Why Retail Traders Underperform

I’ve been meaning to write about this for some time, but other more pressing and timely topics intervened.

Three professors at Purdue and CUNY, wrote a paper entitled “Wisdom or Whims? Decoding the Language of Retail Trading with Social Media and AI” (Chen, Shuaiyu and Peng, Lin and Zhou, Dexin (April 02, 2025), https://ssrn.com/abstract=4867401). The object was to analyze investors’ social media posts to determine how they form their beliefs, select which trading strategies to employ, and their subsequent performance.

If you’re not familiar with academic papers, they are often very difficult to follow and have a style that only professors can love. But unlike many finance papers that have no relation to the real world, this one came up with some results that are worth considering.

Using large language models, the authors began by reviewing 96 million social media posts from 840,000 users on StockTwits. The model classified those that concerned investment strategies (31%) by specific strategy: technical analysis (TA, 29%), fundamental analysis (FA, 44%), or other, e.g. M&A activity (OS, 28%). Interestingly, these percentages agree with the approximately 19% of investors who self-reported their strategies. Not surprisingly, investors tended to use technical analysis when news was scarce and fundamental analysis when information became more plentiful.

After classifying the posts, the authors investigated the performance of the different strategies that the posts revealed. They found that “Technical sentiment negatively predicts stock returns,

particularly among short-term or inexperienced users, whereas fundamental sentiment positively forecasts returns.” They go on to say that “reliance on technical analysis may lead to poor short-term performance.”

Why is this? The authors examined investor sophistication and investment horizons to determine why the results of the two different strategies varied.

First, investor sophistication (self-reported professional experience) “plays a critical role in the overall performance of TA [technical analysis] and OS [other strategies].” Yes, investors who know what they are doing do better than amateurs! But, it’s not that obvious. The authors also found “…little effect of professionals on the predictive ability of FA sentiment.” Further, “This result may reflect a narrower dispersion in fundamental analysis skills among investors on the platform.” In other words, there was little difference in the skill levels of amateurs and pros when it came to fundamental analysis. Many Wall St. Economics PhDs are not going to be happy about that.

Interestingly, the authors examined the 2021 GameStop short squeeze as an example of how skill levels influence returns. It’s worth quoting them in full: “[Our results] indicate a general decline in the informativeness (i.e., positive return predictability) of sentiments across all investment strategies following the GME event. However, this reduction is most pronounced within the TA strategy category. The disproportionate decline in TA informativeness likely stems from the substantial influx of inexperienced retail investors drawn to the market by the high-profile nature of the short squeeze. Our previous finding with investor experience supports this hypothesis: professional investors typically produce more informative and reliable TA discussions. Therefore, we posit that the surge of novice participants disproportionately gravitated toward TA strategies—perhaps due to their greater accessibility and perceived simplicity — leading to a dilution in the quality and predictive value of TA-based sentiment.”

Second, investor horizons (i.e., short-term or long-term holding period) matter: “…the time horizon of an investor fundamentally influence the effectiveness of their chosen strategy, with longer horizons associated with better outcomes.” That’s not that surprising, but bad news if you’re a day trader.

What can we take away from all this? Personally, the findings don’t shock me. Yes, long-term investors tend to do better, whatever strategy they use. Very few traders succeed solely as day traders, or at least those without a clear edge. And without getting into the age-old argument as to whether technical analysis is a legitimate and statistically robust methodology or not, I think everyone would agree that like any tool, it takes a certain amount of expertise and training to use correctly. Otherwise, the results can be spurious, random, and easy to misinterpret. The paper bears this out – technical analysis is not for amateurs.