Wharton’s Nikolai Roussanov discusses his research on the connection between mortgage refinancing and recessions.

For most Americans, their home is their most valuable asset, so it makes sense to borrow against the equity to obtain cash. In lean times, that money can be spent on consumption, which keeps the economy humming along. But if housing values and personal incomes don’t rise, borrowers might find themselves struggling to repay the debt.

That’s what happened a decade ago when the housing market collapsed. New research from Wharton finance professor Nikolai Roussanov shows that the pattern wasn’t particular to the Great Recession. Looking back 30 years, Roussanov and his co-authors found a cyclical pattern of refinancing before each recessionary period. The findings are presented in their latest paper, “Houses as ATMs: Mortgage Refinancing and Macroeconomic Uncertainty,” which is forthcoming in the Journal of Finance. His co-authors are Hui Chen, associate finance professor at the Massachusetts Institute of Technology Sloan School of Management and research associate with the National Bureau of Economic Research, and Michael Michaux, formerly a Ph.D. student at Wharton and assistant professor at the University of Southern California Marshall School of Business. Roussanov recently spoke with Knowledge at Wharton about these findings. (Listen to the podcast at the top of this page.)

An edited transcript of the conversation follows.

Knowledge at Wharton: The title of your paper includes the phrase “Houses as ATMS.” Can you talk about mortgage refinancing and how it relates to that term?

Nikolai Roussanov: The title is derived from a phrase people had been using at the time of the mortgage crisis that eventually led into the global financial crisis and the Great Recession. Even President Obama used that expression once. The notion of using houses as ATMs is the idea of homeowners using mortgage refinancing or home equity loans or home equity lines of credit to literally take cash out of their homes, meaning to borrow against their housing collateral or the equity that they have accumulated in their homes. They are then using that cash for spending or maybe paying off credit card balances, maybe even putting it towards other types of investments such as stocks or bonds and so on.

Knowledge at Wharton: In this paper, you’re trying to look at some of the economic activity around mortgage refinancing before and after the crisis. How did you study this?

Roussanov: To give some background, this idea of homeowners using their houses as ATMs had been around, and a lot of the discussion around the financial crisis was that households had become very highly leveraged over the course of the housing boom, which potentially contributed to the crisis and the Great Recession. The idea was that people had borrowed a lot against their homes when house prices were rising. And when house prices collapsed, they ended up having these unsustainably large debt burdens, which forced them to cut their consumption spending, which probably contributed to the recession and maybe added to the severity of the Great Recession. In fact, a lot of the discussion in macroeconomics and finance literature has been about the role of the great leveraging and then the de-leveraging over the course of the housing boom and bust.

What my co-authors and I looked at in particular was the role of refinancing, but we also took a longer perspective. We looked at the history of mortgage refinancing over the period starting from about the late 1980s through the more recent period. What we showed was that this pattern of home equity withdrawal or cash-out refinancing in the process of refinancing the first mortgage — it was very cyclical. It’s not just this kind of pre-financial crisis period where we see that, but in every recession, and we had three in this sample. We had the 1990-1991 recession, the 2001 recession that followed the dot-com boom, and then the Great Recession of 2007 to 2009.

“If we think about it from a standard economic standpoint, it’s not irrational for people to take advantage of their home equity … if they’re expecting that their incomes in the future will be higher.”

In every one of those recessionary periods, we see that households that were refinancing and taking advantage of lower interest rates going into the recession had been taking out some home equity before that. A lot of those households looked like they were refinancing into mortgages with not lower but even higher interest rates, which seems surprising unless you realize that a lot of the households that had home equity to withdraw weren’t refinancing to lower their rate but rather were refinancing in order to take the cash out of their houses. The question is why?

A lot of the narrative that you hear is that people were borrowing and spending recklessly, and they were taking advantage of these high house prices to do so. This might have been the case during the housing boom. But this was unlikely the case in the earlier recession periods or periods leading up to those recessions. If we think about it from a standard economic standpoint, it’s not irrational for people to take advantage of their home equity, to spend some of it, to convert some of it into cash to use it for consumption spending if they’re expecting that their incomes in the future will be higher. When would it be typical for people to do that? When their incomes are temporarily depressed, which is what happens in a recessionary period.

One thing I think people often forget is that the period in the early 2000s, when we saw prolonged low interest rates and a boom of refinancing that went along with this, was known as a jobless recovery. It was the same after the 1991 recession, where the economy recovered from the recession overall pretty quickly, but job growth was fairly stagnant. Federal Reserve Chairman Alan Greenspan was testifying in Congress in 2004, I think, and showing this as a success of his policy, the fact that keeping interest rates low allowed households to take out cash from their homes through mortgage refinancing or cash-out refinancing and convert it into consumption. In his view, that was helping the economic recovery.

Knowledge at Wharton: Because people were spending the money?

Roussanov:  Exactly. Even though a lot of them weren’t yet seeing their incomes improve. So, the idea from the economic standpoint is that if we think that people’s incomes are temporarily depressed, it’s not crazy for them to use cash that they have locked into their home as a temporary kind of buffer spending.

Knowledge at Wharton: But that’s contingent on their incomes rising. Sometimes they don’t.

Knowledge at Wharton: Exactly. This is kind of what happened during the Great Recession or the collapse of the housing boom that not only made the housing price gains evaporate but also made a lot of the jobs disappear, and those income growth expectations that people may have had didn’t materialize. Of course, what we ended up having was a pretty prolonged period of depressed economic conditions.

But it’s not something that people could have foreseen. It’s easy to say after the fact, “How could we not have seen it?” But again, if you go back to 2005, 2006, the economy is humming along. The housing market is on fire. And a lot of people are saying, “We know house prices don’t go down. They only go up.” As economists, we could say this is irrational, but the expectation was that house prices are going up and the incomes will catch up. It wasn’t that house prices got too high relative to incomes and they’re going to collapse, and that’s how the two will meet. That was clearly not the expectation that people had, even though that’s maybe part of what happened in the end.

“Be very careful when you get that offer in the mail from a bank saying you can refinance your mortgage and can get cash out because your house has appreciated in value.”

Knowledge at Wharton: It feels like before the most recent crisis, people had this idea that you use the value of your house to get you to the next bigger house. Since the Great Recession, people don’t really see it that way anymore.

Roussanov: Perhaps. That’s something that I can’t necessarily speak too much to because in the data we’re using, we don’t necessarily see people trading up or down. What we do know is that the rate of home ownership did go up a little bit over the course of the housing boom. Since then, it has gone down, which might in part reflect the fact that a lot of people are unable to afford the down payment even though interest rates have been very low and mortgages rates have been very low. Just the ability, or maybe the willingness, to make that investment has decreased. And that may have something to do with the fact that these expectations of future income growth — the future opportunities to make those mortgage payments and be able to afford larger homes — are not there any more because we have had a period of stagnant productivity growth, stagnant wage growth. So again, it’s not really surprising that the mindset has shifted a little bit. But we’ll see. Maybe this is a transitory phenomenon.

Knowledge at Wharton: What are the key takeaways for both policymakers and homeowners?

Roussanov: One of the key takeaways is that we shouldn’t be too quick to interpret what we saw during the 2000s, this kind of run up in leverage, as necessarily irrational or a result of some nefarious behavior by unscrupulous lenders. I’m not saying that didn’t happen, but I wouldn’t say that that was the predominant driver of that great leveraging and the de-leveraging that followed, however painful that may have been for a lot of people who were caught off guard by their house value plunging and their debts staying the same in the face of maybe wages not rising like they had expected. Maybe the lesson here is: Be very careful when you get that offer in the mail from a bank saying you can refinance your mortgage and can get cash out because your house has appreciated in value. Be very careful. Think about what’s going to happen maybe five or 10 years down the road, to whatever might happen to your income, to your employment prospects. Those debts that you rack up now will be there for you to repay. And default, bankruptcy or foreclosure is an unpleasant thing that nobody wants to find themselves facing. There could be a situation where you have to repay your debts, and your income is not quite catching up with that, which is not a pleasant situation to find yourself in.

The experience over the last 10 years has taught some people that, but maybe not everyone. Certainly, there is a new generation of first-time home buyers who may be tempted to tap into the cash that’s locked in their homes in the form of equity in some of the areas where house prices have grown. With interest rates once again touching historic lows — mortgage rates in particular — it may be very tempting to take equity out of the home. We have to think about the long term.

And yes, if people think that their incomes in the future are going to be higher than they are now, that they’re going to be stable, maybe that’s not a bad idea to smooth your consumption profile over time. Or if you’re actually planning to invest into your home, doing something that’s going to increase the value of it through some sort of remodeling and home improvement, then it’s not really pure consumption. It’s actually investment. It’s OK to borrow against that, within reason. But you have to think about what’s going to happen down the road with your ability to repay.

Knowledge at Wharton: Was mortgage refinancing created with the idea that you would invest it back into your home, as opposed to using the cash for something unrelated to that?

Roussanov: That’s a good question. I think a lot of the thinking that went into the creation of home equity lines of credit or cash-out refinancing probably had something to do with investment into your home, as opposed to into other forms of spending and consumption, like buying boats and cars and motorcycles. But it’s not necessarily a new-fangled concept, either.

“There is a lot of discussion about the causes of that house price run-up, and there is quite a bit of disagreement.”

If you read classic 19th century literature, wealthy estate owners around Europe, from Russia to England, would be mortgaging and remortgaging their estates, and then eventually getting themselves into debts that they could not repay and having their estates sold at auction. It’s not a result of a crazy financial innovation in the 21st century. This has been around for a long time. People use value or equity in their real estate to borrow against it and spend it to support their consumption, lifestyles and so on. It can get reckless occasionally. And just like some of the heroes of 19th century literature discovered to their great chagrin, you could get yourself into serious trouble doing that. Yet in the 21st century, people are still doing this.

Knowledge at Wharton: It sounds like this past recession was a big anomaly compared to the other two that you looked at, so people maybe couldn’t have predicted that. There has been a lot of talk lately about when the next recession will be. Are we on the cusp of that? There seems to be more uncertainty these days.

Roussanov: Uncertainty is always there with us. Sometimes we perceive it as being greater than at other times. Nobody can predict when the next recession will come. There’s a lot of talk about a recession coming, but that doesn’t mean it is going to happen this year or next year. It will probably happen eventually.

What I can say is the levels of indebtedness and home equity-type borrowing are not as high as they were preceding the Great Recession. Nevertheless, I think whenever the recession comes, we will see greater rates of foreclosure or at least some delinquency in mortgages and other types of consumer borrowing. We certainly see an uptick in consumer borrowing.

Knowledge at Wharton: What’s next for this research?

Roussanov: Our model explains relatively well what happened with mortgage borrowing and mortgage refinancing and with home equity withdrawal and consumption before and after the crisis. The big unanswered question is, why did house prices rise as much as they did before they came crashing down? That’s something that our research does not answer. There is a lot of discussion about the causes of that house price run-up, and there is quite a bit of disagreement. This is something that we would like to understand as well. This idea of using your house as an ATM and valuing the house not just as a place to live but also this potential source of rainy day cash could contribute to the demand for homes, particularly in the areas that are seen as attractive. Maybe that’s something that makes people more willing to pay for a house of a certain size right now, even though it may be more expensive than what they think they can afford or what they even need. That’s not an easy question to answer and something that we are working on. But we’ll see when we get there with that.