In terms of its longevity and profundity, the Great Recession of the last decade was second to none since the times of the Great Depression of the 1930s. Unemployment peaked alarmingly, even as the GDP and median family income declined.

The role of mortgage-backed securities, rating agencies, prudential regulations, and housing policies have been debated at length and quick conclusions rather than a much-needed empirical evaluation have led economists and market pundits to dub the Great Recession a Subprime Crisis.

But what if they are wrong?

A recent paper by MIT researchers refutes this theory, showing that in the run-up to the housing collapse, both credit and defaults expanded proportionally across borrowers of every income level and every credit rating. The authors assert that tag “subprime” is a misnomer and the event could well have been the outcome of the actions of the prime borrowers.

Their theory is one that revolves around “housing expectation,” a phenomenon that happened when lenders caught the over-optimism bug and refused to admit that prices could come down once the bubble stage threw the economic fundamentals to the backburner.

Carried away as they were, lenders were fixated on the increasing collateral value, and this, in no small terms, skewed their mortgage assessments. Lenders felt that the skyrocketing home prices had given the housing collateral a far greater worth than what the loan-to-value ratio suggested, and thus they thought they were indemnified against any possible default.

So, when house prices dropped and the collateral value went bust, lenders did not have a place to run for cover.

The boom and bust cycle was driven more by belief than by some actual credit situation. Lenders that were really sanguine about America’s real estate threw caution to the wind, forgot that there is only so much that an economy can absorb when it closes on a bubble, and kept retaining crucial fractions of assets, thus triggering a mad price escalation. It all had to start going downhill soon enough, and it did.

As the MIT researchers point out, there is nothing to suggest that subprime borrowers were favored in the days of rampant lending. Also, those in the top percentile of household income and living in high house-appreciation areas defaulted a great deal more than expected, and because of the higher dollars attached to their defaults, the Great Recession needs to be reviewed in the light of defaults from prime borrowers.

Reading into the impact of LTV and DTI across population groups

It has been largely argued that the cross-sectional dislocation of credit during the boom period was inclined toward subprime borrowers. But data on homeownership and LTVs clearly suggests that the dislocation was mostly in areas that saw the steepest rise in house prices and where the phenomenon called “churn” was most evident. Churn, or house flipping, led to the conversion of equity into cash via cash-out financing, thus enabling households to sell their property and reinvest in a new one.

This trend was mostly apparent among prime borrowers and higher-income groups and certainly not the marginal borrowers. To add, the debt-to-income ratio also increased across all the population groups. So, while there was a definite rise in consumer sentiment involved, the upswing was not central only to the ambitions of the lower income groups.

For banks, the V of LTV was important, and they over-rated that V on an exponential scale (remember, increasing collateral values based on bullish housing expectations). To the mortgage-backed securities market, it did not matter which demographic or section of society the borrowers came from as long they were happy with the V.

The MIT researchers pull data that clearly shows how DTI increased across income groups. This is another crucial finding that attests the usage of the term subprime as a misnomer.

Churn: A strategy that backfired

When it comes to churning, or a practice associated with speculative buying, the phenomenon was most noteworthy in high-income groups. With banks talking favorably about house prices, optimistic middle and high-income groups flipped houses at a swift rate. Going beyond traditions and using housing as a leverage-based investment vehicle, these groups kept heightening their leverage with each subsequent churn. This strategy backfired when prices fell, and when the lenders came to their senses, it was already over bar the shouting.

It is noticeable that the subprime defaults are not too low even during stable and boom times. What marked the period of the Great Recession was a disproportionately large number of defaults from the middle- and high-income groups.

It is sad in a way, opines the MIT researchers in their paper on the topic, that the subprime view that led to the stringent screening of the low-income groups post 2008 virtually threw this demographic out of the homeownership market.

Sad because the housing expectation theory, which is deservedly gaining traction, clearly believes that shutting out one section from the housing market, especially at a time when the market had bottomed out, was critical to the real estate sector in particular and American macro-economy in general. 

The research findings should act as an eye opener for the lenders and policymakers.

When the next recession occurs, lenders need better data science and must avoid a kneejerk reaction. They need to embrace big data analytics, which helps examine large and varied data sets to uncover information including hidden patterns, unknown correlations and market trends.

Shutting out a certain section of the society from homeownership without the data to back it up would only result in a tougher and longer recovery.