This article delves into the "four-stage reconciliation" mechanism in modern payment systems, and explains in detail the reconciliation relationship between accounting data, payment data, clearing data, and settlement data and its importance. By reading this article, you will have a comprehensive understanding of the core logic and implementation details of payment reconciliation, which I hope will be helpful to you.
No matter which system you are in, the reconciliation system is a very important part, and it is a link
The same transaction will produce multiple sets of data within an organization, and there is a reconciliation relationship between these data, and their reconciliation relationship can be called "four-stage reconciliation"
What is "four-stage reconciliation", the four segments here refer to the "four-segment data", and the "four-segment reconciliation" refers to the mutual verification of the four segments of data, as shown in the following figure
1. What are the four pieces of data?
The four pieces of data are "accounting data, payment data, clearing data, and settlement data", which are the core data of reconciliation, and there is the following symbiotic relationship between them
From the following diagram, it can be seen more clearly that transactions drive payments and accounts, payment drives channels, and the channels themselves generate clearing records and settlement records
It's just that it is necessary to be clear here that the transaction has gone through "billing" in the process of driving the accounting, that is, the merchant handling fee and the channel cost fee, so these two parts of the fee should at least participate in the verification of the fund reconciliation link in principle, and even participate in the verification of the details
Looking back at these four pieces of data, how to label these four pieces of data, that is, you can easily identify what data is through a labeling method, for example, you can call the accounting data Zhang San, the payment data Li Si, the channel data Wang Wu, and the channel settlement data Lao Liu
However, it is clear that this is not very good for communication and data transfer between systems
Another method is to achieve it through a set of digital encoding, and at the same time, the same piece of data, such as payment data, has subcategories such as payment, refund, payment, and payment return, so at least multi-dimensional classification can be achieved in coding, first look at two coding design examples
1) Just like the national level 1 subject number
- Asset class: 1001 is cash in hand and 1002 is bank deposits
- Liabilities: 2001 is short-term borrowings
- Owner's equity: 4001 is paid-in capital, 4002 is capital reserve
2) Also look at the design rules of the second-generation system for business coding
Through such rules, the following business codes are obtained
3) Data segment coding design
Inspired by the above encoding design, we can also arrange each piece of data in this way, so that we get the encoding rules for four pieces of data
According to the actual business type, such as collection, refund, payment, and return, the code of each segment of data is assigned separately (234 segments of data in the figure, the first digit is changed to the corresponding number of segments)
In this case, the four pieces of data are determined
Based on this idea, if there are more services or new types of data, you can continue to increase to 5 segments of data, 6 segments of data, 7 segments of data, and 8 segments of data......
The second and fourth segments of the data will produce three sets of checks
Earlier we introduced 4 pieces of data, so according to their symbiotic relationship, we can generate a three-segment check relationship
1) Account reconciliation
This is the reconciliation of payment records and accounting records to ensure that all payment requests initiated by merchants are completed accounting records, because merchant settlement is based on accounting data, so only to ensure that the data paid to the account is consistent, then the settlement to the merchant can be correct
The meaning of reconciliation is: whether my payment records are all recorded?
2) Transaction reconciliation
It is the verification of the payment record of the platform and the payment record of the channel (channel clearing data), this link is to ensure that the payment situation recorded on my side is consistent with the payment and clearing of the channel, in principle, the payment is initiated by the platform, but the payment processing result is notified by the channel, and the two sides should be consistent
Then the meaning of transaction reconciliation is: whether my records are consistent with the records of the channel
3) Reconciliation of funds
The platform submits a payment request to the channel, and the channel generates the channel's own payment records, that is, the clearing documents obtained by the platform, and some are also called "transaction records", that is, the platform's "receivables and payables"
The channel will eventually settle the funds to the platform based on the clearing records
So the meaning of fund reconciliation is: Is the money that the channel should have settled for me?
Here is an interview question for you: After the account data is settled to the merchant settlement account, then there is a possibility of errors in this step of transportation, what are you going to do?
3. 3 sets of errors were derived
There is an error in the reconciliation, that is, it is not even, there are inconsistencies between the two data, you have me, I have you, you have you, you have I have is different, the status is different, the amount is different, etc
Since we have 4 pieces of data and 3 pieces of checks, there will inevitably be 3 sets of errors
1) Accounting errors - more or less in and in
That is, there is an inconsistency between the accounting data and the payment data verification, that is, there is no correct data in the account, the merchant who has less account does not agree, and the platform with more account entry is too aggrieved
So, for the sake of fairness, what is is how much is how much, not much for a point, not much for a point
2) Transaction errors - even, unilateral or wrong accounts
That is, the check between the platform payment data and the channel clearing data, that is, you have memorized what I asked you, and I have updated it correctly when you notified me
Then this check link will produce transaction errors, that is, unilateral accounts or wrong accounts, unilateral accounts are one party has no party, and wrong accounts are both sides but the amount is inconsistent
3) Funding errors - long or short
That is, the verification of receivables and payables and actual receipts and payments, which is a very important part of the check, because it is related to whether the platform itself has received the money it should have received and whether it has paid the funds that should not be paid
To put it simply, whether there is less money and no more money; Charge less and be sorry for the merchant, and pay more and be sorry for yourself
In short, money is no small matter, and no point can be lost
Fourth, the data segment derived from the error
Have you noticed that the original four pieces of data have produced error data after 3 sets of checks, and if you want to process the error, the error processing data will be generated
In this way, new data will be generated, a derived data, which is also data, and is also equipped with "names", depending on the value assigned by the corresponding data "segment" in the figure above
It should be noted here that the unilateral account is not new data, but the original payment data and channel clearing data, but the other party does not have it
However, when you initiate error processing on a single side, new data will be generated, such as "channel unilateral - platform supplement", in order to distinguish the difference from the normal platform data, you can start a new level to distinguish this data from the data that is supplemented after the reconciliation
5. The purpose of these data "segments".
These data have a great use and play an important role in the subsequent data transfer and accounting process
When you see the "number" of the data, you know what kind of data it is, how it is stored in the system, how it is generated, which accounts to be operated in the clearing and settlement accounting process, and what kind of vouchers the accounting system needs to generate
For example, if a user makes 2 payments of 10 yuan each, 1 transaction is successful, and the other payment is being processed, and the payment core generates payment data
Transaction-driven accounting for bookkeeping, with this data as the accounting data, according to the above figure, the payment record is 2 pieces of data, the collection data is 2001, and the accounting entry rules for the 2001 data are
i.e. data = {payment successful, 2001, 10 yuan}, the generated voucher is as follows
Borrow Liquidation - $10
The loan merchant is to settle 10 yuan
It is represented by a legend as shown in the following figure:
or
This article was written by Everyone is a Product Manager Author [Chen Tian Universe], WeChat public account: [Chen Tian Universe], original / authorized Published in Everyone is a product manager, without permission, it is forbidden to reprint.
Image from Unsplash, based on the CC0 license.