Robert Jarrow and Simon Kwok, "Inferring Financial Bubbles from Option 
Data", Journal of Applied Econometrics, Vol. 36, No. 7, 2021, pp. 
1013-1046.

All files are ASCII files in DOS format. They are zipped in the file
jk-files.zip. Unix/Linux users should use "unzip -a".

The main codes are:	
 optiondata_prep.sas (SAS code, to clean and prepare data for bubble inference)
 bub12_run.m         (Matlab code, to evaluate simple and naive bounds)
 bubcp_run.m         (Matlab code, to evaluate interval estimates)

To run bub12_run.m, the following subroutines are required:
 bub12.m
 bub12_trade.m

To run bubcp_run.m, the following subroutines are required:
 bubcp.m
 bubcp_trade.m 
 anticonv_put.m
 anticonv_call.m
 lpoly.m
 nw_cov.m

Please refer to the comments in individual subroutines for specific usage. 

Four input datasets are required for running the SAS code. All
datasets are SAS data file (.sas7bdat). Input dataset S1, S2 and S4
are obtained from OptionMetrics and Wharton Research Data Services
(WRDS). Input dataset S3 is obtained from Robert Shiller's website.
 
Input dataset S1: callbase.sas7bdat 
 source:           OptionMetrics in WRDS
 fields:
  date             date of transaction (e.g., 04JAN1996)
  exdate           expiry date (e.g., 04JAN1996)
  cp_flag          call/put indicator (C for call; P for put)
  strike_price     strike price (in days)
  best_bid         best bid price
  best_offer       best ask price
  volume           traded volume
  open_interest    open interest
  impl_volatility  implied volatility
  delta            option delta

Input dataset S2: snp500.sas7bdat
 source:           WRDS
 fields:
  date             date
  snp500           S&P 500 index

Input dataset S3: div_shiller.sas7bdat
 source:           http://www.econ.yale.edu/~shiller/data.htm
 fields:
  year             year
  month            month
  dividend         S&P dividend yield 

Input dataset S4: trate_1mo3mo.sas7bdat
 source:           WRDS
 fields:
  date             date
  trate1mo         1-month Treasury bill rate
  trate3mo         3-month Treasury bill rate


Two input datasets are required for running the Matlab codes. Both
datasets are comma-delimited csv files and are generated by the sas
code optiondata_prep.sas. A small extract of Input dataset C1 and a
sample of Input dataset C2 are included for formatting references.

Input dataset C1: allopt_01Janyyyyto31Decyyyy.csv (where yyyy ranges from
1996 to 2015)
 source:     generated by optiondata_prep.sas
 fields:
  date       transaction date (e.g., 04JAN1996)
  cp_flag    call/put indicator (C for call; P for put)
  exdate     expiration date  (e.g., 04JAN1996)
  tau        time-to-expiration (in days)
  strike     strike price
  snp        ex-dividend S&P 500 index closing value
  dy_ma      dividend yield (not used)
  tr         one-month T-bill rate (from Fed)
  money      log-moneyness (not used)
  callprice  call/put price
  volume     trading volume (in number of contracts)
  iv         implied volatility (not used)
  delta      Greek letter delta (not used)

 sorted by:  
  date       ascending order
  cp_flag    C, followed by P
  tau        ascending order
  strike     ascending order

Input dataset C2: allopt_01Janyyyyto31Decyyyy_count.csv (where yyyy
ranges from 1996 to 2015)
 source:     generated by optiondata_prep.sas
 fields:
  date       transaction date (e.g., 04JAN1996)
  _TEMG001   daily count of different option contracts with different tau,
             strike, and types (calls & puts).

 sorted by:  
  date       ascending order

Any question should be sent to the corresponding author:

  Simon Kwok  simon.kwok[AT]sydney.edu.au