Closing a period locks it. After that the numbers for that month stop changing, which is the whole point.
Why close at all
An open period is a period where the figures can still move.
That is fine while you are working in it and unacceptable once you have reported it. A profit and loss quoted to a board in April and different in June is not a reporting problem, it is a credibility problem, and the fix is a lock rather than a promise that nobody will post anything.
What the lock does
No new postings land in a closed period. Not discouraged, not warned about: rejected.
That rejection happens in the database rather than in the interface, which is the distinction that makes it a control. A lock enforced only by a screen is bypassed by anything that does not go through that screen.
Reversals post to the next open period. That is standard practice: if something in January was wrong and January is closed, the correction belongs in February with a description explaining what it corrects. History is not rewritten.
Drafts block the close
A period cannot close while draft journal entries sit in it, and the surface lists the ones in the way.
That is deliberate. A draft entry is neither in the books nor discarded, and letting a period close around one leaves a decision permanently unmade.
The resolution is always one of two things: post it, if it belongs, or void it, if it does not. There is no third option and putting it off is what leaves month ends half-done.
Locking is enforced, not requested
The distinction that makes this a control rather than a convention: a posting into a closed period is rejected by the database.
Not hidden by the interface, not warned about. Rejected at the level everything has to go through, so an integration, an import or a script cannot quietly write into a locked month either.
That is the difference between a close you can rely on and a close that depends on everybody remembering.
The routine
Once a month, and it is short if the month was worked properly.
Check the trial balance balances. If it does not, nothing else matters yet.
Clear the drafts. Post or void each one.
Run the reconciliation. Sub-ledgers against the general ledger, which has its own article. Drift here is the most common reason a close is not clean.
Look at the profit and loss for anything obviously wrong: an empty account that should have activity, an account with far more than expected.
Close it.
Twenty minutes when the month was tidy, and considerably longer when it was not, which is the incentive to work tidily.
Adjustments belong before the close
Anything that makes the period correct goes in before the lock: accruals, prepayments, depreciation, corrections.
That is the actual work of a month end, and it is the reason the close is a step rather than an automatic event. Locking a period that has not been adjusted preserves an incomplete picture rather than a finished one.
What the close is protecting
Three things, and only the first is obvious.
Reported figures, so a number quoted externally stays true.
Comparatives, so this month against last month is a comparison rather than a moving target.
The audit trail, since a period that can still change is a period where an auditor cannot rely on anything.
An organisation that never closes has none of those, and usually does not notice until somebody outside asks a question it cannot answer.
Reopening
A closed period can be reopened.
The surface warns, and the warning is the honest one: any reversal or adjustment bookkeeping that results is your responsibility. Reopening does not undo anything that was decided on the basis of the closed figures.
Reopen for a genuine error discovered soon after. Do not reopen for a routine correction, which belongs in the current period as an adjustment.
An organisation that reopens periods regularly is an organisation whose closes mean nothing.
What blocks a close
Two things, and the surface names both.
Draft journal entries in the period.
An unbalanced trial balance, which has to be resolved before anything else is worth doing.
Neither is arbitrary. Both are states where closing would lock in an unresolved decision, which is exactly what a close is supposed to prevent.
Fiscal year
Which month starts your year is set in accounting settings, and periods follow from it.
Worth checking once if your year does not start in January, because it affects every period boundary and every year-to-date figure.
Closing in order
Periods close in sequence. January before February.
That prevents the situation where a closed February sits on top of an open January, which would mean the earlier period could still move underneath a locked later one and every comparative between them would be unstable.
Who should do this
One person, the same person each month.
Closing is a small task that depends on knowing what normal looks like, and somebody doing it for the third time spots things a rotating cast never will.
What a good month end looks like
The close is quick when the month was worked and slow when it was not, and the difference is entirely in habits during the month.
Payments recorded as they arrive rather than in a block.
Expenses approved twice a week rather than at month end.
Journal entries posted or voided rather than left in draft.
An organisation doing those three has a twenty-minute close. One doing none of them has a two-day close and calls it month end, which is the same work moved somewhere less convenient.
What closing does not do
It does not produce a report. The statements are always available; closing fixes what they will say for that period.
It does not stop the current month. Only the closed one is locked.
It does not check your work. The lock enforces that nothing changes afterwards. Whether what you locked was right is what the routine above is for.
Did this answer your question?
No, ask a person