Page 58 - Invited Paper Session (IPS) - Volume 2

IPS179 Wian B.
3. Description of Data
The data used in this paper is the population of all retail interbank EFT
transactions on a daily frequency facilitated by BankservAfrica for the period
5
July 2017 to October 2018. The Structured Query Language (SQL) database is
currently accessed online using QlikView business intelligence software. Same
bank to same bank transactions are not covered. Interbank transactions are
estimated to represent between 70 to 80 per cent of all EFT transactions,
although it is not known how this proportion varies over time.
There are 54 fields that are populated for each EFT interbank transaction,
some of which are only for internal use by BankservAfrica. The Transaction
Value and Transaction Volume variables, for purposes of this paper, only
include finalized transactions. Transactions which are subject to dispute or are
classified as unpaid6 are not included. The Transacting Bank and Transacting
Bank Type fields are typically analyzed together when looking at transactions.
A bank is either the destination bank (referred to as the homing bank) or the
originator (referred to as the sponsoring bank) of the transaction. The User
Code is a unique identifier assigned to the client initiating the transaction.
There are 44 options available for classifying an EFT transaction, with a
small portion of transactions bearing no classification. The top five EFT
transaction reasons, measured in terms of value and volume over the indicated
period, is shown in Figure 1. Credit transfers, which are payments initiated by
a client to transfer some arbitrary amount to another account at another bank,
represented more than 50 per cent (or approximately R8.9 trillion) of the total
transaction value of all EFTs over the measured period (July 2017 to December
2018). Payments to creditors were the second largest category (slightly less
than 20 per cent), followed by salary and pension EFT transactions. On the
volume side, credit transfers remain the dominant driver, while insurance
premiums also feature prominently.

5 The short period made available through the online interface is most likely due to the high
volume of data. 6 Disputed or unpaid transactions represented slightly more than 5 percent of
all transactions in terms of volume and about 0.5 per cent as a portion of total value.
45 | I S I W S C 2 0 1 9