Housing and Liquidity 
by C. He, R. Wright and Y. Zhu.

Original data used to generate Figure 1 and constructing the data used in the codes

Data:

*Home Mortgages;M$; 
*Home Equity Loans;M$;
Source: Flow of Funds by the Federal Reserve Bank
Note: yearly data

*Home Market Value;B$; 
Source: Land and Property Values in the U.S.
Website: http://www.lincolninst.edu/subcenters/land-values
Note: quarterly data, use third quarter of each year as the data for that year

*Purchase Only House Price Index;
Source: FHFA
Note: Quarterly data, use third quarter of each year as the data for that year

*CPI All Item
Source: BLS
Note: Monthly data, use July as the data for each year

*Owners' Equivalent Rent of Residence; annual data
Source: BLS

*Real GDP;B$2009
*Nominal GDP;B$
*Residential Investment;B$ 
*Quantity Indexes_Investment in Residential Fixed Assets 
Source: BEA
Note: All are annual data

D_new3.mat and Dlong.mat are series for loan-to-equity ratio as 
shown in the file "data.xlsx"

Codes:

In all the codes, c, AA, gama and kapa corresponds to A_{y}, AA, A_{!}, � andA_{h}, respectively.

The following codes are matlab script codes which can be run in matlab except psi_tran_final_new.m.
"  ss_price_D_fix_H.m:  Compute the steady state housing price � and KM consumptin y as a function of D� . This code is used to construct Figure 2.
  

"  Dynamics.m: Contruct the phase diagram as in figure 3a and figure 3b. It also includes construction of two-cycles under the Strategic Bargaining and Kalai Bargaining.
  

"  realistic_dynamics.m: construct pricing paths in which housing price increases for several periods and then drops. Two constructions are included here:

1. A deterministic path in which housing price increases at a rate slower than r=1/beta-1

2. A realization of a nonstationary sunspots equilibrium in which housing price increases at a rate 9%.
The second part of the code computes the transition probability of the nonstationary sunspots at each point of time.
The output prob is the probability of housing price increasing by 9% in the equilibrium at each period.
  

"  dynamic_oneshot.m: Construct and plot the dynamics of the one shot change in D1 as in Experiment 1.
  

"  psi_tran_final_new.m: This function is used to conduct the calibration as described in the paper. 
It calls function search_kapa_fun_new.m which calls function calibration_new.m and markup.m to match markup in KM market. 
The input of psi_tran_final_new.m is a vector of par=(A_{y},A_{!},�), the output of the function is the difference between the model and data as described in the paper. We calibrate the model by choosing par=(A_{y},A_{!},�) to minimize the difference. 
One can choose an initial guess par0 and use the following command to compute the calibrated parameter.

par = fminsearch(@(par) psi_tran_final_new(par),par0)
  

"  transition2_bargainning_final.m: Construct the Experiment 2a
  

"  transition3_bargainning_final.m: Construct the Experiment 2b
 

"  rational_expectation.m: Construct the path under rational expectation as in Experiment 3
  

"  Monetary.m: Compute the equilibrium consumption of y, housing price �, and the interest rate of borrowing and lending money as a function of norminal interest rate in the economy with DD banks. This script is used to construct Figure 10.


" Tauchen.m: Use Tauchen's method to discretize a markov chain. Used in Experiment 3.

" figure_fullpage.m and figure_halfpage.m are used to format the graphs.