{"id":20,"date":"2025-07-05T17:06:19","date_gmt":"2025-07-05T17:06:19","guid":{"rendered":"http:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/?page_id=20"},"modified":"2025-07-05T17:06:19","modified_gmt":"2025-07-05T17:06:19","slug":"iivars","status":"publish","type":"page","link":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/iivars\/","title":{"rendered":"Institutional Investor Classification Data: Variable Definitions"},"content":{"rendered":"\n

<\/p>\n\n\n\n

<\/a>

Spectrum manager number\n<\/strong><\/p><\/p>\n\n\n\n

<\/td>\nThis is the fund manager number used by the Spectrum database, which WRDS labels as \u201cmgrno\u201d.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nManager number version\n<\/strong><\/p>\n\n\n\n

<\/td>\nSpectrum recycles manager numbers. I assign a new version number every \ntime there is more than a two quarter break in holdings information for a\n manager number. I use this information to help update the permanent \nkey (see below).\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nPermanent key\n<\/strong><\/p>\n\n\n\n

<\/td>\nFor a few years, Spectrum assigned fund managers a permanent key to \nallow researchers to merge the 13F data with the mutual fund data. \nSpectrum has discontinued this data item, but I have used it as a basis \nto tie together the holdings history for fund managers that change \nmanager numbers. My RAs and I have created permanent keys for managers \nthat did not have them assigned by Spectrum and have updated them over \ntime. Given this procedure, I cannot guarantee that these histories are\n 100% accurate. Please contact me if you find any errors in this \nassignment and I will update the dataset.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nYear\n<\/strong><\/p>\n\n\n\n

<\/td>\nThis is the calendar year of the classification. In classifying \ninstitutions, I compute averages across the four holdings reports for \neach calendar year.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nType\n<\/strong><\/p>\n\n\n\n

<\/td>\nThis is the legal type of the institutional investor, using the following code:

\n\nBNK = bank trust (Spectrum type code 1)
\nINS = insurance company (2)
\nINV = investment company (3)
\nIIA = independent investment advisor (4)
\nCPS = corporate (private) pension fund (5)
\nPPS = public pension fund (5)
\nUFE = university and foundation endowments (5)
\nMSC = miscellaneous (5)

\n\nAs noted on the WRDS website, the type code variable on Spectrum is not \nreliable after 1998. I have taken the \u201creliable\u201d Spectrum type codes \nand carried them forward in time for institutions still in existence \nafter 1998. For new institutions, my RAs and I have attempted to assign\n a type code based on searches for information about the fund manager. \nIn doing so, we have not attempted to distinguish between type code 3 \nand type code 4. In my research, I merge these two types into one group.\n In addition, we have taken the type code 5 group (\u201cother\u201d) and \nattempted to determine whether the fund manager was a private pension, \npublic pension, or an endowment. All other institutions were classified\n as \u201cmiscellaneous\u201d. Given this procedure, I cannot guarantee that these\n histories are 100% accurate. Please contact me if you find any errors \nin this assignment and I will update the dataset.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

<\/a> Transient\/Quasi-indexer\/Dedicated classification <\/strong><\/p>\n\n\n\n

<\/td>\nThis classification uses the following code:

\n\nDED = dedicated
\nQIX = quasi-indexer
\nTRA = transient

\n\nThis classification is based on the one used in Bushee (2001) and Bushee\n and Noe (2000). Note that I changed my classification scheme after the\n Bushee (1998) paper by dropping the momentum variables to allow it to \nbe used in more general situations. I extended those classifications by\n applying the factor loadings reported in those papers to the more \nrecent data to compute factor scores, which I used to add the new data \nto the existing clusters. If a fund has no classification for a given \nyear, it means that some of the data was missing, the fund has a small \nportfolio (i.e., fewer than four stocks that have available CRSP and \nCompustat data), or the fund has not been listed on Spectrum for two \nyears. As a consequence of these restrictions, not all of the \ninstitutions have been classified using this approach. There are a \nlarge number of unclassified institutions in the past five years, \nprimarily due to growth in the number of new fund managers. For a \npotential solution to this problem, please see the data item below.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nPermanent Transient\/Quasi-indexer\/Dedicated classification\n<\/strong><\/p>\n\n\n\n

<\/td>\nUsing the permanent key variable we created, I find the modal \nclassification for each permanent key and assign that classification to \neach year of data for the fund manager.

\n\nThis approach potentially helps solve the following problems:
\n\nSome fund managers, especially those that manage a large number of \nmutual funds, have classifications that frequently shift across years \n(the first-order autocorrelation in the classifications is generally \naround 0.8). This approach fixes the classification across time to the \nmost common classification.

\n\nBecause fund manager classifications can change, it is \ngenerally not a good idea to compute changes in holdings for a given \ntype as the difference between the percentage ownership by the type in \none period and percentage ownership by the type in a prior period. Such\n a measure would consider a fund that does not change its holdings, but \ndoes change its type, to be a change in holdings by the type. The \npermanent approach eliminates this potential problem.

\n\nAs noted above, in the first two years of a fund’s history, I \ncannot compute its classification. This approach allows me to fill in \nsome of this missing data.\n\n\nThis approach has the following drawbacks:
\n\nSome fund managers likely do change their trading orientation over time. Using the modal classification obscures such changes.

\n\nThe modal classification is based on my permanent key, which \nmay not be 100% accurate. A potentially safer way to compute this \nvariable would be to take the modal classification across manager number\n and manager version number. Note: you do not want to compute the mode \nby manager number alone as this is recycled and may combine different \nmanagers in the measure.\n\nI have tended to use this modal classification scheme in my work, but you should make your own call based on this trade-off.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nInvestment style classification\n<\/strong><\/p>\n\n\n\n

<\/td>\nIn Abarbanell, Bushee, Raedy (2003), we classified institutions based on\n investment styles or preferences for firm size and growth, using the \nfollowing code:

\n\nLVA = Large Value style
\nLGR = Large Growth style
\nSVA = Small Value style
\nSGR = Small Growth style

\n\nAs discussed in that paper, this classification stems from a cluster \nanalysis on a firm size factor and a value\/growth factor. Please see \nthe paper for more details.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nPermanent Investment style classification\n<\/strong><\/p>\n\n\n\n

<\/td>\nThis classification is based on the modal classification for each permanent key. See above for more details.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nGrowth style classification\n<\/strong><\/p>\n\n\n\n

<\/td>\nIn Bushee and Goodman (2007), we classified institutions based on \npreferences for growth or value firms, using the following code:

\n\nGRO = Growth style
\nVAL = Value style
\nG&I = Growth & Income style (middle group)

\n\nAs discussed in that paper, this classification stems from a cluster \nanalysis on a value\/growth factor. Please see the paper for more \ndetails.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nPermanent Growth style classification \n<\/strong><\/p>\n\n\n\n

<\/td>\nThis classification is based on the modal classification for each permanent key. See above for more details.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\nTax-sensitivity classification\n<\/strong><\/p>\n\n\n\n

<\/td>\nIn Blouin, Bushee, and Sikes (2017), we classified institutions based on\n sensitivity to capital gains taxes, using the following code:

\n\nTII = Tax-insensitive institutional investor
\nTSI = Tax-sensitive institutional investor

\n\n\nAs discussed in that paper, this classification is based on measures of \ntax-motivated trading around calendar-year ends and on other portfolio \ncharacteristics. We first classify institutions as tax-senstive on a \nyearly basis, and then use the mode over the prior five years to \ndetermine the TSI or TII classification. Because of the requirement of \nfive-years of data, this method does not classify institutions until we \nhave five years of data. Also, this method does not classify \ninstitutions prior to 1991 because we chose to compute embedded gains \nand losses from 1980 to 1987 before attempting to use them to classify \ninstitutions. Please see the paper for more details.\n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

\n<\/a>\n\n\nExtended tax-sensitivity classification\n<\/strong><\/p>\n\n\n\n

<\/td>\nThis classification makes two modifications to the BBS (2017) approach \nto extend the classification. First, for institution-years with fewer \nthan five years of data, I use as much past data as possible plus some \nfuture data to get five years of data to classifiy the institution. For\n example, for an institution that started in 2001, its 2003 \nclassification would be based on data from 2001, 2002, 2003, 2004, and \n2005. Second, I “smooth” the classification by re-classifying any \none-year blips in the rolling five-year modal classification. For \nexample, if an institution has a time-series like this–TSI, TSI, TII, \nTSI, TSI–I will change the TII to a TSI. There are only 58 \ninstitution-years that were smoothed in this way. \n<\/td><\/tr>
<\/td>Return to data page<\/em> <\/a><\/td><\/tr><\/tbody><\/table><\/figure>\n","protected":false},"excerpt":{"rendered":"

Spectrum manager number This is the fund manager number used by the Spectrum database, which WRDS labels as \u201cmgrno\u201d. Return to data page Manager number version Spectrum recycles manager numbers. I assign a new version number every time there is more than a two quarter break in holdings information for a manager number. I use […]<\/p>\n","protected":false},"author":10,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":[],"acf":[],"_links":{"self":[{"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/pages\/20"}],"collection":[{"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/users\/10"}],"replies":[{"embeddable":true,"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/comments?post=20"}],"version-history":[{"count":0,"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/pages\/20\/revisions"}],"wp:attachment":[{"href":"https:\/\/accounting-faculty.wharton.upenn.edu\/bushee\/wp-json\/wp\/v2\/media?parent=20"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}