Nevertheless it seems like a flaw in the study if it's not counting other liquid sources of cash, such as checking accounts, brokerage accounts, cash itself, etc.