When companies publicly declare that they have suffered a data breach, it's best not to reveal how many individual records were involved if they don't want to take a hit in their stock prices, according to a study.
The Heartland breach last year involving 130 million lost records set off a plunge that reduced its stock price by 90%, and it hadn't fully recovered a year later, according to the Perimeter E-Security "U.S. Data Breach Study of 2009" report. Smaller breaches triggered stock-price drops of 12% on average that were made up for in about 60 days, the study says.
But when companies don't reveal how many records were compromised, there is no discernible impact on the stock price. "When it is a high-profile, largely publicized breach, it seems to impact the stock heavily," the study says. "When a company does not disclose the total number of records lost, there appears to be no statistically meaningful impact to the stock."
Perhaps businesses have already figured this out. The study says that last year in cases when financial data records were lost, two-thirds of the public reports of these incidents did not state how many records were involved, and this seems to be a trend. "In 2008, 42% of incidents did not include the number of records compromised," the data-breach study says.
The trend is true outside financial data, with the overall percentage that don't tell how many records were compromised approaching 40%, the study says.
Because all laws governing the disclosure of breaches are state laws, the rules vary from state to state. Some allow breaches be kept entirely private if there is no "reasonable likelihood of harm" a vague term that is not defined by the laws, the study says. Only five states have no disclosure law: Kentucky, Louisiana, Mississippi, New Mexico and South Dakota.
If lost data is encrypted, that generally exempts the business that lost it from having to disclose that it was exposed. In a few states, if just the encryption keys were compromised or a person with access to the keys is involved, disclosure is required.The study notes that the number of breaches doesn't necessarily correspond to the number of compromised records. For example, 2009 had the most records compromised ever, but had the fewest number of breaches in the past four years. That number was skewed by the Heartland breach, which alone accounted for 130 million out of 215 million records compromised, according to the study.
Methods of reporting can also confuse the actual breaches. In the Heartland case, the breaches occurred in 2007 and 2008, but they were reported in January 2009, so they are attributed to 2009 for analysis purposes, the study says.
Data breaches in general declined from 564 in 2008 to 327 in 2009, so the number of incidents per category of breach of how data was breached generally declined. One exception was the careless disposal of documents, which spiked 130% from 30 to 39. Also up were viruses, which rose 125% from 8 to 10. Stolen laptops remained the most common despite declining 54% from 2008 to 2009, and hacking was No. 2, down 48%.