In effect, the seller of the swap is providing a guarantee that if the bond (that is the subject of the credit default swap) defaults, the seller will pay the buyer a specified sum of money. Numerous credit default swaps were bought/sold in conjunction with mortgage-backed securities that were issued in conjunction with subprime real estate loans in the mid-2000s. Although credit default swaps are often compared to insurance contracts, one important difference is that, with an insurance policy, the policyholder must also own the property being insured. In contrast, the buyer of a credit default swap need not be the owner of the financial instrument for which the swap is providing a financial guarantee. Thus, credit default swaps facilitate speculation (by buyers) as to whether a certain credit instrument will default. Another key difference from insurance is that the seller of a credit default swap—unlike an insurance company—is not required to maintain a specific level of reserves in the event that the subject instrument (e.g., a mortgage-backed security) defaults, and the seller must pay the buyer of the credit default swap. In 2008, AIG Insurance Company's failure to maintain adequate reserves on the billions of dollars in credit default swaps it sold was the major cause of the company's near-collapse.