Inside the mind
of speculation.

PERCEPTION

Every decision
begins with
a signal.

Before a trade comes an interpretation: a price move, a headline, a message from someone we trust.

We ask which signals people notice in speculative markets, how they read them, and what they leave out.

ATTENTION

Attention comes
before choice.

When a stock suddenly trades far above its usual volume, individual investors turn into net buyers. At one large US discount broker, clients were net sellers of the quietest stocks and net buyers of the busiest.

Buying means choosing from thousands of options; selling, only from what one already owns. What catches the eye gets bought.

Barber, B. M. & Odean, T. (2008). All That Glitters. Review of Financial Studies, 21(2), 785–818.Barber & Odean (2008), Rev. Financ. Stud. 21(2). doi:10.1093/rfs/hhm079

Scope: US equities, 1991–1996, three brokerage datasets. Not crypto markets.

Individual investors buy what grabs attention
Buy-sell imbalance of individual investors by abnormal trading volume-20%-10%0+10%+20%+30%Volume decile 1: -18.15% (net selling)Volume decile 2: -8.90% (net selling)Volume decile 3: -6.23% (net selling)Volume decile 4: -2.76% (net selling)Volume decile 5: -0.76% (net selling)Volume decile 6: +1.65% (net buying)Volume decile 7: +5.45% (net buying)Volume decile 8: +9.20% (net buying)Volume decile 9: +13.62% (net buying)Volume decile 10a: +17.72% (net buying)Volume decile 10b: +29.50% (net buying)+29.5%−18.2%Net buyingNet selling← Quietest stocksBusiest stocks →Buy-sell imbalance of individual investors by abnormal trading volume-20%-10%0+10%+20%+30%Volume decile 1: -18.15% (net selling)Volume decile 2: -8.90% (net selling)Volume decile 3: -6.23% (net selling)Volume decile 4: -2.76% (net selling)Volume decile 5: -0.76% (net selling)Volume decile 6: +1.65% (net buying)Volume decile 7: +5.45% (net buying)Volume decile 8: +9.20% (net buying)Volume decile 9: +13.62% (net buying)Volume decile 10a: +17.72% (net buying)Volume decile 10b: +29.50% (net buying)+29.5%−18.2%Net buyingNet selling← Quietest stocksBusiest stocks →
Buy-sell imbalance (%) by abnormal volume decile, large discount brokerage
DecileImbalance
1-18.15%
2-8.90%
3-6.23%
4-2.76%
5-0.76%
6+1.65%
7+5.45%
8+9.20%
9+13.62%
10a+17.72%
10b+29.50%
Buy-sell imbalance of individual investors (buys minus sells, as a share of trades) by the day’s abnormal trading volume. Redrawn from Table 1.

COLLECTIVE BEHAVIOR

Individual minds.
Collective motion.

On-chain markets move through shared stories, visible rankings and rapidly changing expectations.

We ask how individual choices add up to collective conviction, and how much of the outcome is quality and how much is momentum.

SOCIAL INFLUENCE

Seeing others
changes what wins.

14,341 people chose among 48 songs by unknown bands. When they could see what others had downloaded, success became far more unequal than when everyone chose alone, and more so when songs were ranked by popularity.

Across eight parallel worlds that started from identical conditions, the same songs reached different levels of success.

Salganik, M. J., Dodds, P. S. & Watts, D. J. (2006). Experimental Study of Inequality and Unpredictability in an Artificial Cultural Market. Science, 311, 854–856.Salganik et al. (2006), Science 311. doi:10.1126/science.1121066

Scope: online experiment, mostly teenage participants, music downloads.

Visible popularity makes success more unequal
Inequality of success with and without social influence in two experimentsEach of 8 worlds with social signalChoosing alone0.10.20.30.40.50.6Inequality of song success (Gini) — higher is more unequalExperiment 1Download counts shownExperiment 1, world 1: Gini ≈ 0.33Experiment 1, world 2: Gini ≈ 0.38Experiment 1, world 3: Gini ≈ 0.42Experiment 1, world 4: Gini ≈ 0.33Experiment 1, world 5: Gini ≈ 0.38Experiment 1, world 6: Gini ≈ 0.28Experiment 1, world 7: Gini ≈ 0.30Experiment 1, world 8: Gini ≈ 0.30Experiment 1, independent world: Gini ≈ 0.260.260.28–0.42Experiment 2Songs ranked by popularityExperiment 2, world 1: Gini ≈ 0.46Experiment 2, world 2: Gini ≈ 0.51Experiment 2, world 3: Gini ≈ 0.53Experiment 2, world 4: Gini ≈ 0.45Experiment 2, world 5: Gini ≈ 0.50Experiment 2, world 6: Gini ≈ 0.50Experiment 2, world 7: Gini ≈ 0.48Experiment 2, world 8: Gini ≈ 0.56Experiment 2, independent world: Gini ≈ 0.200.200.45–0.56Inequality of success with and without social influence in two experimentsEach of 8 worlds with social signalChoosing alone0.10.20.30.40.50.6Inequality of success (Gini)Experiment 1Download counts shownExperiment 1, world 1: Gini ≈ 0.33Experiment 1, world 2: Gini ≈ 0.38Experiment 1, world 3: Gini ≈ 0.42Experiment 1, world 4: Gini ≈ 0.33Experiment 1, world 5: Gini ≈ 0.38Experiment 1, world 6: Gini ≈ 0.28Experiment 1, world 7: Gini ≈ 0.30Experiment 1, world 8: Gini ≈ 0.30Experiment 1, independent world: Gini ≈ 0.260.260.28–0.42Experiment 2Songs ranked by popularityExperiment 2, world 1: Gini ≈ 0.46Experiment 2, world 2: Gini ≈ 0.51Experiment 2, world 3: Gini ≈ 0.53Experiment 2, world 4: Gini ≈ 0.45Experiment 2, world 5: Gini ≈ 0.50Experiment 2, world 6: Gini ≈ 0.50Experiment 2, world 7: Gini ≈ 0.48Experiment 2, world 8: Gini ≈ 0.56Experiment 2, independent world: Gini ≈ 0.200.200.45–0.56
Gini coefficient per world (approximate, measured from the published figure)
ExperimentSocial-influence worldsIndependent world
Experiment 1 (Download counts shown)0.33, 0.38, 0.42, 0.33, 0.38, 0.28, 0.30, 0.300.26
Experiment 2 (Songs ranked by popularity)0.46, 0.51, 0.53, 0.45, 0.50, 0.50, 0.48, 0.560.20
Inequality of success (Gini coefficient) in eight social-influence worlds and one independent world, two experiments. Approximate values measured from the published Fig. 1.

REWARD & UNCERTAINTY

The space
before a choice.

Anticipation, hesitation, and an outcome that has not happened yet.

We ask how expectation and uncertainty shape trading decisions. We do not measure brain activity and make no claims about it.

PREDICTION ERROR

The signal is
the surprise.

In classic recordings from monkeys, dopamine neurons fired for an unexpected reward, stopped reacting once the reward was predicted, and dipped below baseline when a predicted reward failed to arrive.

The response follows the gap between expectation and outcome, not the reward itself.

Schultz, W., Dayan, P. & Montague, P. R. (1997). A Neural Substrate of Prediction and Reward. Science, 275, 1593–1599.Schultz et al. (1997), Science 275. doi:10.1126/science.275.5306.1593

Scope: single-neuron recordings in monkeys. Not a study of people or markets.

Dopamine neurons respond to surprise, not reward
Schematic of dopamine neuron responses to unexpected, predicted and withheld rewardsCueRewardUnexpected rewardburst at rewardPredicted rewardburst moves to the cuePredicted reward withhelddip when it was dueSchematic of dopamine neuron responses to unexpected, predicted and withheld rewardsCueRewardUnexpected rewardPredicted rewardPredicted reward withheld
Schematic of dopamine-neuron responses in three conditions, after the published Fig. 1. Shows the pattern only; no measured firing rates.

ABOUT THE PROJECT

Markets move.
Minds react.

Solimbic is an independent, early-stage research project. It starts from established findings in behavioral finance and neuroscience and asks how far they hold in speculative, on-chain markets.

It is not a clinical study, a medical product or an academic institution, and nothing here is investment advice. Sources and methods

OVERCONFIDENCE

More trading,
less return.

Across 66,465 US households, those who traded most earned 11.4% a year after costs, while the market returned 17.9%. Before costs, frequent and infrequent traders performed about the same.

The gap came from trading itself: commissions and spreads on every move.

Barber, B. M. & Odean, T. (2000). Trading Is Hazardous to Your Wealth. Journal of Finance, 55(2), 773–806.Barber & Odean (2000), J. Finance 55(2). doi:10.1111/0022-1082.00226

Scope: one US discount broker, 1991–1996; trading costs were far higher than today.

The most active traders earned the least
Average annual return after trading costs, US households 1991 to 19960%5%10%15%20%Households that trade leastHouseholds that trade least: 18.5% a year after costs18.5%Market indexMarket index: 17.9% a year after costs17.9%Households that trade mostHouseholds that trade most: 11.4% a year after costs11.4%Average annual return after trading costs, US households 1991 to 19960%5%10%15%20%Households that trade leastHouseholds that trade least: 18.5% a year after costs18.5%Market indexMarket index: 17.9% a year after costs17.9%Households that trade mostHouseholds that trade most: 11.4% a year after costs11.4%
Annualized net return, 1991–1997
GroupReturn
Households that trade least18.5%
Market index17.9%
Households that trade most11.4%
Annualized net returns, February 1991 to January 1997, as reported by the authors.

SOLIMBIC · $PSYKE

The experiment
starts here.

$PSYKE is the project’s experimental token on Solana. It carries no promised returns, medical benefits or future utility.

Token address and official link not yet published.