A founder under acute pressure, mid-crisis, exhausted, approves a $40,000 software contract in a single call, something that would ordinarily take three weeks of comparison shopping. Later, calmer, they cannot fully explain why it felt urgent at the time. This is usually described, if it is described at all, as "not thinking straight." That phrase is closer to literally true than most people realize, and the actual mechanism is worth understanding in some detail, because the honest version of the science is more useful than the folk version.
Key Takeaway
Financial decision-making under stress is not simply "emotional" in a vague sense; it involves measurable changes in specific brain circuitry. The prefrontal cortex, which provides top-down regulation of impulses, is disproportionately vulnerable to disruption by stress-related neurochemicals, per Amy Arnsten's widely cited 2009 review. Purchase decisions themselves engage identifiable reward and aversion circuits, documented in Brian Knutson's 2007 neuroimaging work. And the stress hormone cortisol has been directly measured shifting financial risk-taking, including in a real London trading floor. The genuinely important nuance, often lost in popular treatment of this topic, is that acute and chronic stress push financial behaviour in different, sometimes opposite, directions, and a simple "stress makes founders reckless" narrative is not what the actual research shows.
What "Neurochemistry" Actually Explains
The rest of this series has largely operated at the level of cognitive bias, describing systematic patterns in how people reason, drawing on the Kahneman-Tversky tradition of behavioural economics. This article operates at a different, complementary level of explanation: not what pattern of reasoning occurs, but what is physically happening in the brain and endocrine system while it occurs. These levels are not competing explanations for the same behaviour; they are different resolutions of the same picture, in the way that "the car ran out of gas" and "the fuel injector stopped receiving fuel" both correctly explain why the engine stopped. Cognitive bias research describes the pattern; neuroscience research describes a mechanism that helps produce it.
The Reward Circuit: Wanting, Buying, And The Pain Of Paying
Knutson and colleagues' 2007 study in Neuron, "Neural Predictors of Purchases," used functional MRI to watch what happens in the brain during an actual simulated shopping task, viewing a product, seeing its price, then deciding whether to buy[1]. The findings mapped the purchase decision onto distinct neural circuits: viewing a desirable product activated the nucleus accumbens, a region central to anticipating reward, while seeing an excessive price activated the insula, a region associated with anticipating pain and loss, and deactivated the medial prefrontal cortex, involved in evaluating whether the price was a good deal. Critically, activity in these regions predicted the actual subsequent purchase decision above and beyond what participants said about their own preferences[1].
This gives concrete neural content to the popular phrase "the pain of paying": parting with money genuinely engages circuitry associated with physical and emotional pain, not merely a metaphorical sense of loss. It also implies something important for the rest of this article: a purchase decision is, at the neural level, a real-time balance between an anticipatory reward signal (wanting the thing) and an aversive pain signal (the cost of getting it), and anything that either amplifies the reward signal or dampens the pain signal will shift the balance toward buying, independent of whether the purchase is actually a good idea.
Cortisol And The Stress Response
Cortisol is the primary hormone released by the hypothalamic-pituitary-adrenal (HPA) axis in response to a perceived stressor, physical or psychological. It mobilizes energy, sharpens certain kinds of attention in the short term, and, centrally for this discussion, has documented effects on cognition and risk-related decision-making that differ depending on how much cortisol is present and for how long. This last point, the acute-versus-chronic distinction, is the single most important nuance in this entire article and is addressed in detail below.
Evidence From A Trading Floor
The most direct field evidence linking cortisol to real financial decision-making comes from Coates and Herbert's 2008 study, published in the Proceedings of the National Academy of Sciences, which sampled endogenous steroid hormones from male traders on a London trading floor under actual working conditions[2]. The study found that a trader's cortisol level rose in step with both the variance of their own trading results and the volatility of the broader market, and that morning testosterone levels predicted that day's profitability. The authors suggested that if these acute hormonal shifts persisted or intensified as volatility rose, they could plausibly shift a trader's risk preferences and behaviour, potentially contributing to the kind of self-reinforcing market instability separate research has since investigated directly[2].
This is not a study of business owners specifically, and generalizing from professional traders to founders should be done cautiously, a point returned to in the limits section. It is, nonetheless, among the cleanest available demonstrations that real-world financial stress produces measurable endocrine changes in real time, under real stakes, not merely in a laboratory simulation.
Acute And Chronic Stress Do Different Things
Here the story gets genuinely more complicated than most popular treatments of "stress and money" acknowledge, and getting this right is the difference between an accurate account and an oversimplified one. Kandasamy and colleagues' 2014 follow-up study, also published in PNAS, directly tested what happens to financial risk-taking when cortisol is experimentally elevated over a sustained period, administering individually tailored hydrocortisone doses to volunteers over eight days to replicate the cortisol levels previously observed in real stressed traders[3]. The finding was not increased recklessness. Participants under sustained elevated cortisol became more risk-averse, and their tendency to overweight small probabilities, a known distortion in how people process unlikely outcomes, became more pronounced[3].
A separate review by Coates, Gurnell and Rustichini synthesizing this line of research put the distinction plainly: acutely elevated steroid hormones may optimize performance on a range of tasks in the short term, while chronically elevated hormones tend to promote irrational risk-reward choices, and the direction of that irrationality is not simply "more risk-taking" across the board[4]. In short, cortisol codes for risk broadly, and its behavioural effect depends heavily on duration and context, not on a single uniform "stress equals recklessness" rule.
The Prefrontal Cortex Goes Offline
A separate, equally well-established line of research explains how stress can produce impulsive, poorly evaluated decisions specifically, even though the cortisol literature above shows chronic stress increasing caution on straightforward risk questions. Amy Arnsten's widely cited 2009 review in Nature Reviews Neuroscience synthesized decades of work on how stress affects the prefrontal cortex (PFC), the brain region responsible for "top-down" regulation of attention, impulse control and working memory[5]. The central finding is stark: even mild, acute, uncontrollable stress can cause a rapid, significant loss of prefrontal cognitive function in both animal models and humans, mediated by stress-related increases in noradrenaline and dopamine in the PFC, and more prolonged stress exposure produces actual structural changes in prefrontal dendrites[5].
The practical significance is specific: the PFC is precisely the region responsible for holding a financial goal in mind, weighing a purchase against that goal, and inhibiting an impulsive action that conflicts with it. When acute stress measurably degrades PFC function, the brain's capacity to perform exactly that evaluation is impaired in real time, independent of whether the person's general risk appetite has gone up or down.
Reconciling The Contradiction
Put the last two sections together and an apparent contradiction emerges: chronic cortisol elevation makes people more cautious on structured risk questions, while stress broadly impairs the prefrontal function needed for careful evaluation. Both findings are well supported, and the honest reconciliation is that they describe different things. The cortisol-and-caution research primarily involves structured decisions under known probabilities, choosing between defined gambles, where a person has time and information to weigh options. The prefrontal-impairment research concerns the capacity for deliberate, effortful evaluation in the moment, including the ability to stop and think before acting at all.
A plausible, and it should be stated clearly as plausible rather than directly proven, reading consistent with both bodies of evidence is this: chronic, grinding financial stress may make a founder generally more risk-averse and defensive about large, deliberate financial commitments, hoarding cash, avoiding investment, exactly the caution the cortisol research would predict, while an acute stress spike, a crisis phone call, a customer ultimatum, a sudden emergency, can simultaneously impair the prefrontal capacity needed to properly evaluate a specific in-the-moment decision, producing an impulsive purchase that looks like recklessness but is better understood as a temporary failure of evaluative capacity rather than a genuine shift toward risk-seeking. Both processes can be operating in the same exhausted founder, on different timescales, which is part of why the folk explanation of "stress makes people reckless" captures something real without capturing it accurately.
Why "Retail Therapy" Is Not Just A Phrase
Given the reward-circuitry findings described above, the folk concept of retail therapy has a genuine mechanistic basis worth taking seriously rather than dismissing as a cliché. Loewenstein's influential 1996 paper on "visceral factors," states of intense feeling such as stress, craving, or emotional pain, argued that these states can systematically override deliberative decision-making, producing behaviour a person would not endorse in a calmer moment and often would not have predicted about themselves in advance[6]. A purchase, per the Knutson findings above, reliably activates anticipatory reward circuitry associated with the nucleus accumbens before the transaction even completes. Under acute stress, with prefrontal evaluative capacity already compromised per Arnsten's findings, that reward-anticipation signal has comparatively less top-down competition, making the anticipated relief of "wanting" a purchase more likely to translate directly into "buying" it, with markedly less of the price-evaluation friction the insula and medial prefrontal cortex would ordinarily provide.
The Decision Fatigue Question, Handled Carefully
No honest treatment of this topic can skip the popular concept of "decision fatigue," or ego depletion, the idea that self-control is a limited resource that becomes exhausted through use over the course of a day, leaving less capacity to resist a poorly-judged purchase later on. The idea, formalized in Baumeister's original resource model of self-control, was influential and widely popularized. It has also, in the years since, run into a serious and largely unresolved replication problem that any accurate account of this territory must disclose.
A large-scale, pre-registered multi-lab replication effort led by Hagger and colleagues, published in 2016 in Perspectives on Psychological Science, involved 23 independent laboratories and 2,141 participants attempting to replicate a standard ego-depletion effect[7]. The result was, in the authors' own words, consistent with a null effect: no reliable evidence that performing an initial self-control task reduced performance on a subsequent one[7]. Notably, 22 of the 23 participating labs had predicted beforehand that they would successfully replicate the effect, underscoring how unexpected the null result was even to researchers sympathetic to the theory. The debate has continued since, with some researchers defending the underlying construct while questioning the specific experimental paradigm used, and others treating the null replication as strong evidence against the classic version of the theory[7].
Given this genuinely unsettled state, this article treats "decision fatigue" as a plausible, popularly resonant hypothesis rather than an established finding, and declines to lean on it as a load-bearing explanation for founder overspending. The mechanisms with stronger evidentiary support, prefrontal impairment under acute stress and reward-circuit activation during anticipated purchases, do not require the specific ego-depletion resource model to be true, and this article's practical recommendations below do not depend on it.
Founder-Specific Amplifiers
Several features of running a business intensify the mechanisms above relative to a typical employee's experience, worth naming explicitly. Founders frequently face genuinely chronic, low-grade financial stress, irregular income, personal guarantees, the entangled personal-and-business financial identity discussed elsewhere in this series, layered with intermittent acute spikes, a client emergency, a cash crunch, a key employee's resignation, rather than either type of stress in isolation. This is precisely the combination the reconciliation above suggests is most behaviourally complex: a chronically elevated baseline that may be producing generalized defensiveness, punctuated by acute spikes that can specifically impair in-the-moment evaluation, potentially producing the seemingly paradoxical pattern of an owner who is simultaneously too cautious about deliberate, planned investment and too impulsive about urgent, in-the-moment spending.
A Worked Case: The $40,000 Software Purchase
Return to the opening scenario. A founder mid-crisis, having just lost a key operational staff member with no notice, receives a same-day sales call from a vendor offering a workflow automation platform positioned as solving exactly the gap the departure created. The founder, per the mechanisms above, is plausibly operating with acutely elevated stress hormones and correspondingly reduced prefrontal evaluative capacity at the precise moment the sales call arrives, while the platform itself, framed as an immediate solution to an acute, painful problem, strongly engages anticipatory reward circuitry. The insula's ordinary "pain of paying" response to the $40,000 price tag, per the Knutson findings, has less prefrontal moderation available to weigh it against calmer, structured evaluation. The contract gets signed same-day.
Six weeks later, calmer, the founder discovers the platform requires substantially more implementation work than represented, addresses only part of the original gap, and a $12,000 alternative solution, found in twenty minutes of comparison shopping once actually undertaken, would likely have served better. Nothing about this founder's general competence or judgment was different between the two moments. What was different, per the research above, was measurable and mechanistic: an acute stress state at the moment of decision, and a calmer state at the moment of evaluation, engaging meaningfully different neural resources for what was framed as, and felt like, the same kind of decision.
What Actually Works, And Why
The practical implications follow directly from the mechanism, not merely from generic advice to "calm down before deciding," though that folk wisdom turns out to be more mechanistically grounded than it sounds.
Build in a mandatory delay for spending above a threshold. Because acute stress specifically impairs the in-the-moment evaluative capacity that a delay allows to recover, even a short, fixed cooling-off period before any purchase above a set dollar amount, structurally separate from the emotionally loaded moment of the sales conversation, gives the prefrontal cortex time to re-engage before commitment.
Pre-commit spending thresholds and approval processes while calm. This is the same structural principle recommended throughout this series for other biases, and the neuroscience gives it a more literal justification here: a policy set during a low-stress period is a decision made with full prefrontal capacity available, designed specifically to bind a future self whose prefrontal capacity cannot be assumed to be intact at the moment of temptation.
Separate the crisis from the purchasing decision. Where possible, structurally decouple "we have an urgent operational problem" from "we are deciding how to solve it right now, today, with this specific vendor." Even a 24-to-48-hour separation between identifying the problem and committing capital to a specific solution allows the acute stress response, which is time-limited by nature, to subside somewhat before the evaluative decision is made.
Treat chronic stress and acute stress as different problems requiring different responses. Given the genuine divergence in their effects, addressing a founder's baseline chronic stress, sleep, workload, financial runway visibility, is a different intervention from managing a single acute crisis moment, and conflating the two, treating every stressed decision as the same problem, likely produces the wrong fix for at least one of them.
A Note On The Testosterone Findings
The Coates and Herbert trading-floor study is sometimes summarized in popular coverage as showing that testosterone drives reckless risk-taking, which overstates what the data actually showed. The study found that higher morning testosterone predicted a trader's profitability that day, and separately identified a "winner effect" in which a run of successful trades was associated with rising testosterone over consecutive days, a pattern also documented in animal research on competitive success[2]. The relationship the study actually documented is between testosterone and successful risk-taking specifically, not between testosterone and risk-taking that happened to be reckless. Whether sustained elevated testosterone eventually tips toward overconfidence and poorly calibrated risk-taking, a plausible extension the original authors themselves raised as a possibility rather than a finding, is a separate, less directly evidenced claim worth distinguishing from what the study established.
What This Does Not Excuse
Understanding the neurochemical mechanism behind a poor financial decision explains it; it does not excuse it, and treating neuroscience as an exculpatory narrative, "my cortisol made me do it," would misuse this research in a way this article does not endorse. The value of understanding the mechanism is entirely forward-looking: it identifies specifically which structural safeguards, delay, pre-commitment, decoupling of crisis from purchase, address the actual vulnerability, rather than relying on the much less reliable strategy of simply resolving to have better judgment under conditions the research shows measurably degrade judgment. Accountability for outcomes and mechanistic understanding of causes are not in tension; the second is what makes genuinely effective prevention of the first possible.
The Limits Of This Analysis
Several caveats matter. The Coates and Herbert trading-floor study involved a small sample of male traders in a specific professional context, and generalizing its specific findings to business owners broadly, particularly across genders, industries and stress profiles, is an inference rather than a directly established result; subsequent research, including Kandasamy et al. (2014), has found gender differences in stress-related risk-taking responses that this article has not fully explored. The prefrontal cortex research, while extremely well replicated across both animal and human studies, describes general mechanisms of stress and cognition rather than financial decision-making specifically, and the extension to purchase decisions, while reasonable given the Knutson findings' independent confirmation of financial-decision neural circuitry, connects two separate literatures rather than reporting a single unified study of stressed founders making purchases.
Finally, and most importantly, this article has been explicit throughout about genuine unresolved tensions in the underlying science, particularly the acute-versus-chronic distinction and the contested state of decision-fatigue research, rather than smoothing them into a simpler, more marketable narrative. The honest state of this field is that stress reliably changes financial decision-making in measurable, mechanistic ways; it does not reliably change it in one simple, uniform direction.
Frequently Asked Questions
Does stress always make people take bigger financial risks?
What actually happens in the brain during a purchase decision?
Is "decision fatigue" a real, scientifically established phenomenon?
Can understanding this neuroscience actually change how I run my business?
Does this research excuse poor financial decisions made under stress?
References
- Knutson, B., Rick, S., Wimmer, G. E., Prelec, D., & Loewenstein, G. (2007). Neural Predictors of Purchases. Neuron, 53(1), 147-156.
- Coates, J. M., & Herbert, J. (2008). Endogenous Steroids and Financial Risk Taking on a London Trading Floor. Proceedings of the National Academy of Sciences, 105(16), 6167-6172.
- Kandasamy, N., Hardy, B., Page, L., Schaffner, M., Graggaber, J., Powlson, A. S., Fletcher, P. C., Gurnell, M., & Coates, J. (2014). Cortisol Shifts Financial Risk Preferences. Proceedings of the National Academy of Sciences, 111(9), 3608-3613.
- Coates, J. M., Gurnell, M., & Rustichini, A. (2010). From Molecule to Market: Steroid Hormones and Financial Risk-Taking. Philosophical Transactions of the Royal Society B: Biological Sciences, 365(1538), 331-343.
- Arnsten, A. F. T. (2009). Stress Signalling Pathways That Impair Prefrontal Cortex Structure and Function. Nature Reviews Neuroscience, 10(6), 410-422.
- Loewenstein, G. (1996). Out of Control: Visceral Influences on Behavior. Organizational Behavior and Human Decision Processes, 65(3), 272-292.
- Hagger, M. S., Chatzisarantis, N. L. D., et al. (2016). A Multilab Preregistered Replication of the Ego-Depletion Effect. Perspectives on Psychological Science, 11(4), 546-573.
This article discusses peer-reviewed neuroscience and behavioural research and is provided for general informational and educational purposes. It is not medical, psychological, or financial advice, and does not diagnose any condition. If stress is significantly affecting your health or decision-making, consider speaking with a qualified medical or mental health professional in addition to a financial advisor.