Study: Corporate executives listen to analysts—and do the opposite

In the collective effort to set accurate prices, financial markets look to two main gestures from the experts: analyst recommendations, and corporate insiders' buying and selling of their own company's stock. We know that these two groups are aware of one another; in fact, associations between them can sometimes be questionably cozy.

For investors and market observers, this raises the question: Are analysts and corporate insiders in harmony, or dialogue? Is there any added value to considering the decisions of both, as opposed to just one? Jim Hsieh, associate professor of finance at George Mason University School of Business, says that "most finance papers talk about the so-called 'substitution hypothesis'–analyst and insider information tend to substitute for one another."

However, Hsieh's recently published paper in the Journal of Banking & Finance (co-authored by Lilian Ng of York University and Qinghai Wang of University of Central Florida) finds that these two camps of experts often behave in opposite ways, fueled by an apparent direct–yet one-way–interaction that holds meaning for investors.

Hsieh and his co-authors compared analyst recommendations covering thousands of firms to insider transactions (harvested from SEC filings) for the same companies over the period 1994-2016. After controlling for attributes like firm size and share price momentum that might otherwise motivate analyst/insider behavior, the researchers could see a general pattern in the response of corporate insiders to changes in analyst ratings.

"Insiders and analysts tend to disagree, most of the time," Hsieh says. "When analysts upgrade, insiders tend to sell and vice versa. If analysts and insiders have the same information, their actions should be in the same direction, but we don't see that."

The paper speculates that insiders purchase their company's downgraded stock to send a contrasting signal to the market, in hopes that it will cancel out the negative assessment from the analyst. There was a direct relationship between the size of the downgrade and the amount of stock insiders purchased, implying a signaling intent.

On average, the signaling strategy worked, resulting in a 0.33 percent boost in monthly abnormal returns. And because returns remained elevated three to six months after the insider purchase, the signals seemed to convey genuine information about the company's strengths, rather than being a smoke-and-mirrors attempt to manipulate the market.

"If there's no information to support the transaction, the price will drop," Hsieh says. "But we don't really see that. The stock price continues to rise. It's not consistent with the idea that insiders buy the stock to temporarily bolster the price."

The reverse situation, in which insiders sold stock following an analyst upgrade, could also have a signaling motivation. Insiders looking to cash in or diversify their portfolio may find the post-upgrade period a good time to sell, since the analyst thumbs-up may dampen the negative signal broadcast by the sale.

When insiders echoed analysts' signals–selling after a downgrade, or buying after an upgrade–the positive/negative effect on the share price was significantly heightened. For Hsieh, this is yet more evidence against the substitution hypothesis–if the two signals were redundant, their agreement would have little to no impact on prices.

Interestingly, analysts–unlike the broader investor community–did not appear to heed the signals embedded within insider transactions. Whether insiders bought or sold stock did not affect analysts' subsequent change in the rating of the company. The public conversation between analysts and insiders turned out to be one-sided, as analysts' attention was apparently elsewhere.

"Analysts have been widely criticized for ignoring important signals that could move stock prices," Hsieh says. "We don't know why analysts don't take advantage of the signals from insiders' trades. But we are not surprised by our results since other finance studies have shown that analysts only use a limited number of signals, such as momentum, firm size, and some accounting measures. But it is still puzzling that they ignore one of the most obvious signals (insiders' trades) that could significantly improve their recommendations."

Still, changes in analyst ratings, in and of themselves, provided informative cues to the market–particularly downgrades, which were associated with a 0.18 percent decline in monthly returns.

In Hsieh's opinion, this research suggests that despite close professional relationships that may exist between analysts and senior leaders, the two camps use discrete data-sets to make decisions. "Analysts provide value-pertinent information to the markets through analyzing public information while insiders are likely to exploit private information," the paper states. Ironically, this diversity of focus benefits the market on the whole, even as the mixed signals it creates may complicate investor choices.

"Investors need to look at how insiders and analysts interact and how they behave in response to each other," Hsieh says. "They need to pay special attention to insider buy-in, which is such an important signal. We also have to look at why insiders sell, but that's less informative than insider buying."