Flux & Variance

What is flux analysis in accounting?

Flux analysis compares each account balance to a prior period and explains what the movement says about the business. It's a standard part of month-end close, and almost nobody learns it in school.

The short version: you take each account, compare the current period's balance to a prior period's, and for anything that moved enough to matter, write an explanation of what caused it. Done across the trial balance, those explanations add up to a picture of what happened to the business during the period, told through the financial statements.

"Flux" is short for fluctuation. You'll also see it called fluctuation analysis, and it overlaps heavily with what people call variance analysis — the terms get used interchangeably in practice, with some distinctions worth knowing about further down.

Why you probably weren't taught this

Accounting coursework covers how transactions get recorded, how statements are built, and how to audit them. It rarely has you sit with a real trial balance and work out why an expense account is up 40 percent from last month.

So most people meet flux analysis on the job, usually by being assigned it. That's a genuinely disorienting way to encounter a task — you're handed a workbook, told to explain the variances, and left to work out both what a good explanation looks like and how to find one.

If that's where you are right now, the confusion isn't a sign you missed something. There was nothing to miss.

What it's actually for

Two purposes, and they run at the same time.

It's a control. Someone reviews the movements and signs off that they're explained and reasonable. That's what makes it appear in a SOX control matrix or an audit workpaper — it's a detective control, catching errors that made it through everything upstream by asking whether the resulting balance makes sense.

It's how the business gets explained. Somebody above you needs to know why margin moved, why headcount cost jumped, why receivables grew faster than revenue. Flux commentary is often where those answers first get written down in a form anyone can check.

The second purpose is the one that makes the work interesting and the one that's easy to lose sight of when you're just trying to finish.

Flux analysis vs variance analysis vs fluctuation analysis

In everyday use these are close to synonymous, and nobody will correct you for mixing them. Where people do draw a line:

• Flux analysis tends to mean actual against actual — this period against a prior period. Month over month, or the same month last year, or year to date against the prior year to date.
• Variance analysis is the broader term and often includes actual against budget or forecast, which is a different question: it tests a prediction rather than comparing two things that really happened.
• Fluctuation analysis is just the long form of flux, more common in audit language than in close language.

Worth knowing which one your organization means, since the comparison decides what counts as an answer. The full comparison is here, and the comparison types are covered on the hub.

What doing it actually looks like

Some version of these steps, whatever the software:

1. Get the comparison. Pull balances for both periods and calculate the movement, in dollars and usually as a percentage.
2. Decide what needs explaining. Most places set thresholds — a dollar amount, a percentage, or both — so you're explaining the accounts that moved materially rather than all of them.
3. Find the reason. This is the actual work. Pull the detail behind the account — organized by customer, vendor or category — and find what drove the movement.
4. Write it down. Name the drivers, quantify them, and point at where someone could verify each one.
5. Get it reviewed. Someone checks it, usually sends some of it back, and you revise.

Step three is where the time goes, and it varies enormously by account. Cash is often close to self-explanatory — you can see deposits against payments and the story is mostly there. An accrual account with a dozen entries posted against it is a different experience entirely, because the ledger tells you that entries posted, not what each one was for.

The mistake almost everyone makes first

The natural way to learn this is to look at what last month's explanation said and write this month's version of the same thing. It works, it's fast, and it's how most people get through their first few closes.

An account can move by a similar amount for an entirely different reason. Copying forward captures the format and misses the change, and it's specifically hard to catch because nothing about the output looks wrong — it reads fine, it's in the right style, the number ties.

Using last month as a starting point is fine. The step people skip is asking whether the driver is actually the same one, before reusing the sentence that describes it.

What makes it hard at first, and what helps

The first few times through, it's mostly volume and unfamiliarity. There's a lot to get through, you don't yet know which accounts are quick and which are slow, and you're figuring out the software and the expectations at the same time as the analysis.

That part improves on its own. Accounts get much faster once you know how their entries get booked, which is why the second and third month are dramatically easier than the first.

The harder habit to build is not letting the task collapse into just getting it done. It's easy to end up inside the ledger — clicking through detail, matching numbers, filling boxes — and lose the thread of what any of it means for the business. The explanations that get sent back are usually the ones written from inside that tunnel: technically accurate, internally consistent, and saying nothing anyone can use.

The question worth keeping in view is what someone reading this would actually want to know about the business. That's the difference between describing a movement and explaining one.

Where to go from here

• Month-end flux, start to finish — the comparison types, what makes commentary hold up, and where each piece is covered in depth.
• Why flux commentary gets sent back — the three most common reasons, and what reviewers are checking.
• The commentary test — a free tool that grades your driver lines against four computed checks. Runs in your browser.
• Income statement and balance sheet flux — why they need different data pulls.

If you work this in Excel: the free flux template flags what needs explaining and shows the residual your drivers don't cover. The CloseOps Flux & Variance System ($79) starts from your trial balance instead: confirm each account's classification and it builds the income statement and balance sheet flux statements, then ranks what to investigate.

Related: a worked example, end to end, month-end flux, start to finish, four of the five tests your commentary fails are arithmetic, and why your commentary keeps getting rewritten.

Found something wrong or missing? Tell me here — anonymous, thirty seconds.