Wednesday, December 28, 2016

Housing: Part 198 - Bank Credit through the recession

I think this graph is useful in thinking about the recession and housing.

Source
Excess reserves really can be thought of as treasury bills.  If the Fed unwound the entire mass of excess reserves simply by exchanging it for treasury bills so that instead of reserves, the banks held treasury bills, would anything change?  It seems to me that this would be functionally equivalent to what we have today.

So, really, interest on reserves, the QEs, etc. can all be thought of as ways to maintain deposit levels at the banks while the banks accumulate treasuries and agency securities.

This might be all well and good in a crisis (notwithstanding the question of whether the crisis could have been averted in the first place).  But, the natural market demand for mortgage debt will be high as long as our geographical centers of economic opportunity are governed by limited housing.  Since we blamed credit itself for the problem of high home prices, we developed a national policy framework of smashing down the level of mortgages outstanding.  But, we never actually tackled the real source of the problem, which is the Closed Access cities.

So, since 2008, regulatory pressures and federal control of the GSEs have been the main tools for suppressing mortgage growth.  So, we have basically diverted 10% of bank assets out of residential investment and into treasuries.  (Did I hear somewhere there was a problem with a savings glut and low interest rates?)

Source
So, since 2007 (2006, really) we have had recession level GDP growth and recession level residential investment.

One other item that I think is of interest here is that real estate lending began to drop at the banks by early 2007.  When the private securitization market collapsed in summer 2007, real estate lending also stopped growing at banks.

Source
I used to think that the collapse of private securitizations was due to a liquidity crunch and was the cause of subsequent dislocation.  To some extent, it was.  But, I think the chain of events was a little more complicated, and the collapse of private securitizations was actually more of an effect than a cause.  In fact, the drop in mortgage lending was a lagging factor.  Total mortgages outstanding didn't peak until 2Q 2008.  By then, home prices had collapsed by more than 10% and homeowners had lost trillions in home equity.

The drop in mortgages didn't happen because there was a lack of money.  It happened because there was a lack of faith in home equity.  The first collapse was in home equity.  More than $1 trillion of that collapse happened before 1Q 2007 when home prices were still basically at their peak.  This was a shift of existing homeowners out of the market.

It's ironic that every guy at the end of the bar knows that what happened is that all those silly, greedy speculators thought that home prices could never fall, and that they kept pushing prices up when the drop was inevitable.  On the contrary, for more than a year before prices collapsed, home owners were fleeing the market in anticipation of a price collapse.  When the collapse in mortgages finally happened, it wasn't because mortgage originators ran out of suckers.  It was because rating agencies, lenders, and qualified home buyers all became convinced that housing was doomed to collapse.  The Fed's response was, "Yeah.  Probably right.  We'll be letting that happen."  And the collapse was the last thing to happen.

So, originations from both the banks and from private securitizations were drying up by early 2007, because nobody was willing to be a lender or a borrower with those expectations.  When private MBS securities collapsed later in the year, it was because of expected future defaults that were presumed to be inevitable because of those expectations.  The Fed confirmed that they would enforce, or at least allow, those expectations to come to fruition.  Because we all just knew that supporting anything short of a complete collapse would be irresponsible.  We have come to conceive of systematic instability as a public good.

Notice that graph of GDP and residential investment.  They collapsed together by mid 2006, even though mortgage growth was still healthy.  The first collapse was a real collapse - an unnecessary real collapse.  And the real collapse is what eventually led to the nominal collapse.

I happen to think that the market monetarists are on to something, and that nominal instability can create instability in real production.  But, in this particular case, the real economy was the first mover.  This was still a monetary phenomenon in many respects.  Bernanke himself takes "credit" for that drop in residential investment because the Fed had raised rates into mid 2006, which we might call the interest rate channel of monetary policy.  Then, the Fed clearly signaled in 2007 and 2008 that they were willing to watch the bottom drop out of housing.  That is what we might call the expectations channel.  And even after the disaster of September 2008 when the Fed finally committed to stability - still over the objections of many flavors of liquidationists - most of the collapse in the low end of the housing market was imposed through punitive and erroneous regulatory policies.

But, all that being said, much of the damage might have been avoided with a simple NGDP growth peg.  If the Fed had had an NGDP growth target, they would have had to support the nominal economy despite themselves.  It's possible, I think, that the effect this would have had on expectations might have been enough to prevent the collapse in home price expectations that had caused the initial collapse in home equity to begin with.

Source
Now, the lack of housing is the main source of inflation, so we could functionally cut inflation by encouraging investment or we could dysfunctionally cut inflation by taking away money.  In a regime of excess reserves, I think this means cutting deposit growth.  That seems to be happening since the first rate hike.  Housing starts seem to be leveling off while shelter inflation increases while non-shelter inflation is falling back toward 1%.  Will it continue now that we have had another rate hike, or will the surprising financial optimism that has accompanied the Trump election overcome it?

19 comments:

  1. Kevin, the only thing I would have expressed differently is that an NGDP growth level target might have been less subject to political pressure, than simply a growth target. For instance, consider that the Fed might have assumed in this instance, growth potential to be less likely, so changes the growth target it already has, downward. It might not be so easy to do that if they had a growth level target rule, in the first place.

    ReplyDelete
  2. I dont see how an NGDP target is feasible. A mandate is only as firm as it's political support. An NGDP target of 6% would theoretically allow for a period of 0% real growth and 6% inflation. Regardless of whether that were warranted, I believe the mandate would be changed via politics.

    So it's possible that the mandate would never be credible and therefore wouldn't work. It's also possible that the mandate would be credible at first and the economy would adapt (e.g., a lot of business would lever up I believe) - then when the mandate was changed the system would basically implode.

    You might argue that the same thing is possible under the current mandate however: 1) it has a decent amount of flexibility and therefore can be sensitive to politics, 2) it has decades of familiarity behind it, and 3) i would argue high inflation is uniquely capable of producing political volatility.

    ReplyDelete
    Replies
    1. I think your basic intuition here is correct. The Fed would have needed to be more accommodative in 2006-2007 under an NGDP target rule, and it would have probably stabilized the housing market. When the country demands instability, there may not be much you can do. Still, it would be best to have policy inertia supporting stability.

      Regarding economic adaptations, corporations today, in enterprise value terms, have very low leverage. And, I don't think leverage and expectations of stability are particularly correlated. In fact, we should expect long term expectations of stability would draw investors away from debt into equity, because the perceived downside cyclical risks would be lower. This certainly describes the late 1990s, where debt/enterprise values were low, real interest rates were high, and the equity risk premium was relatively low.

      Delete
    2. My view is that if you take volatility out of the business cycle, businesses will lever up. I think this is consistent with how banks behaved in the Greenspan era (lower volatility + some sort of "put" = add volatility). When that quasi-mandate couldn't be achieved the banks imploded.

      Also, keep in mind that under NGDP targeting the Fed would've been tighter starting all the way back in 2004. So it's pretty hard to say how the housing market would have developed (but I agree, it might have been smoother overall).

      Delete
    3. Well, these issues are at the heart of my project. Leverage was from real estate financing, not from business leverage. Mortgage growth was due to a lack of supply, not risk-taking. NGDP targeting might have led to somewhat less nominal economic growth in 2004, but the bust wasn't inevitable, and it has nothing to do with brief NGDP growth of 6% or 7%, which is quite moderate by any long-term standard.

      These 198 posts are basically an empirical investigation that reaches these conclusions.

      Delete
  3. One quibble. Treasuries with a 6 month term now pay 61 bps. IOR is 75 bps for overnight money with no risk of principal loss. This is crony capitalism - ie, I can't get 75 bps and you can't, but a select class (banks) can.

    ReplyDelete
    Replies
    1. I suppose. But, in a competitive banking environment, deposit rates should equilibrate to a competitive return on capital, so I think, for the most part, this could be considered a subsidy to savers in general. Banking is certainly less competitive in terms of new entrants, etc., since the recession and Dodd Frank, but rates on deposits should still be competitive.

      Delete
    2. Good point. Note though that deposits can be very sticky. I'd bet that the average rate paid on deposits moved less than 10bps over the 12 months following the first 25bps rate increase.

      Delete
  4. Another great post.

    By the way, this post ties in with the commentary of Adair Turner that 80% of commercial bank lending is on extant real estate assets, which are inflating in value.

    Why are extant real estate values inflating? Supply and demand, and supply is cramped.

    I won't even mention the endogenous creation of money, but it is a topic worth pondering. Okay, I will mention it and I think Kevin Erdmann should do a post on the endogenous creation of money.

    One can read the macroeconomic blogs and see endless commentary on the minimum wage, immigration or free trade. The topic of property zoning remains all but verboten.

    ReplyDelete
  5. Another great post.

    By the way, this post ties in with the commentary of Adair Turner that 80% of commercial bank lending is on extant real estate assets, which are inflating in value.

    Why are extant real estate values inflating? Supply and demand, and supply is cramped.

    I won't even mention the endogenous creation of money, but it is a topic worth pondering. Okay, I will mention it and I think Kevin Erdmann should do a post on the endogenous creation of money.

    One can read the macroeconomic blogs and see endless commentary on the minimum wage, immigration or free trade. The topic of property zoning remains all but verboten.

    ReplyDelete
    Replies
    1. I think the endogenous money aspect is interesting. I think we might say there is monetary offset there, too. I think over the past 20 years, what we have been dealing with is monetary offset that is more than 1:1. It more than offset fiscal stimulus in 2008 and in 2014. It more than offset credit stimulus in 2006-2007.

      Delete
    2. Correction - in 2014, it more than offset fiscal austerity.

      Delete
  6. I think you're right.

    But what I'm pondering is the endogenous creation of money the spigot can be turned off nearly overnight, if banks get spooked that property values are falling.

    The fiscal response is necessarily slow, awaiting the next annual budget.

    The Fed's usual is lower interest rates, but it is very reluctant to go to quantitative easing, let alone helicopter drops.

    So let us assume that real estate lending quickly contracts in 2017. Federal spending will hardly budge--- there will be some automatic stabilizers---and the Fed will be flat-footed.

    From this perspective, it will be the reduction of endogenous money creation that causes a recession.

    That is a prospect in 2017 or perhaps the next year.

    ReplyDelete
    Replies
    1. Have you ever seen this?
      https://sffed-education.org/chairthefed/
      The funny thing is that the way to win is to make a bold move when shocks hit. Like if there is a demand shock, lower rates by 4% immediately.

      Delete
    2. That's exactly right! Yet it is 180 degrees from current thinking. The Fed is more concerned with each move being "communicated" and gradual than with producing a stable NGDP path. The Fed is more concerned with having a reputation for being a perfect forecaster. They say they are data dependent, but the history of rate moves shows that they absolutely aren't. A data dependent Fed should have a history of rate moves that are uncorrelated and of varying sizes (better to have a few zero moves than up 10 bps, down 12 bps, up 8, etc.). And heaven help us if the Fed ever cut rates by 205 bps at once (like it should have on 9/15/2008).

      Delete
    3. In their defense, it is a lot easier to make bold moves in a simulation where variables are definitive than it is in real time.

      Delete
    4. Sigh. I know you're right.

      Delete
  7. Kevin and bill: yes there is tremendous conservatism in monetary policy.

    I suspect the Fed should go to big helicopter drops (my favorite, helicopter drops into the Social Security fund to finance a tax holiday) in Q1 of any downturn. Or earlier. Perhaps there should be a "Taylor Rule" dictating helicopter drops when certain conditions are met.

    Fat chance.

    ReplyDelete