Financial Crisis of 2008 through the lens of data

Part 1: Tracing the roots

Kathy Tran Anh Ngan

Published in

NYU Data Science Review

8 min readNov 17, 2023

2008, a year that lasted forever

“The housing market is rock solid.” That’s what they all said in the years leading to 2008 — from Bank of America to Wall Street investors, to the legendary Federal Reserve Chairman Alan Greenspan. No one saw it coming, as housing mortgages were believed to be the safest bet or to quote The Big Short: “Who the hell doesn’t pay their housing mortgage, right?” [1]. Yet, they were all wrong. It collapsed, fast enough to leave us in a suspenseful panic but slow enough for the panic to remain in our hearts even long after its occurrence.

The financial crisis of 2008 has always been a hot topic for its massive impacts, the biggest of which was the collapse of financial institutions that were universally considered “too big to fail.” It was written in a New York Times article in 2010 that: “Alan Greenspan, the former chairman of the Federal Reserve, proclaimed last month that no one could have predicted the housing bubble” [2].

Is that really true?

The author of the aforementioned article, Michael J Burry, was the manager of a hedge fund based in California called Scion Capital when the crisis occurred. In the mid-2000s, before anyone else could, he recognized the possibility of a housing crisis and decided to bet against the housing market, which was another way of saying: betting against the world. The crisis occurred, just as his prediction, and he ended up making 100 million dollars in this deal.

So what did Dr. Burry do to predict what was considered unpredictable? The answer is simple: he looked at the data.

Where the story began…

Almost every single financial crisis started with a “bubble” — the phenomenon of when a certain asset’s prices skyrocket. The symbolism of a “bubble” suggests that its outburst is unpredictable and that a burst, or when the price reaches its peak and starts decreasing rapidly, would cause significant financial damage to the economy. There were several instances in history when rising prices proved to be a sign of financial crash. Two of the most popular examples would be the Great Depression of 1930, caused by a bubble in American stocks, and the Dot-com crash in 1995, characterized by a bubble in internet-based start-up stocks. Therefore, one could point out that the first signal of the housing bubble would be the rapidly rising prices of houses from the 1970s to the mid-2000s. Figure 1 shows that the average housing price grew from $39,000 in 1975 to approximately $240,000 in 2005, which indicates a 515% increase over the course of 20 years.

Housing mortgages at the heart of the bubble

In the midst of soaring housing prices, housing mortgages emerged as the most popular type of loan. A housing mortgage, a contract that the bank gives to a borrower to purchase a house, allows the bank to take the home if the borrower fails to pay back the loan — referred to as a default. But the housing mortgage was not the problem. The main problem was subprime mortgages — the housing loans made to people with high risks of default.

Normally, when borrowers request a mortgage at a bank, the bank will use several criteria to assess the borrowers’ ability to pay back the loan including credit score, income, employment, payment history, etc. Generally, banks will only lend loans to people with high scores on these criteria, which indicates that the borrowers have a high possibility of paying back the loan. However, the 2000s witnessed a surge in what we call unethical lending practices: banks giving mortgage loans to people with high risks of default in order to maximize the number of mortgages they could sell. Figure 2 illustrates the credit scores of mortgage borrowers from 2003 to 2020. We can see that compared to the years after the financial crisis of 2008, the years from 2003 to 2007 experienced a much higher percentage of borrowers in the credit score group of 620–659 (yellow) and less than 620 (dark blue) — the two lowest groups. The light blue and dark grey sections represent borrowers with credit scores above 720, the strongest loan payers. Yet, between 2004 and 2007, these groups only accounted for no more than 50% of the borrowers, indicating the credit weakness of the mortgage borrowers at the time.

Such low credit scores of loan borrowers were a major signal of a housing bubble burst because they indicated a high risk of default on housing mortgages, as most of the borrowers had unreliable financial conditions.

The motivation behind unethical lending practices

There were multiple reasons for banks giving out a large number of mortgages to borrowers with a high risk of default. During this time of increasing house prices, housing mortgages were a “bank’s favorite” because they were considered safe with high returns. In other words, if the borrower defaulted, the bank could always sell the house for a higher price since house prices were increasing so rapidly. Such guarantee in profit allowed banks to be more reckless in evaluating customers’ financial reliability.

Simultaneously, the government also put effort into encouraging homeownership, especially among low-income people and minority groups through their affordable housing rules. For instance, the Department of Housing and Urban Development (HUD) required Fannie Mae and Freddie Mac to purchase mortgages loaned to low and moderate-income borrowers. Such policies encouraged lenders to offer mortgages with lower down payment requirements, making it easier for borrowers with limited savings to enter the housing market.

But perhaps the most convincing reason behind the banks’ recklessness was the fact that they would sell the bonds to a third party. When commercial banks lent out mortgages to borrowers, they usually did not intend to keep these mortgages, but instead, they would sell them to investment banks, which would then sell shares of pools of these mortgages to another party — usually investors. As a result, the party that took on the risk of mortgage defaults was usually the investment bank or investors, serving as motivation for commercial banks to lend out as many loans as possible without fearing the risk of defaults. In fact, no bankers at that time would reject any housing mortgage applicants no matter how unfavorable their financial status was, as they knew an investment bank would buy that mortgage the next day.

Altogether, these factors led to a significant increase in the number of subprime mortgage loans in the 2000s. As depicted in figure 3, the percentage of subprime shares rose from roughly 6% in 2000 to about 21% in 2006, representing an approximately 250% increase. As banks loosened financial requirements, almost anyone could get a home loan, causing homeownership to rise noticeably from 7% in 1997 to 20% in 2006 — a 185% increase.

The trap of adjustable-rate mortgages

Another aspect of unethical lending practices was how banks convinced low-income borrowers to buy the mortgage. With adjustable-rate mortgages, borrowers paid interest on the loans as it varied over time. These kinds of mortgages typically offered lower initial interest rates compared to fixed-rate mortgages, making them more affordable for borrowers in the short term. As we can see in figure 4, compared to previous years in the 20th century, the 2000s experienced much lower interest rates, which decreased considerably from 5% in 2000 to an all-time low of 0.75% at the end of 2002 and remained relatively low at around 2% to 3% until it gradually went up from 2005 to 2008.

By using adjustable-rate mortgages, banks would tell low-income borrowers to be assured of their ability to pay back the loan because the interest rate was at a record low at the time. However, what banks did not tell the borrowers was that if the Fed increased interest rates and house prices went down, borrowers would not be able to refinance (the process of obtaining a new mortgage loan to replace an existing one in the hope of a lower monthly payment rate). In other words, if home prices stopped going up, these borrowers would be stuck with paying a high-interest rate on their mortgage.

The book The Big Short describes this situation: “You’re basically drawing someone in by telling them, ‘You’re going to pay off all your other loans — your credit card debt, your auto loans — by taking this one loan. And look at the low rate!’ But that low rate isn’t the real rate. It’s a teaser rate.” [3].

“They defraud the American people and prey on their dreams of owning a home.”

-The Big Short-

Stay tuned for part 2: From Burst To Dust

It’s not the bad times that indicate danger, it’s the good times that do. The graphs that illustrate the events leading up to the outburst of the crisis all show positive signals for a stable and healthy housing market: skyrocketing house prices, falling interest rates, and rising home ownership. Yet, the real “bomb” has yet to be revealed. In part 2, we will explore the core and peak of the crisis as well as the housing bubble outburst that terrified the entire world in 2008.