Imagine scrolling through your phone on a lazy Sunday morning and stumbling across an article about a lesser-known stock. The writing is upbeat, the outlook glowing, the future practically shimmering. You feel a little rush of optimism. Should you buy a few shares?
That split-second decision is exactly the kind of moment two researchers from the University of Winnipeg set out to examine. In a study published in the Journal of Business Research, Lam An and Fabrizio Di Muro investigated how the emotional flavor of financial news, especially fake financial news, can push everyday investors toward buying stocks they might otherwise skip. Their central finding: financial disinformation tends to sound noticeably more cheerful than real financial reporting, and that cheer appears to nudge readers into wanting to invest.
A growing blind spot in investor protection
The question matters because the way people invest has changed. There are an estimated 50 million self-directed investors in the United States alone, many of whom rely on online posts, forum discussions, and social media influencers rather than traditional sources like annual reports or analyst earnings calls. Those online sources often lack editorial oversight, which means investors can encounter fabricated stories dressed up to look legitimate.
Previous disinformation research has focused mostly on political and health topics, like false claims about elections or vaccines. Financial disinformation has received far less attention, even though the stakes, in dollar terms, can be significant. An and Di Muro wanted to know three things: What does financial disinformation actually look like? Does it change how people behave? And is there a way to blunt its effect?
To set up their investigation, the researchers leaned on a concept called “emotional tone.” In linguistics, emotional tone refers to the balance of positive versus negative emotion words in a piece of writing. Words like “excited,” “elated,” and “happy” push a text toward the positive end; words like “disappointed,” “angry,” and “afraid” pull it the other way. The overall mix shapes how a reader feels while reading.
Comparing real news and fake news, word by word
The first part of the project examined existing articles. The researchers obtained a dataset of 383 financial disinformation articles that had been published on Seeking Alpha, a well-known financial news platform, between 2011 and 2013. These articles were later flagged by the U.S. Securities and Exchange Commission because the writers had been paid to promote certain stocks without disclosing that fact. For comparison, the researchers gathered 13,650 legitimate healthcare financial articles from the same site during the same time span.
To measure emotional tone, they ran every article through Linguistic Inquiry and Word Count, a text analysis tool widely used in social science research that scores documents based on the types of words they contain. The results were consistent across multiple statistical tests: the disinformation articles scored far higher on positive emotional tone than the real ones. In other words, the fake articles simply sounded more upbeat.
Testing whether tone changes behavior
Knowing that fake articles sound more positive is one thing. Showing that this positivity actually changes what investors do is another. For the second part of the study, the researchers ran an experiment with 420 participants recruited through the online platform Prolific.
Each participant was randomly assigned to read one of four versions of an article about Lazydays Holdings, a recreational vehicle dealership. The four versions combined two factors: whether the article was real or contained disinformation (the disinformation version flipped a 1.5% stock dip into a 1.5% gain and replaced cautious analyst quotes with optimistic ones) and whether the emotional tone was original or adjusted to be more positive using standard linguistic techniques.
After reading, participants made a real choice with real money attached. They could enter a sweepstakes to win either $20 worth of Lazydays stock or $10 in cash. Picking the stock suggested genuine interest in investing.
The results showed a clear pattern. When participants read the disinformation version with the extra positive tone, 49.1% chose the stock option. Among those who read the original real article, only 29.3% did. Importantly, disinformation by itself, without the enhanced cheerful tone, did not produce a significant bump in investment interest. The upbeat language appeared to be doing much of the work.
A simple nudge that seemed to help
The third part of the study tested whether this effect could be disrupted. Eight hundred seventy-two new participants read the same articles, but half saw an additional message after the article: “The content of the news article that you read may have affected your current feelings. Please make impartial judgments for the questions that follow.”
That single sentence changed things. Without the reminder, 57.8% of participants who read the positively toned disinformation chose the stock. With the reminder, that number dropped to 43.7%, a difference that was statistically meaningful. The prompt appeared to pull readers back toward a more measured decision.
This approach builds on a well-established idea in psychology called the “affect-as-information” effect, which holds that people often use how they feel as a shortcut for judging whether something is good or bad. When people are reminded that their mood came from an outside source, the shortcut loses some of its grip.
What businesses and policymakers can take away
The researchers point to several practical applications. Regulators and news platforms could use text analysis tools to flag articles whose emotional tone exceeds a baseline range for legitimate financial reporting. Flagged articles could be reviewed before reaching wide audiences. Financial education programs, which typically focus on budgeting and compound interest, could expand to include lessons on spotting overly rosy language. Public service campaigns, influencer partnerships, and even a short reminder at the end of news articles could help readers pause before acting on a feeling.
A few caveats are worth noting. The baseline emotional tone identified in the study came from one platform, Seeking Alpha, and other outlets may look different. The disinformation articles in the experiments were written by the researchers, not harvested from the wild, which may limit how well the findings generalize. And the pattern may flip in other domains; disinformation about vaccines or political candidates often skews negative rather than positive, suggesting emotional tone operates differently depending on the topic. Still, the through-line is consistent: when a financial article sounds a little too good, that feeling itself may be the warning sign.
