In the mortgage mess it seems that everyone was an agent.
- The buyer obtained 100% mortgages and therefore had little or no money at risk.
- The mortgage broker sold the mortgage to a Wall St. bank. Their risk was limited to the short time they held the mortgage.
- The Wall St. bank combined the mortgage into securities and sold the securities. Their risk was limited to the short time they held the mortgage.
- The rating agencies had no money at risk.
- Often the buyers were fund managers, pension fund managers, bond fund managers, hedge funds, etc. They had little or none of their own money at risk. In fact some of the bond fund managers were obligated to buy high yield bonds.
The "invisible hand" of Adam Smith assumes that prices are set by the market forces of supply and demand. There are free market fundamentalists that still believe that this "invisible hand" actually works to set prices so that supply and demand reach equilibrium. And this does work fairly well for consumer commodity products and services.
But it assumes that the buyers and sellers are rational actors, they are making rational decisions in their own best interest and that they both buyers and sellers have the same information. The agency problem is that if the buyers and sellers are agents and are not making decisions in the interest of the real buyer or real seller then the invisible hand simply isn't there. This is only one problem with the invisible hand. There are others which we may get to.
Contrast the players involved with the housing and credit bubble with the Bailey Savings and Loan system. In this system a local bank lent money to a developer - as a construction loan. Some of the money came from their depositors but some of the money was the bank's own money. The ratio between the depositors money at risk and the bank's own capital is leverage and the leverage was probably around 10 to 1. The banker would make the construction loan based onthe developer's past performance, their credit worthiness and the bank's assessment that the housing that was built could be sold. The developer had to put some of their money at risk as well.
When the contruction was complete the developer would sell the homes. The home buyers would get mortgages from the same bank. The bank would assess the value of the house, the credit worthiness of the buyer and would expect the buyer to put some money at risk - at least 20%. The homes are sold and the construction loan is repaid.
The banks hold the mortgage, collect fees and interest and pass on some of the interest (always at a lower rate) to their depositors. The difference in the interest rates results in profits to the bank less reserves for the possibility of some defaults. But even in the case of default the bank is protected since they assessed the value of the house and the buyer had 20% of their money in the house. Only if the value of the house decreased more than 20% was the bank at risk.
Now compare that simple model with the model used during the credit and housing bubble.
Since most of the players were agents and not principals they due diligence of assessing the overall housing market, the specific home assessment, and the credit worthiness of the buyer, the developed, the mortgage broker and the Wall St. bank were all neglected.
The buyer borrowed money from the mortgage broker.
The mortgage broker borrowed money from the Wall St. bank. They paid the money back when they sold the mortgage to the Wall. St. bank.
The Wall St. bank borrowed money from institutional investors and the ratio of money at risk to their own capital wasn't 10-1 it was 35 or 40-1. They paid the money back when the sold the mortgage securities to their investors.
And the agents became rich by paying themselves large salaries and bonuses while they put their corporations and the shareholders - the real principals at risk.
So when it all fell apart the shareholders or Countrywide Financial, Citigroup, Lehman Bros, Bear Stearns, Bank of America all went south but the executives walked away with their bonuses intact.
