The first six characters are known as the "base" (or "CUSIP-6"), and uniquely identify the issuer. Issuer codes are assigned alphabetically from a series that includes deliberate built-in "gaps" for future expansion. The last three characters of the issuer code can be letters, in order to provide more room for expansion.
Issuer numbers 990 to 999 and 99A to 99Z in each group of 1,000 numbers are reserved for internal use. This permits a user to assign an issuer number to any issuer which might be relevant to his holdings but which does not qualify for coverage under the CUSIP numbering system. Other issuer numbers (990000 to 999999 and 99000A to 99999Z) are also reserved for the user so that they may be assigned to non-security assets or to number miscellaneous internal assets
The 7th and 8th digit identify the exact issue, the format being dependent on the type of security. In general, numbers are used for equities and letters are used for fixed income. For commercial paper the first issue character is generated by taking the letter code of the maturity month, the second issue character is the day of the maturity date, with letters used for numbers over 9. The first security issued by any particular issuer is numbered "10". Newer issues are numbered by adding ten to the last used number up to 80, at which point the next issue is "88" and then goes down by tens. The issue number "01" is used to label all options on equities from that issuer.
Fixed income issues are labeled using a similar fashion, but due to there being so many of them they use letters instead of digits. The first issue is labeled "AA", the next "A2", then "2A" and onto "A3". To avoid confusion, the letters I and O are not used since they might be mistaken for the digits 1 and 0.
The 9th digit is an automatically generated check digit using the "Modulus 10 Double Add Double" techniqueInside the CUSIP numbering system. To calculate the check digit every second digit is multiplied by two. Letters are converted to numbers by adding their ordinal position in the alphabet to 9, such that A = 10 and M = 22. The resulting string of digits (numbers greater than 10 becoming two separate digits) are added up. The ten's-complement of the last number is the check digit. In other words, the sum of the digits, including the check-digit, is a multiple of 10. Some clearing bodies ignore or truncate the last digit.
NOTE: In addition to digits 0-9 and letters A-Z, three characters *, @, and # may also be used in some rare cases. When calculating the check digit for such issues, use the numeric values of * = 36, @ = 37, and # = 38.
CINS adds a single country code letter to be the beginning of an otherwise similar CUSIP. These are not standard country codes, for instance Norway is "R". A table of the country codes appears on the CUSIP web site.
Examples
Apple Inc: '''037833100'''
The low-numbered issuer number, 037833, is a side effect of the company name starting with the letter "A". Their stock is the first issue they released, and is thus numbered "10".
The check digit is calculated by first collecting the even and odd digits, converting any letters to numbers if need be (not in this case):
(0, 7, 3, 1), (3, 8, 3, 0)
The second set is then multiplied by two:
(6, 16, 6, 0)
The individual digits are then added together:
(0 + 7 + 3 + 1) + (6 + (1 + 6) + 6 + 0) = 30
The check digit is the ten's complement of the last digit of 30, which is 0, so the check digit is 0. The ten's complement of a digit is the result of subtracting that digit from 10, except that the ten's complement of 0 is 0.
Wal-Mart: '''931142103'''
Similar to the Apple example, but with a name appearing near the end of the dictionary. The check digit is (9 + 1 + 4 + 1) + (6 + 2 + 4 + 0) = 27, the ten's complement of 7 is 3.
US Treasury Note: '''912827XN7'''
Here we have a combination of letters and numbers. The letters A to Z have the values 10 to 35, so X is 33 and N is 23. The check digit is (9 + 2 + 2 + (3 + 3)) + (2 + (1 + 6) + (1 + 4) + (4 + 6)) = 43, the ten's complement of 3 is 7.
As code
===In Python (programming language)|Python===
import string
- constants
cusipchars = string.digits + string.ascii_uppercase + "*@#"
cusip2value = dict((ch, n) for n,ch in enumerate(cusipchars))
cusipweights = 1,2,1,2,1,2,1,2
def checksum(cusip):
digits = [(w*cusip2valuech) for w,ch in zip(cusipweights, cusip)]
cs = sum(for x in digits) % 10
return str((10-cs)%10)
def check(cusip):
return len(for ch in cusip if ch in cusipchars)
