Back in the late 1990's when I was allocating economic capital, residential mortgages had the lowest amount of risk assigned to any major loan category. Most asset classes had about 7-9% allocations, mortgages about 4%. But this made sense in 1999, because back then people had 20% down payments and verified income commensurate with their debt payments. One would have never guessed, back in 1999, that among all the things banks could invest in, mortgages would be the trouble area. The Commercial Real Estate debacle of 1990, or the oil patch lending problems in the 1980's, or lending to South America, was considered the biggest risk.
I'd like to think that I would have increased my capital allocation through the 2000's based on my estimates of the new lending standards, or the increase in housing prices, but that's not obvious. After all, in 2005 even Robert Shiller was not predicting a fall in housing, only noting that it could not continue its torrid upward pace and that some areas might have some agressive declines. The LOB heads (line of business heads) would be arguing the new lending standards were basically the same, in that poor people are even more likely to sell everything to meet their monthly payments. The data at that time would have backed them up through 2006. Further, we will be sued or will not meet our CRA lending targets if we don't apply them, so it would have been hard to say that risk went up 100% because of what our erstwhile regulators were encouraging us to do. Regulators don't encourage banks to invest more in risky areas, after all.
In 1999, internet lending was primarily called 'structured' finance, and derided as 'air ball' lending because you lent based on the equity of the firm--an 'air ball' to experienced lenders. The collateral would only be counted on when it was by definition worthless. Banks had minimal exposure to this bubble, because they saw it coming.
We should not kid ourselves with all this focus on recent disastrous mortgage practices of banks. Having recently lost a boatload on residential mortgages, I suspect this will not be our problem area going forward. What is the current equivalent of residential mortgages circa 1999, the largest low risk asset on bank books? The only asset that did not explode in terms of its spread was US Treasuries. Given our new spending regime, poor demographics for social security, my best guess is that the banking sector's currently most riskless product is a good bet for our next lending cancer.