The FAST Standard Is Rubbish. There, I Said It.

If you’ve spent any time in project finance or infrastructure modelling, you’ve heard of FAST. Flexible, Appropriate, Structured, Transparent. It sounds lovely. It has a certification programme. It has signatories — big-name firms with logos on a website. It hasn’t been meaningfully updated since July 2019.

And I think it’s rubbish.

Not all of it. Some of the principles are perfectly sensible — consistent formulas, clear labelling, don’t merge cells. But these aren’t ideas FAST invented. They’re things competent modellers have been doing for decades. The parts that are uniquely FAST — the rigid rules, the dogmatic structure, the sign convention — are where the problems start. Let me explain.

The Sign Convention Is Backwards

This is the hill I will die on.

FAST rule 1.01-08 mandates the “normally positive” convention on workings sheets. Every figure sits in the model as a positive number. Revenue? Positive. Costs? Also positive. Capital expenditure? Positive. Debt repayment? Positive. The direction — whether money is coming in or going out — is supposedly communicated by the label, not the number.

Then, on presentation sheets, rule 1.01-09 kicks in: you switch to inflow/outflow convention, where outflows become negative. So somewhere in your model, you need sign switches — and FAST even has a rule for how to do them (3.04-04: use -1 * explicitly, never just a minus sign).

This is, to put it politely, insane.

Think about what cash actually does in a business:

• Revenue makes cash go up. It should be positive.
• Operating costs make cash go down. They should be negative.
• Drawing a bank loan makes cash go up. It should be positive.
• Repaying a loan makes cash go down. It should be negative.
• Capital expenditure makes cash go down. It should be negative.
• Selling an asset makes cash go up. It should be positive.
• Collecting a receivable makes cash go up. It should be positive.
• Paying a supplier makes cash go down. It should be negative.

This isn’t complicated. It’s how money works. Positive means cash increases. Negative means cash decreases. Every person who has ever looked at a bank statement understands this intuitively.

The “normally positive” approach fights this intuition at every turn. You enter your operating costs as 500,000 — a positive number — and then somewhere downstream you have to remember to flip the sign before it hits the cash flow statement. You enter your capital expenditure as 10,000,000 — positive — and flip it later. You enter your debt service as 2,000,000 — positive — and flip it later.

Every single one of those sign switches is an opportunity for error. And not a small, easily- spotted error. A sign error doubles the impact — what should be a £5m outflow becomes a £5m inflow, a £10m swing that can hide in a model for months.

The Extra Row Problem

Here’s what makes the FAST sign convention particularly dangerous: it contradicts FAST’s own simplicity principle.

FAST rule 3.01-03 says formulas should be “as simple as possible.” But the normally- positive convention requires an additional sign-switching row for every outflow item that needs to appear on a financial statement. That’s an extra row, an extra formula, and an extra place where things can go wrong.

In a model with, say, 40 line items that need sign-switching before the financial statements, that’s 40 additional rows of -1 * formulas. Forty new touchpoints where someone can link to the wrong row, forget to add the switch, or accidentally switch something that shouldn’t be switched.

The inflow/outflow convention — costs entered as negatives from the start — eliminates all of those rows entirely. Your SUM formulas work naturally. Your cash flow statement adds up without intervention. The model is shorter, cleaner, and has fewer places for errors to hide.

FAST’s own simplicity principle argues against its own sign convention. The irony is remarkable.

This Isn’t Theoretical — It’s How the Big 4 Actually Work

I spent years working in financial modelling at a Big 4 accountancy firm. We used the negative convention for costs. Not because someone wrote a blog post about it — because it’s the approach that survives contact with reality when you’re building models under deadline pressure, handing them to partners who need to trust the numbers, and defending them to clients who are making nine-figure decisions.

And it’s not just us. Over those years I reviewed models built by all four of the Big 4 firms, across transactions of every size and shape. Every single one of them used costs as negative. Not one used the FAST “normally positive” convention on their workings sheets. These are firms that collectively employ thousands of financial modellers, that have had decades to test which approaches work and which ones cause problems. They all landed in the same place — and it’s not where FAST tells you to be.

“But the numbers look ugly with negatives on the input sheet!” I hear the FAST loyalists cry. So what? Your model exists to be correct and auditable, not to look pretty on the inputs tab. A cost of -500,000 on an input sheet is unambiguous. A cost of 500,000 that silently becomes -500,000 three sheets later is a lawsuit waiting to happen.

Rules for Rules’ Sake

FAST is a rules-based standard. It tells you exactly how many columns to use for labels (rule 2.02), exactly where to put your time series (rule 2.04), exactly how to name your sheets (rule 2.01). It has over 100 specific rules, many with sub-rules.

The ICAEW Financial Modelling Code, by contrast, is principles-based. It says: here are the things that matter — clarity, consistency, integrity — and here’s the reasoning behind each principle. Then it trusts you, the modeller, to apply those principles in a way that fits the specific model you’re building.

This distinction matters enormously in practice.

A rules-based standard is brittle. It works perfectly for the type of model the rule-writers had in mind (in FAST’s case, project finance models for infrastructure) and falls apart the moment you try to apply it to something different. A three-statement operating model for a retail chain doesn’t need the same structure as a wind farm cashflow model. A PE acquisition model doesn’t need the same layout as a government PFI. But FAST’s rules don’t flex — they’re rules, not principles.

The ICAEW Code was built through consensus across the industry rather than by a single organisation promoting its own methodology. The contributor list includes people from Grant Thornton, KPMG, Mazars, Operis, and a dozen other organisations — the same firms whose logos appear on the FAST signatory page, incidentally.

The Calculation Block Bloat

FAST’s insistence on “one formula per row” (rule 3.01-01) and its block-based calculation structure leads to models that are extraordinarily long. A calculation that would take three rows in a conventionally structured model becomes seven or eight rows in a FAST model when you add the call-up row, the calculation row, the check row, and the various structural elements FAST requires.

Defenders of FAST will say this makes the model more transparent. In practice, it makes the model harder to navigate. You spend half your time scrolling through structural scaffolding instead of looking at the actual logic. A model that would be 2,000 rows becomes 5,000 rows. That’s not transparency — it’s clutter wearing transparency’s hat.

The Document Is Incomplete

Section 4 of the FAST Standard, which was supposed to cover “Model Outputs,” is marked as “to be completed.” It has been “to be completed” since 2019. The version history shows the last substantive update was July 2019 — nearly seven years ago.

If you’re going to position yourself as The Standard, you need to actually finish writing it.

The Certification Incentive

Let’s address the elephant in the room. FAST offers a paid certification programme. There is an inherent tension in an organisation that both sets the standard and sells qualifications against it. I’m not suggesting bad faith, but the incentive structure is worth noting: the more prescriptive and specific the standard, the more there is to test, and the more valuable the certification becomes.

A principles-based standard like the ICAEW Code doesn’t lend itself to a multiple-choice exam. You can’t easily test “do you understand and can you apply the principle of clarity?” in a standardised assessment. But you can test “did you put the time series starting in column H?” That’s not a coincidence.

So What Should You Use Instead?

The ICAEW Financial Modelling Code. It’s free, it’s complete, and it was written by a genuinely representative group of senior practitioners across the industry.

It gives you the principles you need — clarity, consistency, integrity, simplicity — without the dogma. It trusts you to apply judgment. It doesn’t mandate a sign convention; instead, it says pick one, document it, and be consistent. It doesn’t tell you which column to start your time series in; it tells you to make your model navigable and logical.

It’s the difference between being given a set of laws and being given a set of values. Both can produce good outcomes. But only one of them survives contact with the messy, varied, unpredictable reality of building financial models for real businesses.

The Bottom Line

FAST was a useful contribution to financial modelling when it first appeared. The idea that models should follow consistent principles was genuinely important, and FAST helped normalise that conversation.

But the standard has calcified. Its rigid rules, its backwards sign convention, its incomplete sections, and its failure to evolve have left it behind. The ICAEW Financial Modelling Code does what FAST tried to do, but does it better — with more flexibility, more completeness, and more intellectual honesty about the trade-offs involved in building real models for real decisions.

If you’re setting a modelling standard for your organisation, read both documents. Then use the ICAEW Code. Your models will be better, your auditors will be happier, and you’ll never have to explain to a client why their £50m cost just appeared as a positive number.

If you want to read the source material yourself: the FAST Standard is available on their website, and the ICAEW Financial Modelling Code is a free PDF from ICAEW. Make up your own mind. Just don’t enter your costs as positive numbers.

---

Will Wardle is the founder of Financial Modelling Consultants. He spent years building and reviewing financial models at a Big 4 accountancy firm, where he developed the firm’s peer review methodology and trained 30 modellers in its application. He has since reviewed models built by all four Big 4 firms across sectors including banking, oil and gas, retail, infrastructure, energy, and real estate.