate rewards but delay payments or losses, even if it results in larger future costs. Real-life examples? Credit Card Payments (Loss Scenario) Loan Repayments (Loss Scenario) Saving for Retirement (Gain Scenario) Lottery Winnings (Gain Scenario) Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice- Other Anomalies 2) Magnitude Eect: Which would you prefer: A: $10 right now B: $20 in 3 years Which would you prefer: A: $1,000,000 right now B: $2,000,000 in 3 years Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Intertemporal Choice- Magnitude Eect The magnitude eect in intertemporal choices refers to the observation that people tend to discount smaller amounts of money more steeply than larger amounts. People are more willing to take $10 over $20 in 3 years but are willing to wait for the $2,000,000 in 3 years over the immediate $1,000,000 People tend to be more patient as the amounts at stake rise Faiza Majid FINS3655 Behavioral Finance  Risk Taking Faiza Majid FINS3655 Behavioral Finance Neural Substrates of Risk Attitudes: Theory Higher baseline dopamine (for genetic reasons) and/or higher phasic dopamine (in response to environmental factors)  Stronger anticipatory reward signals  Higher risk taking Higher serotonin in response to risk  Stronger fear signals  Lower risk taking attitude. The combination of the two dimensions determines ones risk. Faiza Majid FINS3655 Behavioral Finance Neural Substrates of Risk Attitudes: Evidence Anticipatory activation of the nucleus accumbens precedes risk seeking mistakes in an investment task (Kuhnen & Knutson, 2005)  Reward system activated by DA release when gain anticipated Anticipatory activation of the fear system (anterior insula) precedes risk aversion mistakes in the same study. Faiza Majid FINS3655 Behavioral Finance Neural Substrates of Risk Attitudes: Evidence Testosterone [a steroid hormone] would promote risk taking by inuencing the reward system Increase in circulating testosterone would increase DA release in the nucleus accumbens. High freq traders morning testosterone predicts his days protability (Coates, 2008) High freq traders 2D:4D ratio predict protability (Coates & Rustichini, 2009) The lower the trader’s 2D:4D, the more money he made when increase in circulating testosterone Signicant relation between circulating testosterone and risk aversion in career choices (Sapienza ea, 2009) Women subjects were administered testosterone during play of the IGT: Played more with the risky decks after receiving testosterone (Van Honk ea, 2004) Faiza Majid FINS3655 Behavioral Finance Testosterone and Risk Taking Winner Eect: In games, it has been observed that  () testosterone after winning (losing). Since elevated testosterone  chances of winning, winning contributes a later win (Coates, 2010) Application to nancial markets 1 ↑ testosterone after a gain ⇒ ↑ risk taking ⇒ ↑ risk taking in bull markets (above average gains) 2 ↓ testosterone after a loss ⇒ ↑ risk aversion ⇒ ↑ risk aversion in bear markets back Faiza Majid FINS3655 Behavioral Finance Other Factors Inuencing Risk Taking Dispositional Factors Gender Age Genetics Situational Factors Vividness (Will see in the next lecture) Mood Life experiences Faiza Majid FINS3655 Behavioral Finance Dispositional Factors Determining Risk Taking: Individual Dierences Faiza Majid FINS3655 Behavioral Finance Gender and Risk Taking Women are more risk averse than men (e.g., Barber & Odean 2001, among many others) Hormonal argument: there must be testosterone-induced fundamental sex dierences in nancial risk-making More women = less testosterone = less risk-taking, so bring on the women (Coates, 2011) If Lehman brothers had been Lehman sisters, run by women instead of men, would the credit crunch have happened? (The Guardian, 2009) Faiza Majid FINS3655 Behavioral Finance Gender and Risk Taking Risk taking is NOT a 1-dimensional personality trait: there are insurance-buying gamblers and skydiving wallowers (Slovic, 1964) Taking physical risks vs. taking nancial risks vs. taking social risks Risk taking is highly domain-specic (Weber at al 2002) and the evidence suggests women are actually more risk takers than men for social risks (Byrnes et al 1999).... What about for nancial risks? Faiza Majid FINS3655 Behavioral Finance Gender and Risk Taking Meta study collates lab data from 18 studies of sex dierences in nancial decision-making (Julie Nelson 2014, Journal of Economic Methodology 21(3), 211-231) Shows how stereotyping and conrmation bias have led researchers to publish only the studies that conrmed their stereotypes re-gender dierences in risk taking Absence of evidence is even more blatant after excluding all problematic data E.g., data obtained from teenage subjects (see in 2 slides why we need to exclude them), from self-reports (the believed appropriate response likely dier in men and women!), and from experiments that failed to provide signicant incentives and/or that used hypothetical questions rather than real decision-making. Once we control for wealth constraints and investment knowledge, no gender dierences in risk taking among managers and professional investors. (See Johnson Powell 1994 and Atkinson et al 2003) Faiza Majid FINS3655 Behavioral Finance Risk Taking Age and Risk Attitude Adolescents have a propensity to take excessive risks One root cause of this: lower self-control abilities (Walsh, 2005 and Ameriks et al., 2007) Genetics and Risk Attitude Individuals with higher baseline dopamine levels for genetic reasons are expected to be more risk-takers than the rest of the population Part of trading performance would thus be genetically based. There is some evidence for this hypothesis: see Sapra et al 2012 and ongoing research Faiza Majid FINS3655 Behavioral Finance Environmental (Situational) Factors Determining Risk Taking: Individual Dierences Faiza Majid FINS3655 Behavioral Finance Risk Taking Faiza Majid FINS3655 Behavioral Finance Ambiguity Aversion Faiza Majid FINS3655 Behavioral Finance The Known Unknowns In most experiments seen, decision maker knows the odds In the real world, the odds are unknown! Savage, 1964: agent has beliefs about the outcome probabilities Ambiguity vs Risk Ambiguity comes from not knowing the odds whereas risk comes from the odds themselves. Faiza Majid FINS3655 Behavioral Finance Risk vs Ambiguity Imagine an unbiased dice; decision-maker knows its unbiased. No ambiguity. But there is risk as next outcome is totally unpredictable (each outcome has probability 1/6) Now imagine a biased dice: number 1 occurs with probability 0.8, number 2 with probability 0.2; the other numbers never occur. Decision-maker knows the probabilities. No ambiguity. Risk is lower than in previous case as more likely that 1 is going to occur Faiza Majid FINS3655 Behavioral Finance Risk vs Ambiguity Now what about the same biased dice as in the second case, but this time the decision-maker is not told the probabilities; can learn about them through sampling one number every 10 sec. Decision-maker faces both ambiguity and risk Ambiguity: maximal at the beginning, then decreases as learning proceeds Level of risk perceived by the agent: depends on the agents beliefs Maximal at the beginning if the agent is agnostic about the ps, much reduced once learning has proceeded Faiza Majid FINS3655 Behavioral Finance Basic Ellsberg Setup An urn containing 90 balls 30 are known to be red Remaining 60 are either blue or yellow, but the proportion is unknown. The player is then asked to choose between various gambling options all dealing with the same jar Faiza Majid FINS3655 Behavioral Finance Faiza Majid FINS3655 Behavioral Finance Ellsberg: Choice I Option A: If a red ball comes up, you get $100 (otherwise nothing) Option B: If a blue ball comes up, you get $100 (otherwise nothing) Which would you bet on, red or blue? Or are both equally good options? Most people prefer to bet on red. Faiza Majid FINS3655 Behavioral Finance Ellsberg: Choice II Option C: If either a red or a yellow ball comes up , you get $100. Option D: If either a blue or a yellow ball comes up , you get $100. Which of these options would you choose? Or are both equally good options? Most people prefer option D. Faiza Majid FINS3655 Behavioral Finance Ellsberg Paradox THE PARADOX? The EU of betting A is greater than the EU of B i the EU of C is greater than the EU of D. Ambiguity aversion is a preference for known risks over unknown risks. How can neuroscience shed light on this? Faiza Majid FINS3655 Behavioral Finance Do People Care About Ambiguity? Ambiguity aversion is a preference for known risks over unknown risks. This behavior was rst introduced through the Ellsberg paradox (people prefer to bet on the outcome of an urn with 50 red and 50 blue balls rather than to bet on one with 100 total balls but for which the number of blue or red balls is unknown). How can neuroscience shed light on this? Faiza Majid FINS3655 Behavioral Finance Why Do People Care About Ambiguity? Fear system alerts us when information is missing Hsu et al., (2005) contrast ambiguous bets vs. risky bets Level of ambiguity correlates with activation of amygdala and OFC OFC-damaged subjects do not distinguish between the risky bet and the ambiguous bet For more evidence that the brain distinguishes between risk and ambiguity: see Huettel et al., (2006) and Payzan-LeNestour et al., (2013). Faiza Majid FINS3655 Behavioral Finance