Two-thirds of Canadians surveyed by the Bank of Montreal are not able to keep their emotions in check when they make investment decisions and a majority have invested on impulse.
The BMO Psychology of Investing Report also finds four-in-10 respondents to the Pollara poll say emotions play a role in their investment decisions.
And 60 per cent say they've invested on impulse at least once.
The most common emotions respondents reported were anticipation, trust and fear.
The report also found that many Canadians are unaware of how to make an investment and lack the confidence to do so.
About 44 per cent say that they are not confident when investing, while only one-quarter do careful and extensive research when making an investment decision.
The report is the second in a series of Canadian mindsets when it comes to personal finances. A previous report found 60 per cent of Canadians engaged in impulse shopping and mood-lifting impulse purchases cost Canadians $3,720 annually.
Like shoppers, investors need to ensure they're in the right state of mind in order to make the best decisions that they do not later regret, said Serge Pepin, vice-president of investment strategy at BMO Asset Management Inc.
"While we're only human, wise investing means more than simply following your heart," he said.
Mr. Pepin added that reading the business section of newspapers and magazines, researching online and seeking advice from experts can help ensure they are making the most informed investment decision.
The report also revealed that 41 per cent of respondents don't believe they're investing enough — with 90 per cent of those saying they haven't saved enough to invest and 83 per cent saying they want to pay off debt before investing.
The survey results are from online interviews with a random sample of 1,000 Canadians 18 years of age and older, between Nov.6 and Nov. 8.
The polling industry's professional body, the Marketing Research and Intelligence Association, says online surveys cannot be assigned a margin of error because they do not randomly sample the population.