The trial balance

In depth

Why the columns must agree

The agreement is algebra, not luck. Every journal entry has equal debits and credits, and posting copies them, so the ledger’s debit postings and credit postings have the same total: for Lantern’s March, $61,110 each. Write $D_i$ and $C_i$ for the debits and credits posted to account $i$. Its balance, counted as positive when it is a debit balance and negative when it is a credit balance, is $D_i - C_i$. Adding over every account:

Working out a balance nets the smaller side of an account against the larger, which removes the same amount from both columns. Lantern’s Cash, Accounts Receivable and Accounts Payable netted $10,250, $3,000 and $1,000, $14,250 in all, which is why $61,110 of postings on each side becomes $61{,}110 - 14{,}250 = 46{,}860$ in each column.

What a balancing trial balance proves, and what it doesn’t

Equal totals prove three things: the debits posted equal the credits posted, each balance was worked out and placed in a column consistently, and the columns were added correctly. They prove nothing about whether the right things were recorded. A trial balance can balance and still contain errors, and some of the most expensive errors in accounting are exactly the ones it can’t see.

Bookkeeping courses in the British tradition give these undetectable errors standard names. The Cambridge International AS & A Level Accounting syllabus for 2026–2028 lists errors of omission, commission, principle and original entry, reversals and compensating errors, and ACCA’s study material defines all but the last in the same terms. The American textbooks cited in this lesson describe the same mistakes without the labels. Here they are with Lantern examples, each a hypothetical mistake in March’s records.

Error of omission

A transaction left out entirely. If the utility bill of 31 March had never been recorded, Utilities Expense and Accounts Payable would each be $260 too low, and both columns would total $46,600.

Error of commission

The right amount on the right side, but in the wrong account of the same type. Posting March rent to Utilities Expense leaves Rent Expense at zero and Utilities Expense at $1,460; both are debit balances, so the totals don’t move. Posting an invoice to the wrong customer’s account is the same error inside a subsidiary ledger.

Error of principle

The right side, but the wrong type of account, which breaks an accounting principle. Suppose a $400 repair to the espresso machine were debited to Equipment instead of Repairs Expense: assets would be $400 too high and expenses $400 too low, and the totals would still agree.

Error of original entry

A wrong amount in the journal, posted faithfully to both accounts. If the 20 March wages had been journalised as $1,060 instead of $1,600, Wages Expense would be $540 too low and Cash $540 too high. Both are debit balances, so the debit column would still total $46,860.

Reversal of entries

The right accounts and amount with the debit and credit swapped. Recording Meridian’s $3,000 payment as a debit to Accounts Receivable and a credit to Cash would leave Cash at $20,650 and Accounts Receivable at $7,200, and both columns at $46,860.

Compensating error

Two or more errors whose effects cancel. If Wages Expense were overstated by $100 and Catering Revenue were also overstated by $100, both columns would rise to $46,960 and still agree.

Duplicated entry

A whole entry recorded twice adds equal debits and credits again. Journalising the $350 advertising payment twice would leave Advertising Expense at $700 and Cash at $26,300, with the totals unchanged.

Every one of these leaves a debit for every credit, and that common thread gives you a test for any mistake you meet: if it keeps debits equal to credits, the trial balance can’t see it. Finding such errors needs other checks: reconciling Cash with the bank statement and each control account with its subsidiary ledger, comparing balances with what you expected, and following cross-references back to the source documents.

Errors the trial balance does reveal

Anything that puts more on one side than the other shows up as a difference between the totals. The common culprits leave differences with recognisable shapes:

Using the difference to find the error

When the totals disagree, work through the checks in a fixed order, cheapest first. The sequence below follows the one in Edwards and Hermanson’s textbook (see the references), with one check, step 2, moved forward from its later list of causes:

1. Re-add both columns. Addition slips are common and quick to rule out.
2. Look for an amount equal to the difference: a balance missing from the list, or a journal line with a blank posting reference.
3. Halve the difference and look for that amount on the wrong side. A $300 debit listed as a credit makes the debit column $300 short and the credit column $300 long, a gap of $600. Start by checking that every balance sits on its normal side.
4. Divide the difference by 9. If it divides exactly, look for a transposition or a slide.
5. If none of that finds it, work backwards: compare every trial balance figure with its ledger account, recompute each ledger balance, check each posting against the journal through its posting reference, and finally check that each journal entry balances.

Notice which column is short. If the debit column is the smaller one, the error either understated a debit or overstated a credit, which halves the number of places to look.

Why transpositions and slides leave a multiple of 9

A transposition error swaps two digits: 54 written as 45, or 1,600 as 1,060. Take a two-digit amount with tens digit $a$ and units digit $b$. Its value is $10a + b$; with the digits swapped it is $10b + a$. The difference is

which is always a multiple of 9, and the multiplier, $a - b$, is the gap between the two swapped digits. When the pair sits further left, every place value is multiplied by 10, 100 or more, so the difference becomes $9(a - b) \times 10^k$. Writing 1,600 as 1,060 swaps the 6 in the hundreds place with the 0 in the tens place: $9 \times (6 - 0) \times 10 = 540$, and indeed $1{,}600 - 1{,}060 = 540$.

A slide error puts the decimal point in the wrong place, so every digit shifts: $350 recorded as $35, or as $3,500. If the smaller figure is $y$, the larger is $10y$ and the difference is $10y - y = 9y$. A slide of two places, $350 recorded as $3.50, gives $100y - y = 99y$, also a multiple of 9.

A deeper reason covers every rearrangement of digits, not just neighbours. Each power of ten is one more than a multiple of 9 (10 = 9 + 1, 100 = 99 + 1, 1,000 = 999 + 1), so any whole number equals a multiple of 9 plus the sum of its digits. Two numbers made of the same digits have the same digit sum (1,600 and 1,060 both have 7), so they leave the same remainder when divided by 9, and their difference is a multiple of 9. Adding or removing zeros, which is what a slide does, doesn’t change the digit sum either. When amounts include cents, apply the test to the difference in cents: $350 against $3.50 differs by 34,650 cents, which is $9 \times 3{,}850$.

Three trial balances in one cycle

The trial balance you have just prepared is the first of three that a full accounting cycle produces. Each is prepared the same way, at a different point, and the later lessons build them for Lantern:

Lantern closes its books every month so that you can see the whole cycle in one month; most businesses close once a year.

A trial balance is not a financial statement

A trial balance looks official, with a heading, a date and double-underlined totals, but it is an internal working document. It mixes accounts that belong on different statements: assets and liabilities from the balance sheet, revenue and expenses from the income statement, dividends from the statement of changes in equity. It has no subtotals, doesn’t show net income, and doesn’t separate current from noncurrent items. Its jobs are to check the ledger and to feed the statements; turning Lantern’s adjusted trial balance into financial statements is the subject of lesson 10.