Trevor Grossman, PhD, Licensed Psychologist, CEREVITY
The claim you have heard is that tired traders take more risk. The evidence for it is a good deal weaker than the confidence with which it gets repeated, and what it gets wrong matters more than what it gets right, because it sends people looking for the wrong problem. Sleep debt does not mostly show up as appetite for risk. It shows up in the anxiety running underneath every high-stakes decision.
A 2025 review in Psychonomic Bulletin & Review pulled together 25 studies on sleep deprivation and risky decision making. Across 47 separate results, sleep deprivation increased risk-taking in 13, decreased it in 6, and produced no significant effect in 28.
Both halves of that deserve stating. Most results found nothing. Among the results that did reach significance, better than two to one pointed toward more risk-taking, so anyone claiming the effect is imaginary is also overreaching. The authors’ actual conclusion is that the effect is complex and depends on context, with sex, gain and loss framing, psychotropic drug use, the duration of deprivation and the choice of task all changing the direction.
Which is a genuinely unsatisfying answer, and it is the honest one. It also points somewhere more useful than the headline version, if you trade for a living. The thing that reliably breaks under accumulated sleep debt is not appetite for risk. It is the ability to follow your own rules.
The one study run on actual traders
Frank Song and Matthew Walker followed 17 full-time traders at a New York proprietary trading firm for 42 days, collecting daily surveys, and published the results in PLOS ONE in 2023.
Two findings are worth memorizing.
Each cup of caffeinated beverage cost about 10.4 minutes of sleep duration. Over a week that is more than an hour of sleep. Caffeine’s effect on how well the traders thought they had slept pointed in the expected direction but did not reach statistical significance, which in a sample this size means the subjective signal was too weak to detect rather than absent.
Each glass of alcohol predicted about a 3-point drop in subjective sleep quality on a 100-point scale. And when alcohol followed daytime caffeine, the evening alcohol partially offset caffeine’s effect on sleep duration.
That last one is the finding that should bother people. The two habits pull in opposite directions on different measures. Caffeine takes time off the clock while barely registering in how the night felt, alcohol measurably degrades the quality of what remains, and the combination can produce a night that feels unremarkable while two separate components of it are being eaten. If you are auditing your own sleep by how it felt, that is the least reliable instrument available.
Seventeen traders at one firm, all male, is a small sample and I am not going to pretend otherwise. Its value is that it followed real traders through real working weeks rather than testing students in a lab, and its findings sit comfortably inside a much larger literature on caffeine and alcohol generally.
What actually degrades
In clinical work with traders, the thing that erodes under accumulated sleep debt and chronic activation is almost never risk appetite in the abstract. It is adherence.
Nearly every trader I have worked with can state their rules. Position sizing, maximum daily loss, what happens after two consecutive losers. They can state them cleanly on a Sunday. The question is whether the rules survive Thursday afternoon of a bad week, and what determines that is not conviction. It is whether the nervous system running the account has any capacity left, and strain quietly worsens decisions long before anyone admits to it.
The specific failure has a shape, and it has a name that traders use before any clinician does. It is day three of a drawdown. The trader knows the rule says size down. What they feel is that sizing down guarantees the week ends red, and that ending the week red means something about them rather than about a distribution. So they size up to get it back today. Revenge trading is the term, it is usually described as a discipline failure, and they can tell you afterward that it was a bad decision, and they will do it again in six weeks under the same conditions.
It also shows up after a very good year, not only a bad one. A large realized gain rearranges a person’s relationship to risk, and what a windfall does to the person holding it turns up clinically far more often than the base rate of windfalls would suggest.
None of that is a discipline problem. It is what happens when a loss sequence, or a win large enough to change the stakes, gets processed as a verdict on the self, and when the capacity to sit with discomfort has been spent before the decision arrives. That pattern is not unique to trading. Trading just marks it to market daily, which at least makes it visible.
Three things that hold up under load
Separate the outcome from the read on yourself. This is the whole game. A trader who can experience a losing week as information rather than as evidence about their worth keeps access to their own process. One who cannot will override it, and the override always feels like conviction in the moment.
Treat the caffeine and alcohol routine as part of the process, not as personal life. The study measured roughly ten minutes of sleep per cup, on average, without breaking it out by time of day. If you are drinking several, that is a systematic input to your decision quality, and it belongs in whatever review you already run on your trades rather than in the category of things that happen outside work.
Get the review from someone with no position. Colleagues, risk managers and partners are all useful and all conflicted, because your state is information about their exposure. That is precisely why traders tend to underreport, and why a clinician who already works with this population is a different kind of conversation than the one on the desk.
The evidence does not say sleep loss makes you reckless. It says your judgment about your own state is the first thing to go, which is a worse problem, because it is the instrument you use to check everything else.
Trevor Grossman, PhD is a licensed psychologist in California, seeing clients by telehealth through CEREVITY, a nationwide network of independent licensed clinicians. He works with traders, executives and entrepreneurs on decision fatigue, high-stakes stress and performance under sustained load. This article is general information, not clinical advice and not investment advice.
