How Could a Declining Housing Market Hurt the Recovery?, Kiyotaki-Moore Edition.

In responding to Kevin Drum, Ryan Avent wants to know the channels in which falling house prices could cause a deeper recession.  Shouldn’t it net balance out?

Well, first off falling house prices de facto increase the leverage of households and, as the Eggertsson/Krugman model shows, the effect of the following deleveraging shocks are to push interest rates further negative.

But I’m going to try and build this up because this story is important.  There’s three steps to approaching this: balance sheet effects, asset fire sales, and then putting them together into a model with asset market feedbacks.

Balance Sheet

Let’s say I try to keep a consistent net worth and a consistent leverage ratio at all times. As an example, let’s say I have a house worth $100 and I keep a leverage ratio of 3.

Assets = 100. Debt = 75. Equity = 25.

Now let’s say that the value of my house drops $10. Now I have:

Assets = 90. Debt = 75. Equity = 15.

My leverage ratio has now increase from 3 to 5 (75/15), meaning I have much more leverage than I did before. I can’t function that way so I need to cut my debt load down to 45. I do this by consuming less and deleveraging for a period. I can also do this by selling off units of the asset, which is housing. I would have to sell $30 of housing in this example to get back to my ratio.  Notice how leverage can cause major swings in these examples.

(Since it’s going to hang over the rest of this, the first complication is why assume a consistent leverage ratio/net worth for households? Corporate Finance theory can get us to arguments about why keeping a consistent net worth for entrepreneurs is optimal for investors; arguments like this can be generalizable to households on the part of their access to credit. The Eggertsson/Krugman paper does a similar move: “We will assume, however, that there is a limit on the amount of debt any individual can run up. Implicitly, we think of this limit as being the result of some kind of incentive constraint; however, for the purposes of this paper we take the debt limit as exogenous.”)

Fire Sales

The next problem is the problem of fire sales. If there is a positive correlation of distress between buyers and sellers of a good then assets will sell at a discount relative to value.  In English, if you are selling your home because you are unemployed and/or overlevered, and unemployment and overleverage is not just an individual phenomenon but characteristic of the economy as a whole, potential buyers are also likely to have weak earnings, worries about employment, constrained access to credit markets and worries about net worth. As such the home will sell at a large discount.  You are currently seeing things like recoveries of 40-60% in foreclosure, gigantic haircuts on foreclosed properties, so this is happening. However this drives down the value of housing…

Asset Market Feedback

What happens when you put the balance sheet mechanism together with the fire sale mechanism?  You get the Kiyotaki-Moore model from their paper Credit Cycles (1997).

Notice the overlap.  Declining value puts pressures on balance sheets which puts selling pressure on assets, which if sold at fire-sale values marks down the net worth of assets, which increases the leverage of other homeowners that they see their property is worth less, which puts pressures on balance sheets, etc.

Or think of it as this chart from the Kiyotaki-Moore paper:

Which they describe as:

The transmission mechanism works as follows. Consider an economy in which land is used to secure loans as well as to produce output, and the total supply of land is fixed. Some firms are credit constrained, and are highly levered in that they have borrowed heavily against the value of their landholdings, which are their major asset.

Other firms are not credit constrained. Suppose that in some period t the firms experience a temporary productivity shock that reduces their net worth. Being unable to borrow more, the credit-constrained firms are forced to cut back on their investment expenditure, including investment in land. This hurts them in the next period: they earn less revenue, their net worth falls, and, again because of credit constraints, they reduce investment. The knock-on effects continue…

For the market to clear in each of these periods, the demand for land by the unconstrained firms has to increase, which requires that their opportunity cost, or user cost, of holding land must fall…The fall in land price in period t has a significant impact on the behavior of the constrained firms. They suffer a capital loss on their landholdings, which, because of the high leverage, causes their net worth to drop considerably. As a result, the firms have to make yet deeper cuts in their investment in land. There is an intertemporal multiplier process…

In fact, two kinds of multiplier process are exhibited in figure 1, and it is useful to distinguish between them. One is a within-period, or static, multiplier. Consider the left-hand column of figure 1, marked ‘‘date t ’’ (ignore any arrows to and from the future). The productivity shock reduces the net worth of the constrained firms, and forces them to cut back their demand for land; the user cost falls to clear the market; and the land price drops by the same amount (keeping the future constant), which lowers the value of the firms’ existing landholdings, and reduces their net worth still further. But this simple intuition misses the much more powerful intertemporal, or dynamic, multiplier. The future is not constant. As the arrows to the right of the date t column in figure 1 indicate, the overall drop in the land price is the cumulative fall in present and future user costs, stemming from the persistent reductions in the constrained firms’ net worth and land demand…

Land in that example is isn’t a house but “land”, something that is productive and can also function as collateral (think airplanes for airlines).  This mechanism is important for the fire sale channel part of this – when an airline goes into crisis chances are other airlines are in crisis and the only buyers are unproductive users of the collateral, say someone who wants a private jet for fun (and will pay less because they’ll profit less from their use).

So is land being productive (an output), necessary for feedback from the future, a reasonable assumption?  Well no, but we do see that high leverage is well-correlated with a general lack of demand, so mechanisms that produce higher leverage will feedback through weaker local economies, which will cause more fire sales, which will produce higher leverage and have potentials for spirals.  Here’s Atif Mian and Amir Sufi on the link between leverage and weak demand.  Here’s the graph for auto sales:

And employment growth:

Avent’s question assumes a frictionless market.  What is necessary for these issues to complicate the picture?  Credit frictions (our people can’t carry arbitrarily high leverage), debt which produces large swings, fire sales and low recoveries, those low recoveries being observed as as the value for the balance sheet and deleveraging causing unemployment.

Is it a slam dunk channel?  No.  But it does have me worried that a drop in housing will hit the economy hard.  And it makes me think the lack of mechanisms to put some checks on firesales – say modification of mortgages in bankruptcy – and going with the “liquidate the homeowners” approach was a major mistake for the Obama administration.

This entry was posted in Uncategorized. Bookmark the permalink.

13 Responses to How Could a Declining Housing Market Hurt the Recovery?, Kiyotaki-Moore Edition.

  1. Mark T says:

    Alythough I generally agree with this post, I would not describe the Obama administration approach as being “liquidate the homeowners”. It was more of a muddle through, triage approach than something so extreme.

    I would also note the phrase “high leverage is well-correlated with a lack of demand” applies on a fiscal level as well.

  2. mattrognlie says:

    Nice post.

    I will say, though, that while Kiyotaki-Moore is one of my all time favorites in the macro-finance literature, I have trouble mapping it very precisely onto the real world. In Kiyotaki-Moore, no “land” is ever wasted; it is either used by “farmers” or “gatherers”, and binding credit constraints mean that farmers use land more productively at the margin. A shock (and the subsequent feedback) has a real impact because it shifts land from farmers to less efficient gatherers.

    In actual recessions, however, it doesn’t seem that inputs are so much used by some alternative, less efficient sector where credit constraints are not binding as they are not used at all; capacity lies idle and workers sit unemployed. To be fair, in the less popular “extended model”, Kiyotaki and Moore introduce firm-specific investments called “trees” associated with the land that drive a wedge between the internal and external value of land. Although this change serves another purpose in the model (reducing the unrealistically high leverage ratio in the model from its baseline level of 1/R, where R is the interest rate), I suppose that a large enough shock could make farms poor enough that they are forced to sell their “trees”, and that the resulting inefficiency could be interpreted as the temporary decline in output from a financially distressed company liquidating its firm-specific assets. But I don’t think that this is really the main impetus behind declines in output during a recession—output falls because factories and people are run at less-than-full capacity, not because firms literally acquire liquidated assets and take a long time to learn how to use them. We could stretch the interpretation a lot and say that firms are liquidating human assets, but this analogy wouldn’t really hold up (there are too many differences between the characteristics of employees and the “land” discussed in the paper). In sum, I think that we need some channel for the real impact of fire sales and deleveraging that isn’t present in Kiyotaki-Moore.

    Of course, as you mention, Kiyotaki-Moore is also a story about supply rather than demand. Thus we also need some way through which a deleveraging cycle can affect demand. At the aggregate level, this isn’t quite as obvious as it sounds: the real interest rate should adjust to clear the market, and the only thing stopping it is the zero lower bound faced by the Fed on nominal interest rates. (The Eggertsson and Krugman paper essentially provides a reduced-form example of a deleveraging shock in order to make this point, and to subsequently bring out New Keynesian machinery–with all its usual implications.) I think that effects at the regional level, however, are both more interesting and less well-understood, and I’m glad that your post emphasizes this aspect. The standard “everything would be fine if only the Fed didn’t face the zero lower bound” argument works wonderfully if we’re looking at aggregate outcomes in a relatively homogeneous country, but given the large region-specific shocks caused by deleveraging it doesn’t seem to apply anymore, and I don’t have very good intuition about what should happen.

  3. Your “Balance Sheet” section ignores the fact that not everyone is a home owner. For us non-home owners, NPV rent payments are a liability on our balance sheet and they’re going down with lower house prices. Plus you’re making an analogy between K-M’s temporary productivity shocks and house prices. Is this analogy appropriate?

    • That was too dense: my first point was that decline house prices is an decreasing asset value to some people (leading to leveraging) but its a decreasing debt for the rest of us (leading to deleveraging). The macro implications are not obvious.

      Second, I should have said: your analogy is not apt! And not because the model uses the word “land”. Its not apt because decreasing housing prices are not temporary.

      • Mike Easterly says:

        This is a very interesting point, one worth considering, and I’m quite impressed with you for coming up with it. Still (and you knew there was a “but” coming), I’m skeptical. Are the two effects likely to be commensurate? Observationally, I haven’t seen rents go down that much, but home prices have fallen off a cliff. Plus, you and I, as renters, are likely to start from a position much less leveraged than home owners, and we’re less likely to take on new leverage now to support consumption, especially with the economy in the state it’s in. (I may be way off base here; I have no idea what our fellow renters’ borrowing behavior is.) Thus our deleveraging could be miniscule compared with the increased leverage of home owners.

        I hope this makes sense. As you’ve indicated, it’s hard to discuss this in coherent prose. (My math may also be off on this one; sadly, it’s still early in the morning for me cognitively.)

  4. Pingback: FT Alphaville » The deflation risk is still out there, SocGen says

  5. LosGatosCA says:

    I’d say it was a major, major mistake.

    Mortgage modifications, especially through federally subsidized buy-downs, would have been triple bang for the buck. Employment for the folks involved in administering the programs, lower principle on outstanding home loans, and contributing to the bank balance sheet clean up, by reducing outstanding loans, reducing the non-performing loan volume.

    Instead the bankers got their cash directly and still have all the crappy, misvalued loans on their books, the borrowers are underwater and being foreclosed and no one was ever hired to administer the programs.

    On top of all that, Bernanke was a terrible, or uninterested, negotiator so all the loans directly to the banks were provided without any conditions for improved regulatory oversight or virtually any consideration at all.

    Could not have been done much worse, starting in Fall, 2007 through the present.

  6. Living as I do on “Main Street” and not in an ivory tower, I just have to say these models seem out of touch with the reality of the housing meltdown.

    Housing prices have been smashed – SMASHED – by the crisis – at least in my neck of the woods (Chicagoland.) Consumers do not look at their house as a piece of their leverage equation. It serves primarily as their home and secondarily (for decades until the meltdown) as a steady and reliable investment.

    No one I would know would ever look at the decreased market value of their house and say “I would have to sell $30 of housing in this example to get back to my [leverage] ratio.” And in reality, Mike, a house cannot be sold in bits and pieces. So this leverage equation is not useful in illustrating the impact of the housing crash.

    What consumers are saying is that “in order to get to a new job elsewhere, I need to take a serious hit to my bottom line.” In my area, prices have crashed 25% – 33% since the high of 2006 – with the bottom nowhere in sight. This impacts the homeowner in many, many ways, not just in their leverage ratio.

    If a homeowner is older and eager to retire, they either have to sit in a big over priced elephant for some years or lose significant dollar amounts by selling their house at a loss, which could seriously diminish their standard of living during retirement.

    And if the homeowner is targeted for a promotion in another part of the country, they have to factor in the dollar amount of the loss of their primary asset (home) with the benefits of the new job. Inventory is excessive right now and homes are sitting on the market for months. Salaries and relo benefits for the middle class generally don’t cover two housing costs (old mortgage and new rent.) These are serious impediments to an employee’s mobility.

    That’s why many are saying “no” to the move.

    The Kiyotaki-Moore model also seems less relevant to the housing market.

    “Being unable to borrow more, the credit-constrained firms are forced to cut back on their investment expenditure, including investment in land. This hurts them in the next period: they earn less revenue, their net worth falls, and, again because of credit constraints, they reduce investment. ”

    Single-family homes are not revenue producing entities. What that model fails to recognize is that the housing crisis is yet another blow to America’s middle-class. These are people already struggling with stagnant wages – as they also see significant increases in health care and college costs.

    The middle-class is being threatened with extinction. That’s the true cost of the housing crisis.

    To see their net worth evaporate in the massive scam we just witnessed is a terrible shock. Foreclosures are at their highest – as are underwater mortgages. [Who ever heard of an underwater mortgage before the crisis?]

    To see bankers get bonused and homeowners reap nothing but loss is a serious political problem.

    To see that the financial community’s complete disregard for business ethics and loan standards is apparently legal and aboveboard [who’s been charged with lax lending standards?] smashes just not the price of homes, but the consumer’s trust in home ownership. Many consumers are questioning whether they’ll ever own a home again – it’s a huge cost with no gain at this point. [Rentals are much bigger than home purchases in my area. Relo experts I’ve talked to also point to a growing demand for rentals.] And that will have social and political implications for years to come.

    You can’t shrink down the housing crisis into a mathematical equation. It would be nice to wrap it up neatly like that, but it’s far too emotional a topic – and far too messy a catastrophe for such neat and orderly equations. What’s painfully obvious to most people I know: the drop in housing values “will not hit the economy hard” – because we’ve already taken the hit – and it’s rocked us to the core.

  7. SummerW says:

    I have done a research paper for this topic also.
    Lots of male workers in U.S. are working for construction, so now, when there is no one buying and selling house. They became employees.

    If the situation is not getting any improvement, it is going to have some serious consequences.

  8. SummerW says:

    UNemployees****

  9. JamIN says:

    How could a falling housing price hurt the economy already in recession !
    Kiyotaki–Moore model of credit is more about profit and saving rather than going even or economical in bad times which is what may be remedial.
    For example it is necessary to look at the entire chain of events of the housing sector.As mentioned already the cycle has seen its hay days and couldnt hold on for long. Why?
    And Clearly it was highly dependent on the big bully macro economic features, which i say assumin small factors of seasonal fluctuations are unable tp play their part.
    In fact, if the housing economy goes even with chunky earning in this hovering economic phase for a time,and let go of the notion of forever living head held high, things should play back at our hands along with the economy. I am an optimist.
    However, deriving margins out of housing sector may well prove what we are avoiding- a crash landing.To be more precise a crash landing with the ecomony onboard.

  10. JamIN says:

    As someone pointed above, future is not constant. Forgive my understanding but any artificial leveraging, will only make it out of tune of the present crisis.
    I would rather admit I say this keeping in mind the vastness of the crisis which is well in tune with the overall US economy and not vice versa.

  11. meina77 says:

    This is true and I agree, and as Jam said, future is not constant. But some factors as bad economy can have an effect on housing economy, and especially if there is high unemployment in these areas, therefore forcing residents to sell their houses and prices to go down. However, some people take this as an advantage who can’t find a good house at a reasonable cost. So actually at the end economy will balance each other. This is just my point of view.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s