Step 1: A home-buyer approaches a bank for a mortgage on a house, or a refinancing of an existing mortgage.
Step 2: Rather than send an appraiser to the house to determine how much should be lent against it, the bank enters data on the prospective home buyer and the house into an automated database such as the emili system operated by CMHC. For example, it asks how risky a $400,000 mortgage would be if lent against a home at 22 Elm Street.
Step 3: Using the information the bank provides on the home buyer, such as income and credit rating, and data on the house in question, such as average home prices for the area and the size of the property, the automated system begins crunching the numbers.
Step 4: Within minutes or seconds, the system tells the banker whether the proposed loan amount is within an acceptable range. The ranges are low-risk, acceptable risk or high risk. Some bankers compare it to a traffic light, with green, yellow and red alerts.
Step 5: Industry sources say the system can be gamed to write bigger loans. Since the system merely tells the banker whether a proposed amount is within an acceptable range, the banker and the client can continue entering higher numbers to find what the upper end of the 'acceptable' category can support.