Open any stock trading app and you’ll see a familiar layout: a prominent price chart tracking the ups and downs of an asset, buttons for placing orders, a list of your holdings, and somewhere, often tucked away, an order book showing current bids and offers. The assumption baked into most of these designs is that traders want to see where prices have been. But do they actually look at that chart when making decisions?
A new study published in the Journal of Behavioral and Experimental Finance put that question to the test by strapping eye-trackers to experimental traders and watching where their gaze landed. The answer, according to the researchers, was not the price chart. It was the order book.
The question behind the experiment
Matthias Herrmann-Romero of the University of Liechtenstein and colleagues at the University of St. Gallen, the University of Innsbruck, and MCI in Austria wanted to understand how traders allocate their limited attention among the many data sources thrown at them on a trading screen. Attention, they note, can matter in two different ways. Some information is consulted rarely because it barely changes (like an asset’s fundamental value), while other information demands constant monitoring because it moves in real time (like a price chart). Measuring what traders consider “important,” then, requires more than one lens.
The team decided to use two complementary approaches. First, eye-tracking technology to capture where traders’ eyes actually rested. Second, a version of the experiment where traders had to actively choose which information elements to display, revealing what they deemed indispensable even if they didn’t stare at it constantly.
Designing the trading lab
The researchers recruited 151 university students and ran them through 16 laboratory sessions between April and November 2022. Participants traded a risky asset in a simulated market while a Tobii eye-tracker recorded their gaze at 70 times per second. A chin rest kept their heads still for accurate measurement.
The trading interface offered seven elements: a central order submission field for placing trades, plus six information boxes showing the order book (all outstanding bids and offers from other traders), a price chart, an inventory of the trader’s own holdings, a signals field with information about the asset’s value, a time clock, and the trader’s own order history. The researchers kept the visual design deliberately plain, using gray backgrounds and standardized box sizes, and randomized where each element appeared on the screen to rule out position effects.
Participants were split between two conditions. In the “complete” condition, all six information elements were visible at once. In the “limited” condition, traders could select only three of the six at the start of each period. The researchers also ran two different market setups: one prone to price bubbles (based on a classic design by Smith, Suchanek, and Williams) and one focused on aggregating private information (based on Plott and Sunder). Before trading, participants completed tests of cognitive reflection, fluid intelligence, financial literacy, and social preferences.
Where traders actually looked
After the order submission field itself, which absorbed about 45 to 55 percent of fixations for reasons of central placement and operational necessity, the order book dominated visual attention. It pulled in between 23 and 32 percent of all fixations across every market design and interface tested.
The price chart, by contrast, received less than 5 percent of visual attention, even in the bubble-prone market where one might expect traders to scrutinize past prices for patterns. It was rarely selected by traders who had to pick their three preferred elements, and it almost never appeared in the final fixations right before an order was placed. Out of nearly 1,000 observed sequences of the last three fixations before a trade in one condition, only 33 included the price chart.
The researchers interpret this as evidence that traders in these settings did not rely on historical price patterns. Instead, they focused on the liquidity and intentions revealed by the limit orders sitting in the order book, which offers forward-looking cues about where the market might move.
Looking versus choosing
The two measurement approaches mostly agreed with one another, but not perfectly. The inventory field is a telling exception. When given the chance to pick three elements, traders selected the inventory box more often than anything else. Yet it captured only about 9 percent of their visual attention during trading. The authors read this as a distinction between what traders consider essential to have available and what they actively process in the moment. You want to know your holdings are there, but you don’t need to stare at them.
The order book showed the opposite pattern: heavily fixated upon but selected slightly less often than the inventory. Put another way, eye-tracking and active selection each captured a different facet of importance.
Cognitive skills and attention
The study also looked at whether personal characteristics correlated with where traders focused. Participants who scored higher on cognitive reflection tests tended to spend more time looking at the order book and less on the order submission field, at least in certain conditions. In the information-aggregation market, traders with higher fluid intelligence showed a similar pattern, focusing more on the order book than on the narrower order submission field that only displays the best bid and ask. This echoes earlier findings that cognitively stronger traders are better at extracting signals from overall market activity.
The researchers found weaker and less consistent patterns for gender, financial literacy, and social preferences. Some correlations appeared in specific treatments but did not replicate across market designs or across the three ways of measuring fixations (count, duration, and frequency).
Attention and performance
Does where you look affect how much money you make? The evidence here was mixed. In the bubble market, traders who spent more time on the order book earned more, while those who fixated on the order submission field earned less. In the information-aggregation market, attention to the signals table (which carried private information about the asset’s value) correlated positively with performance, which makes intuitive sense given how the market was designed.
Traders also shifted their attention in predictable ways within each trading period. Early on, they looked more at static information like the signals table. As the period progressed, their gaze drifted toward dynamic elements like the order book and the time clock. The researchers read this as traders front-loading the consumption of information that doesn’t change, then watching real-time updates as the clock ran down.
Practical takeaways and caveats
The researchers suggest that trading platforms might reconsider their default layouts. Many experimental studies and real-world platforms devote substantial screen real estate to price charts, yet the traders in this experiment barely glanced at them. Making the order book more visually prominent, the authors argue, could help users focus on the information most linked to better outcomes.
Several caveats apply. The asset’s fundamental value was unusually transparent in both experimental designs, which may have reduced the usefulness of price history compared to real markets where the true value is genuinely uncertain. Real-world price charts also span much longer time horizons and aggregate far more trading activity. The participants were university students, not professional traders, and the laboratory setting differs from the complex, colorful interfaces of actual brokerages, where research has shown that design choices like color can sway behavior. The authors note that their order book served a dual purpose, acting as both an information display and a place where traders could cancel their own limit orders, which makes it hard to fully separate attention driven by information-seeking from attention driven by execution.