Tuesday, March 21, 2017

Leaks

I got to spend a year (July 2015-June 2016) working as a Senior Advisor at the Department of Housing and Urban Development.  I wasn't particularly high up in the pecking order, but I got to work with a number of people from HUD, Treasury and the White House who were.

Here's the thing (sorry for using a Sorkinism): all of these people--every, single, one--liked President Obama; were proud to work for President Obama.

Did they think he was perfect, or always made the correct decision?  Of course not.  But I have to say, in all the meetings in which I got to participate, there was reasoned deliberation, and comportment really mattered.  The atmosphere was professional and respectful.  And I think because of this, no one wanted to embarrass the president--certainly no one wanted to go out of his or her way to damage the president.

Just saying...

Sunday, February 19, 2017

Troika

Troika is the forecasting process for the federal government; it is called Troika because it is a joint project of Treasury, the Office of Management and Budget, and the Council of Economics Advisors.

Last year, while I was a Senior Advisor at HUD, I got a peek into Troika; I was invited to participate in a meeting to offer a perspective on the US housing market.  I am not going to say much about the details of the meeting, except that the Troika process is very much based on econometric modeling, that the modelers are really good at their jobs, and that the debates about the models are exactly the sorts of the debates one would wish government officials to have.  To give one example, at the meeting I attended, there was a debate about a parameter estimate.

The debate arose from the following conditions: suppose economic theory implies that a parameter b = b*. The estimated parameter b = b*+a.  The standard error of the parameter is 2a.  The debate was whether the forecast should be based on b*, or b*+a.  Needless to say, on could make reasonable arguments either way (showing that no matter how good a modeler is, she needs to rely on judgment at some point).

This is how government forecasting has been done--empirically, rigorously, and without an agenda. It saddens me to think that this is under attack.





Monday, January 23, 2017

Why I like what Quicken is doing

There is a piece in the New York Times from yesterday that sort of implies that Quicken Loan's rapid rate of growth (they are now the second largest FHA lender after Wells-Fargo) must mean the lender is up to no good.  But unlike Countrywide and WAMU, whose growth in the previous decade was the result of unsound lending practices, Quicken has developed a business model that, in my view, can result in lending that is sounder than traditional lending, expanded access to credit, and reduce loan applicant frustration.

I suppose I should say here that I have no financial interest in Quicken (it is closely held, so I couldn't even if I wanted to). I met its CEO, Bill Emerson, once, and spoke to him on the phone once, and we had nice conversations, but I would hardly say we now each other socially (for all I know he wouldn't even remember talking with me).  I have also had cordial conversations with other Quicken executives, which I think gave me a little insight into how the company operates.

Yesterday's piece notes that Quicken is viewed more as a technology company than a mortgage company, but it doesn't expand on what that means.  Here is what I think it means--it uses technology to improve quality control and compliance, and to do its own underwriting.  Specifically, when a potential borrower applies for a loan using the Rocket Mortgage app, she gives permission to Quicken to download financial information from the IRS, bank accounts, and other accounts. Because the information flows directly from the source, loan applications are complete and accurate, and hence comply with an important requirement for FHA loans.

The information is then run through the FHA TOTAL scorecard, where it receives an accept or refer (a refer means that for a loan to be approved, it can be manually underwritten, but is often rejected) and through Quicken's own underwriting algorithm.  The executives I spoke with at Quicken told me that the algorithm is updated frequently.  My guess--I don't know this for a fact--is that the algorithm's foundation is the sort of regression that I discussed in a previous post.

As noted in that post, statistically based algorithms can both improve access to credit and the performance of loans.  As the pool of potential borrowers becomes less and less like previous borrowers (in terms of source of income, credit behavior, family participation in loan repayment, etc.), using data to continuously improve and refine underwriting will be important for sustaining the mortgage market.  To the extent that Quicken is doing this, it makes the mortgage market better.

This is not to say it would be good for Quicken ultimately to dominate the market (such dominance is never healthy).  It would be nice to see fast followers of Quicken to enter the market.  But I suspect the reason the company has grown so rapidly is that it has built a better mousetrap.


Monday, January 09, 2017

Danny Ben-Shahar leads me to reflect on whether data should be treated as a public good

Danny Ben-Shahar gave a really nice paper (co-authored with Roni Golan) at the ASSA meetings yesterday about a natural experiment in the impact of information provision on price dispersion.  I want to talk about it, but first a little background.

Price dispersion is an ingredient in understanding whether markets are efficient.  When prices for the same good vary (for reasons other than, say, transport costs or convenience), it means that consumers lack the information necessary to make optimal decisions, and the economy suffers from deadweight loss as a result.

Houses have lots of measured price dispersion, even after controlling for physical characteristics. Think about a regression for a housing market, where

HP = XB + h+e

where HP is a vector of house prices, and X is a matrix of house characteristics.  The residual h+e has two components—unmeasured house characteristics, h, and an error term, e, which reflects “mistakes” consumers of houses make, perhaps because of an absence of information.  The h might reflect something like the quality of view, or absence of noise, etc.

When we run this regression, we can compute a variance of the regression residuals.  Because we can only observe h+e, we cannot know if this variance is the result of unobserved house characteristics, or of consumer errors.  But if h remains fixed, and there is an information shock that reduces consumer errors,  e will get smaller, and so will the regression variance.

Here is where Danny’s paper comes in.  In April 2010, authorities in Israel began publishing on-line information about house transactions, and in October 2010, they launched a “user-friendly web site.”  (Details may be found in the paper).  The paper measures the change in measured price dispersion before and after the information was publicly available, and, at minimum, found reductions in dispersion of about 17 percent. The paper takes pains to make sure their result isn’t a function of some shock that happened simultaneously to the release of the information.  For example, they show that price dispersion fell less in neighborhoods with well-educated people.  This could either reflect that (1) well educated people were better informed about housing markets to begin with, and so got less benefit from the new information or (2) that a greater share of the residuals in well-educated neighborhoods comes from non-measured house characteristics.[i]  In either event, the result is consistent with the idea that the information shock is what contributed to the decline in measured price dispersion.

So more information really does seem to produce a more efficient housing market.  The policy implication may be that data, in general, should be a public good.  Data meet half of Musgrave’s definition of a public good—they are non-rival (one person’s use of a data-set does not detract from another person’s use).  And while data are excludable (services such as CoreLogic show this to be true), their creation produces a classical fixed-cost marginal-cost problem.  The fixed cost of producing a good dataset is very large; once it is created, the marginal cost of providing the data to users is very low.  This suggests that the efficient price of data should be very low. 

Currently, data services have something like natural monopolies, with long downward sloping average cost curves.  Theory says that this means they are setting prices such that marginal revenue equals marginal costs, instead of setting price equal to marginal cost.  All this implies that data are underprovided.  Danny and Roni’s work shows that this under-provision has meaningful consequences for the broader economy.




-->



[i] BTW, this second interpretation is mine (I don’t want the authors on the hook for it if they disagree).