Protecting the individual investor has long been a mandate of the U.S. Securities and Exchange Commission, a directive that was reaffirmed with the passage of the Dodd-Frank Act following the financial crisis a decade ago.
One way the SEC has tried to act on that mandate is by making sure that individuals are able to obtain information about the health of the companies they’re investing in, such as requiring firms to release quarterly earnings. But is making investors aware that the information is available — and making it easy to acquire — enough? New Wharton research indicates that in this push to protect investors, the SEC and others are missing key impediments that keep individuals from using that information to improve their returns.
In “Why Do Individual Investors Disregard Accounting Information? The Roles of Information Awareness and Acquisition Costs,” Wharton accounting professor Christina Zhu and co-authors Elizabeth Blankespoor, Ed deHaan and John Wertz of the University of Washington use an innovation that made earnings information more widely and easily available – the Associated Press’s (AP’s) rollout of robo-journalism earnings articles – to test which frictions are really keeping less sophisticated investors from using accounting to improve their trading behavior. The paper was recently published in the Journal of Accounting Research.
“Prior literature shows that when individuals fail to use accounting information, trading becomes hazardous for their wealth,” Zhu says. “They trade on trends and underperform because they’re not using value relevant information to improve their trading decisions.”
Zhu and her co-authors delve into that question further by examining why investors fail to do this. Two commonly held theories are first, that investors are unaware that the information exists (awareness costs) or that they are aware, but feel it would take too much time and effort to obtain the data (acquisition costs).
“Prior literature shows that when individuals fail to use accounting information, trading becomes hazardous for their wealth.”
The Role of Robo-journalism
The researchers studied nearly 30,000 earnings announcements by more than 2,000 firms that had received no coverage from the AP before the wire service began introducing automated earnings articles in October 2014. Prior to this rollout, earnings articles were written by individual journalists. The earnings article algorithm incorporates data from companies’ news releases, analysts’ reports and stock performance and allowed the AP to produce 12 times the number of earnings stories as before, according to a 2016 AP article about a previous, related study by Zhu and her co-authors. More than 60% of the firms studied by Zhu and her co-authors began receiving earnings coverage in the year following the introduction of the robo-journalism stories.
“There are several features of this setting that were beneficial for our research question,” Zhu says. “The robo-journalism articles involve no human decision to write the specific article about a specific firm. The robo-journalism articles were rolled out on a staggered basis and had nothing to do with whether a particular earnings announcement was positive or negative or contained more information than another.”
The articles were all written the same way, rather than a human journalist deciding which pieces of information to highlight and what to omit, Zhu adds. The only difference between some articles and others is that a subset included information about whether the earnings beat analysts’ projections or not.
“This is actually a feature of our design because an analyst consensus lowers acquisition costs for investors,” she says. “So if we find that an investor reading this article does not trade on earnings, then we know that lowering acquisition costs is not enough.”
Leaving Money on the Table
Indeed, although the articles made it easier for investors to acquire earnings information, the researchers found no evidence that they actually used it in their trades. Instead, they responded more to the trailing stock returns included in the articles.
“If we find that an investor reading this article does not trade on earnings, then we know that lowering acquisition costs is not enough.”
“We find that, on average, individual investors make zero excess profits when trading in response to the returns presented in the articles,” Zhu says. “Combined with the insight that there are transaction costs associated with trading, their losses will include both the time and money required to monitor and read these disclosures, as well as trading costs.”
She adds that implementing a trading strategy based on unexpected earnings would generate excess profits for investors of between 0.2% and 1.7% over a few days after the earnings announcement. “The investors trading in response to these articles are using the resources to acquire information and trade, but are leaving a significant amount of money on the table by not using accounting information in their trades.”
So if awareness and acquisition costs aren’t the issue, what else is going on? The researchers say that investors may be facing a third type of cost: integration costs, or the task of evaluating, combining and incorporating the information they find into valuation models and their trading decisions. “Because integration follows awareness and acquisition, integration costs can cause investors to forego fundamental analysis, even absent awareness and acquisition costs,” the researchers write. “Second, individuals may suffer from behavioral biases such as overconfidence in flawed trading strategies.”
What that means for regulatory bodies such as the SEC is that it’s not enough to simply tackle the problems of lack of awareness or the cost of simply obtaining information, Zhu says. “Integration costs could be targeted with more education about how to actually use accounting information in valuation models and in trading models,” she notes. “Investors need to understand how they would possibly use that information — i.e., whether a positive earnings surprise means you should buy versus a negative earnings surprise means you should sell. That seems to be something investors don’t know much about, based on our results.”
Another tough question to consider, Zhu says, is whether individual investors should be encouraged to buy and hold passive index funds, a strategy that tracks a market-weighted index or portfolio, rather than actively participating in the market. She said future research could include looking at how to reduce investors’ integration costs or behavioral biases.
“Now that we have this framework, it would be useful to investigate other things that the SEC might do, or what others can do about this,” she says. “The next step is to really focus on integration costs or behavioral biases. Now that we know what costs are not binding and now that we know what regulations are not likely to be effective, the next step is understanding what will be effective.”