Konstantinos Metaxoglou and Aaron Smith, "State Prices of Conditional Quantiles: New Evidence on Time Variation in the Pricing Kernel", Journal of Applied Econometrics, Vol. 32, No. 1, 2017, pp. 192-217. All data files are ASCII files in DOS format. They are zipped in the file ms-data.zip. Unix/Linux users should use "unzip -a". ************************************************** Risk-neutral Distribution Implied by Option Prices ************************************************** Relevant file(s): logistic_params.csv Variable names: datenum : date in YYYYMMDD format alpha : call dummy omega : mixture weight mu1 : scale parameter for mixture component 1 mu2 : scale parameter for mixture component 2 sigma1 : location paramter mixture component 1 sigma2 : location paramter mixture component 2 Note: As Section 3.4 of the paper discusses, the mixture of the logistic distributions requires CBOE data for S&P500 options, which can be purchased, but cannnot be provided by the authors. However, we do provide the parameters of the logistic distributions for each of the 268 trading dates in our sample in the file listed above. ******* Table 1 ******* Relevant file(s): qreg_vol_cv_and_vix_diff_stata.csv Variable names: datenum : date in YYYYMMDD format ret_co : close-open returns as defined in equation (28) vol_cv : squared root of continuous variation of daily returns (CV) vol_diff : CV-VIX vol_vix_cboe : VIX Note: We used Stata to estimate our quantile regressions. You can replicate our estimates in Panel A of the table using the qreg command in Stata with robust std. errors. The Stata and Matlab codes for the statistics reported in panels B and C are available from the authors upon request. ************** Tables 2 and 3 ************** Relevant file(s): a. reg_EPK_vars_vol_cv_and_vix_diff_stata_series_orig.csv b. reg_EPK_vars_vol_cv_and_vix_diff_stata_series.csv Variable names in file a: datenum : date in YYYYMMDD format epk0010 : dependent variable for 0-10% regressions epk0025 : dependent variable for 0-25% regressions epk0050 : dependent variable for 0-50% regressions epk75100 : dependent variable for 75-100% regressions epk90100 : dependent variable for 90-100% regressions ln_vol_cv_std : Log CV ln_jumpt_diff_std : Log tail jumps ret_ccl_std : Returns sent_cons_std : Consumer sentiment ln_dp_std : Log dividend yield cay_std : CAY ln_unc_idx_sum_std : Log uncertainty index sum dfsp_std : Default spread tmsp_std : Term spread ln_liq_oit_0010_std : Log open interest for 0-10% regressions ln_liq_oit_0025_std : Log open interest for 0-25% regressions ln_liq_oit_0050_std : Log open interest for 0-50% regressions ln_liq_oit_75100_std : Log open interest for 75-100% regressions ln_liq_oit_90100_std : Log open interest for 90-100% regressions ln_liq_dol_0010_std : Log dollar volume for 0-10% regressions ln_liq_dol_0025_std : Log dollar volume for 0-25% regressions ln_liq_dol_0050_std : Log dollar volume for 0-50% regressions ln_liq_dol_75100_std : Log dollar volume for 75-100% regressions ln_liq_dol_90100_std : Log dollar volume for 90-100% regressions Note: The suffix _std indicates that the variable has been standardized to have mean zero and std.deviation of one. The original variables can be found in file b. We have also provided the jumpt variable, which can be used to generate ln_jumpt_diff=ln(( vol_cv+ jumpt)/ vol_cv). The results reported in the table are based on a GLS estimator discussed in section 3.3. of the paper. The Matlab code for the GLS estimator is available from the authors upon request. The data sources are provided in the online Appendix A.6 of the paper. ******** Figure 1 ******** Relevant file(s): fig_epk_vol_cv_and_vix_diff_stata.csv Variable names: y1: bottom quartile y2: top quartile x : log volaility ******** Figure 3 ******** Relevant file(s): fig_options_strikes.csv Variable names: datenum : date in YYYYMMDD format strikes_all0100 : bottom 10% strikes strikes_all0250 : bottom 25% strikes strikes_all0750 : top 25% strikes strikes_all0900 : top 10% strikes ******** Figure 5 ******** Relevant file(s): quantiles_vol_cv_and_vix_diff_stata.csv Variable names: datenum : date in YYYYMMDD format pctile0100 : 10% quantile pctile0250 : 25% quantile pctile0500 : 50% quantile pctile0750 : 75% quantile pctile0900 : 90% quantile lambda0100 : SPOCQ(0,10) lambda0250 : SPOCQ(0,25) lambda0500 : SPOCQ(0,50) lambda0750 : SPOCQ(0,75) lambda0900 : SPOCQ(0,90) ******** Figure 6 ******** Relevant file(s): fig_vol_cv_and_vix_diff_stata_qpk_crisis.csv Variable names: year : year quarter : quarter epk0010 : QPK(0,10) epk0025 : QPK(0,25) ******** Figure 7 ******** Relevant file(s): a. fig_epk_vol_pers_6bins_qreg_vol_cv_and_vix_diff_stata_gls2.csv b. fig_epk_yr_pers_all_6bins_qreg_vol_cv_and_vix_diff_stata_gls2.csv Variable names in file a: period : low volatility (low vol)/ high volatility (high vol) identifier ticks : 1 "0-10%", 2 "10%-25%",...6 "90%-100%" epk_mean : average QPK estimate epk_low : average QPK lower bound (estimate -2 std.errors) epk_upp : average QPK lower bound (estimate +2 std.errors) Variable names in file b: period : 1990-2012 ticks : 1 "0-10%", 2 "10%-25%",...6 "90%-100%" epk_mean : average QPK estimate epk_low : average QPK lower bound (estimate -2 std.errors) epk_upp : average QPK lower bound (estimate +2 std.errors) Please address any questions to: Konstantinos Metaxoglou Department of Economics Carleton University 1125 Colonel By Drive Ottawa ON K1S 5B6 konstantinos.metaxoglou [AT] carleton.ca https://sites.google.com/site/kostasmetaxoglou/