Paid market data can make investors overconfident, study finds

Investors may put too much trust in market data for one simple reason: they paid for it. A study by researchers in Germany and Canada found that people who bought trading information were more likely to treat it as useful, even when the data did not improve their decisions.

The finding matters for retail investors because financial data has become a product category of its own. Brokerage apps, newsletters, data vendors and research platforms all sell signals that claim to help users make better calls. Some data can be valuable. The study shows the price tag itself can change how people read that data.

The effect comes from a behavioral economics idea known as the price-value heuristic. A heuristic is a mental shortcut. When people cannot easily judge quality, they often use price as a stand-in for value. The classic example is wine: people may expect an expensive bottle to taste better than a cheaper one, even if a blind tasting would not support that view.

Financial information is a natural place for that shortcut to show up because investors often cannot tell, in real time, whether a signal is genuinely useful. That is especially true for alternative data, a broad term for nontraditional information used in investment decisions. Hedge funds and other investors often pay for these data sets in the hope of gaining an edge.

What the researchers tested

In a paper available on SSRN, researchers from Germany and Canada recruited 774 volunteers and asked them to trade stocks in an experiment. Participants saw historical price charts. Some also received order imbalance data for free, while others had to pay for the same type of information.

Order imbalance measures whether there are more buy orders or sell orders in a market. When the imbalance is large, it can point to buying or selling pressure that may move prices. When the imbalance is close to zero, it carries little information because neither side is putting much pressure on the market.

The key test was whether paying for the data changed how people used it. According to the researchers, it did.

• Participants who paid for the order imbalance data were 11 percentage points more likely to judge it informative, even when it was objectively uninformative.
• When the data had little signal, paying users still tried to use it. The researchers found their forecast error rose by 1.08 percentage points in those cases, a relative increase of about one-eighth.
• Participants who paid for the data put an average of 7.1 percentage points more of their money into risky assets.
• The study also found that paying users adjusted their portfolios less in response to market changes. The elasticity of their investments fell by about 15%, with stronger effects among people with lower financial literacy.

The investor takeaway

The study does not say all paid financial data is useless. It says the value depends on whether the data contains real information for the decision at hand. If a data set is clear and relevant, it can help. If it is noisy or uninformative, paying for it may make investors more confident in a weak signal.

That is the practical warning for anyone using premium market tools. A higher price can feel like proof that a signal deserves attention, but the experiment found the act of paying can make people rely on data more heavily than its content justifies.

This story draws on original reporting from Klement on Investing.