The estimates used in my paper Trading Costs and Returns for US Equities: Estimating Effective Cost from Daily Data are available for noncommercial research purposes. These data substantially replace those associated with the earlier (February, 2005) version of the paper, but I am maintaining the older data for researchers who have already built these estimates into their analyses. Datasets and programs associated February 2005 draft

If you use these data, please drop me an email at jhasbrou@stern.nyu.edu. I'll notify you of any updates. I would, as well, be grateful to know about any problems or other things you notice about the data. I make this dataset available on a "best efforts" basis. I believe, but cannot ensure, that the calculations are correct.

The new data include estimates based on two models described in the paper:

*Basic Market-Adjusted Model*. This version is very similar to that of the earlier paper. I've added a market factor to the price change regressions, so the liquidity dataset includes an estimate of beta. Estimates based on this model are flagged "BMA".*The Latent Common Factor Model*. This model allows for common stochastic variation in effective costs. Estimates based on this model are flagged "LCF".

Datasets with annual liquidity estimates by firm

- LiqEstimatesAug2006 (SAS Dataset) contains the 1926-2005 liquidity estimates (from CRSP, BMA, LCF, etc.). One record for each permno/year. Not all of the liquidity measures in the file ended up being reported in the final version of the paper.
- CompSampleAug2006 (SAS Dataset) liquidity estimates (from TAQ and CRSP) for the comparison sample (approx 300 firms per year, 1993-2005). These are annual estimates. Monthly and daily estimates are available for the 2005 version of the paper.

Datasets with estimates of common factor for effective costs.

The LCF model also generates an (estimated) common factor for effective costs (denoted zt in the paper). The series is available from 1926-2005 at a daily, weekly and monthly frequency. Beginning with 1993, the data include the cross-firm average effective cost for the comparison sample. The monthly dataset contains innovations in zt based on an AR(1) model. These are denoted "zu". Important usage note: As described in the paper, the common factor estimates are normalized within each year. (You can't directly compare magnitudes computed in different years.)

Details follow.

Variable descriptions: LiqEstimatesAug2006

Variable | Description/Source | |

Identifying or descriptive variables | Permno | CRSP Permno |

Year | End of the year used for estimating the liqudity proxies (1926-2005) | |

Ticker | CRSP | |

Exchcd | CRSP | |

Comnam | CRSP | |

CRSP/Gibbs liquidity measures | c_BMA | Gibbs estimate of c, from Basic Market-Adjusted model |

beta_BMA | ... beta ... | |

sdu_BMA | ... std. dev. of u ... | |

gamma0_LCF | Gibbs estimate of gamma0, from Latent Common Factor model | |

gamma1_LCF | ... gamma1 ... | |

beta_LCF | ... beta ... | |

sdu_LCF | ... std. dev. of u ... | |

Other liquidity measures | I1 | Amihud illiquidity measure, average over year of 1000000*abs(ret)/(abs(prc)*vol); |

I2 | Square root variant of above; average over year of 1000*sqrt(abs(ret)/(abs(prc)*vol)) | |

NI | Number of days used in calculation of I1 and I2 | |

L1 | Amivest liquidity ratio, average over year of .000001*abs(prc)*vol/abs(ret) | |

L2 | Square root variant of above; average over year of .001*sqrt(abs(prc)*vol/abs(ret)) | |

NL | Number of days used in calculation of L1 and L2 | |

PropZeroRet | Proportion of days with zero returns (relative to the number of days with non-missing returns) | |

cmz | Moment estimate of c, based on all reported prices, including quote midpoints | |

cmzAlt | Moment estimate of c, excluding quote midpoints | |

PSGamma | Pastor-Stambaugh gamma | |

Supplementary variables | Price | End of year price per share |

MktCap | End of year market capitalization | |

MERankFF | Market capitalization rank (1-20) based on Fama-French NYSE breakpoints | |

nPrc | Number of non-missing price observations in the year | |

nMid | Number of prices reported as quote midpoints (i.e., <0 on CRSP) |

Variable descriptions: CompSampleAug2006

Variable | Description/Source | |

Identifying or descriptive variables | Permno | CRSP Permno |

Year | End of the year used for estimating the liqudity proxies (1993-2005) | |

Ticker | CRSP | |

Exchcd | CRSP | |

Comnam | CRSP | |

CRSP/Gibbs liquidity measures | c_BMA | Gibbs estimate of c, from Basic Market-Adjusted model |

Other CRSP liquidity measures | I1 | Amihud illiquidity measure, average over year of 1000000*abs(ret)/(abs(prc)*vol); |

I2 | Square root variant of above; average over year of 1000*sqrt(abs(ret)/(abs(prc)*vol)) | |

NI | Number of days used in calculation of I1 and I2 | |

L1 | Amivest liquidity ratio, average over year of .000001*abs(prc)*vol/abs(ret) | |

L2 | Square root variant of above; average over year of .001*sqrt(abs(prc)*vol/abs(ret)) | |

NL | Number of days used in calculation of L1 and L2 | |

PropZeroRet | Proportion of days with zero returns (relative to the number of days with non-missing returns) | |

cmz | Moment estimate of c, based on all reported prices, including quote midpoints | |

cmzAlt | Moment estimate of c, excluding quote midpoints | |

PSGamma | Pastor-Stambaugh gamma | |

TAQ liquidity measures | c | Effective cost, trade-weighted average |

lambda_tSignSqrtDVol | Price impact coefficient | |

Supplementary variables | Price | End of year price per share |

MktCap | End of year market capitalization |