Without turning around, Eric Denna threw a chalkboard eraser at me as hard as he could. At least it seemed that hard. It whizzed past my head, bounced off the back wall of the classroom, and dropped onto the floor. Eric was a brand new professor. This was his first course. Perhaps that explains some of his enthusiasm, and the strength of his convictions. From the first day of class I sensed he was ready to change the world. And we were his tools.
I was in my junior year in the accounting program in the fall of 1988, and Eric was my professor for an introductory course in accounting information systems. He had just graduated from Michigan State University. His mentor had been William E. McCarthy and his specialty the REA Accounting Model. REA stood for Resources, Events and Agents. Eric was drilling it into our heads that an accounting system could be constructed without specifically recording revenues, expenses or any of the other accounting system classifications. That would change the definition of the accounting model. He wanted to make sure we understood that. So while he was writing REA on the board to begin the catechism, I flippantly said it was all clear to me: REA stood for Revenues, Expenses and Assets. The eraser was his answer.
The Book of Record
What would accounting be without revenues or expenses? Eric taught us that revenues, expenses, liabilities, assets and owner’s equity were simply “classifications,” “views” or perspectives of business events that were of interest to the accountants. But other views of these business events exists that were equally valid for other users. By forcing the accounting view of the data on top of the data capture, other uses or views of the data are destroyed. He said there was an entirely different way the system could be constructed to report the same results. He argued that that approach would allow the system to do much more.
What does that mean? Let’s start simple by continuing our example, by defining what “the data” is, or perhaps more descriptively, what the “book of record” is. The term “book of record” is used by accountants to define the most important part of the accounting system; that which is audited and found to be accurate in “all material respects.” The choices for the book of record from our accounting system are:
- The information on the balance sheet and income statement,
- The information on the trial balance,
- The rows in the General Ledger, or
- The journal entries.
Options one and two don’t make a lot of sense; they are outputs from the book of record, not the book of record. Once the statements are produced they are static and don’t change; they are not updated through the accounting cycle.Most accountants would define the General Ledger (the GL) as the book of record. It is the last step in the accounting cycle before the creation of the trial balance and financial statements. It is an ongoing part of the process.However, every balance in the GL has to be supported by journal entries. This then suggests that actually the journal entries could be the book of record. Instead of using the General Ledger as the source of producing the trial balance and the statements, it is possible to use the journal entries. But, to use the journal entries as the source of producing the trial balance and the statements, one would need to keep all the journal entries ever created and add them up to produce the trial balance as of a point in time.Make no mistake; this isn’t an easy thing to do. In our personal financial example, if some portion of the cash you have on hand was earned on your first summer job, you would need to have journal entries for it and all the other receipts and expenditures since then to calculate your cash balance today.That seems a bit unrealistic. But if it were possible, would there be benefits? Does that system do more? Would it allow other “perspectives” or “views” of the financial information as Eric suggested?
Flexible Income Statements
One thing it would enable is creation of income statements for any period of time, eliminating the need for the closing entries. By using the general ledger, an income statement can be produced every day in April, the balances would always start from the first day of the month.For example, if we wanted our accountant to give us an income statement for an odd period, say April 16th through May 5th, she couldn’t use the GL for the income statement balances. Ignoring the closing entry, the following journal entries show that the salary revenue for that period should be $2,243.65.
But if the General Ledger were used, it would show zero on May 5th.
If we produced the income statement by adding up the journal entries within the date range, rather than from the GL, we could produce this and any other income statement.
Journal entries could be used to generate the balance sheet in a similar manner; the journals can be added up to determine the financial position at any point in time. There is a catch, though. All income statement accounts must also be included in the balance sheet calculation in order to show net worth or retained earnings. This would be necessary if there were no closing journal entries.If we decided to call the journals the book of record, perhaps we have a bit more flexible reporting system as we have seen. These benefits aren’t really all that great. But this way of thinking helps to highlight a key concept, which can lead to even greater flexibility in the accounting system.
The “E” in REA Theory stands for “Event”. The idea behind the theory is that events, business events, provide the basis for business reporting. A business event is anything the business wants to plan, execute, control, or evaluate. For example, in our personal financial world, being paid our salary is an event we would likely want to plan, execute by seeing that we actually get paid, control where the money is deposited, and evaluate any deductions taken out such as taxes when we do our income taxes.It is likely that every journal entry made in our personal financial system is an event; we wouldn’t have taken the time to make a journal entry of it if we didn’t want to plan, execute, control or evaluate it. Thinking of the journal entries as events helps to understand the theory. Events are sometimes more commonly called transactions.
Transactions Versus Balances
This difference, between viewing the GL or the journal entries as the book of record highlights a key contrast in the approach to financial reporting: the difference between transactions and balances. Balances tend to be a point in time picture, much like the balance sheet. The GL shows the balance for each account after the posting process. The GL balances are immediately informative; there is no calculation to be done to make them useful. However, they aren’t very flexible; they are mostly useful for the information they contain at that moment.Transactions aren’t as informative. They can be used as building blocks and combined in many ways to present information. But, they have to be added up to form balances before they become interesting.1To show this graphically, the following are the transactions—the offset side of the journal entries—affecting the credit card payable account.
The following are the resulting balances at the end of each day from posting these transactions to the General Ledger Credit Card Payable Account Balance:
What do the general ledger balances tell us? We can see the date, the account number and name, the day’s change in the balance, sometimes called the movement, and the balance at the end of the day.
Notice that the number of balances is less than the number of journal entries because we purchased tires and had the alternator replaced both on the same day. From the balances we can’t see what we did on that day, just the net effect of all that happened. Another difference is that the journals have a description. Because more than one thing can happen on a day, the description would be meaningless summarized on the ledger.Why would that be important? Notice that many of the journals deal with the car. Might it be useful to know when we last purchased tires; perhaps we could negotiate a discount if they had not lasted as long as expected? We can find this information on the journal entries, but not on the ledger.And that is a major point Eric was trying to teach us. More information is available from the journal entries than the ledger. The financial reports aren’t geared to tell us about car maintenance; they tell us about money. When we summarize transactions, like we do when we update the ledger, we lose information that might be useful to us or others. The accounting “view” of the data precludes other uses of it.Balances have an affinity for the balance sheet. Transactions are sometimes called movements, in that they represent the changes in balances. They have an affinity for the income statement—the statement that explains changes in the balance sheet.These affinities and relationships aren’t always so straight forward. Sometimes balances are used as transactions, in that they are manipulated to form new balances and new information as we will see later. And although the income statement shows accumulated “transactions,” it is produced as of a specific point in time using revenue and expense balances from the ledger. Thus the distinction between balances and transactions isn’t always tidy and our language in describing them isn’t very exact, but being able to recognize and use each is important.
Thus far in our financial system, we have been interested in one category of “things” called account. In our little financial example, there isn’t much question about the “who” is involved. We have assumed that our financial system is just for us as an individual, but business financial systems aren’t just for one person. A business is most often composed of groups of people, and the system has to produce reports for these groups. In business, the “who” is very important and a lot of work goes into keeping track of it all. Let’s extend our example financial system in perhaps a bit of a farfetched way to show how this impacts the system, and how this can cause balances to become transactions if approached in a particular way.Let’s assume that our financial system is for our family, which is composed of more than one person. Each family member in our little system might use some ID, like a social security number, to record on the journal entries which ones are theirs. In our example, as in many traditional financial systems, a concept of a cost center is used to track the “who”.2
This cost center simply becomes another column on the journal entries containing numbers similar to the Account column. This column would also be added to the General Ledger. Thus when we post entries, we post them to a combination of account and cost center. At any point in time we can find what the balance is for a particular account for a particular cost center by looking in the general ledger. We can then produce a trial balance which is summarized by cost center as well as by account. Using these pieces of information, we can then produce financial statements for each person.
Let’s suppose our journal entries for salaries identify whether they are for Dad, Mom or the first child.
When we make a trial balance from a ledger that includes cost center as a posted element (a column), it would look like this:
Now we have an interesting problem. We have added cost center to our whole system, and that allows us to produce information at a lower level of detail. But have we lost the ability to produce the summary results? Can our system still produce the consolidated family financial statements we showed in Accounting above? The trial balance now has multiple rows for one single account; one row for Dad’s checking account and another for the Mom’s and a third for the Child’s3. There is a need to add together these rows before we can place them on the consolidated family financial statement.This could be accomplished in a couple of different ways. One way would be to create another ledger, a summary ledger. In the summary ledger, we might only have account, and not record cost center, just like the ledger in our example above. We then post each journal entry to two ledgers, once to the detailed ledger, and then once to this summary ledger. We can then produce two trial balances, one from the summary that feeds the consolidated statements, and one from the detailed ledger that feeds the individual statements.4This approach raises an interesting question: Which ledger, the summary or the detail, is the book of record?Another approach is possible though. We could post only to the detail ledger, and then produce the detailed trial balance that includes cost center like that shown above. We could then take a temporary copy of the last detailed ledger rows and blank out cost center. We could then summarize these rows, by adding together like rows where two people had activity for the same account, to create a new trial balance for the consolidated entries. The new rows would look like the following5:
This second approach is interesting. Here we have taken the balances as of a point in time in the detailed ledger to make transactions that we then summarize to create new balances.This now raises another interesting question: Which journal entries would be considered the book of record: The detailed journal entries that include cost center, or do we also include the new summary journals we’ve just created?
This same approach to report on new attributes is used through IT systems; when a new attribute is needed, more detail is created. Yet at report time, the detail has to be summarized out. Today’s Finance have need to add attributes like Book Code (IFRS or Local GAAP), products, customer types, even customers for high value accounts. This constant contest, between the need for additional attributes on the one hand and the need for an instantaneous answer on various reports on the other, is what McCarthy and Eric were trying to get at.
Actually, a third approach is also available. We could add up each of the detailed journal entries, which includes cost center, and create both trial balances, by simply ignoring the cost center value when making the second trial balance. This can be done because we have computers. Eric started to intertwine accounting and my computer classes, and show they had a closer relationship than I had realized.
Additional options are available for a definition of the book of record besides just the GL or the double-entry journal entries. A different set of “journal entries” could also be considered the book of record. But now we start to depart from the standard process of accounting.