How to Write Fraud and White Collar Crime Fiction: A Guide 2026

Learning how to write fraud and white collar crime fiction comes down to two things: a scheme the reader can actually follow, and a person worth following through it. Financial crime is documentary rather than physical, so your tension has to live in ledgers, filings, and small discrepancies instead of in a body on the floor. That is harder than it sounds, and it is where most drafts quietly fall apart.

Nearly all the craft advice out there for crime fiction stops at the fire, the body, and the detective who will not let anyone sleep. Fraud plots run on different materials: an audit trail nobody has read, a compliance officer with a bad feeling, a spreadsheet that does not add up in the third column, a spouse who has stopped asking where the money came from. What makes the subgenre work is that nobody inside the story realises a normal life has been dismantled, so the damage stays invisible right up until it does not.

The second thing that makes it work is the reader’s complicity. A con works by taking advantage of the expectations people bring to a situation, which is a sentence worth keeping above your desk while you draft. Readers know, on some level, that they are being worked. Keep that knowledge alive and you can get away with a lot of technical detail.

What follows is the sequence I use: premise, motive, cost, research, clue chain, pressure, reveal, revision. Eight stages, each with something you can check before moving to the next one.

What You Need

Six things, and none of them are software. You need a bounded financial premise, a motive with a contradiction inside it, a research base you can cite, a timeline, a map of the evidence, and a written line between storytelling and instruction.

• A bounded financial premise. Bounded means you can say the scheme in one sentence without it becoming an operational description. If you cannot, you have a mood rather than a plot. Naming the type early also tells you which research you owe yourself, because an insurer’s claim process and a fund’s reporting calendar are entirely different reading.
• A character want and a contradiction. Who wants money, and what excuse do they give themselves for wanting it? A fraudster needs a want the reader understands and a justification the reader recognises from their own life. Sympathy for the person, no excuse for the act, is the line that holds. Cross it in either direction and the book stops working.
• A research base you trust. Public filings, enforcement announcements, regulatory guidance, annual reports, and at least one proper conversation with someone who has done the job. Fiction can be approximate; it cannot be confidently wrong, because confident error is what readers notice and then write to complain about.
• A timeline. Every fraud story is really a schedule: when statements close, when audits begin, when bonuses land, when a filing is due. Build the calendar before you build the plot, or you will discover in chapter nine that your character was out of the country on the one day the money moved.
• An evidence and clue map. One page listing what a competent investigator would see, when they see it, and which wrong conclusion it supports at the time. This is the document that stops your twist from feeling like a trick rather than a payoff.
• A written ethical line. A short note to yourself about what you will not put on the page. It is easy to set in advance and surprisingly hard to hold under deadline pressure.

How to Write Fraud and White Collar Crime Fiction Step by Step

Eight stages, in order. Each one has a checkpoint, and the checkpoint is the point. Most fraud drafts collapse because a stage got skipped and a later one papered over the gap.

1. Choose a specific kind of financial wrongdoing

Choose the type of financial wrongdoing at the level of consequence and discovery, not at the level of method. Embezzlement, an investment fraud, market abuse, insurance fraud, invoice manipulation, romance and investment scams, identity theft, and fraud committed inside a community institution all produce different victims, different regulators, and different discovery processes. The type sets your setting, your antagonist’s job title, and the shape of your paper trail before you write a single scene.

What makes a premise distinctive is rarely the scheme itself. It is the betrayal at the centre of it: a business partner, a spouse, a mentor, an entire retirement fund. Give the story an institution the reader already has feelings about, a school, a charity, a family firm, a small pension scheme, and the stakes stop being abstract.

Fraud type	Usual motive	Who it hits	How it surfaces	What it gives the story
Embezzlement	Status, debt, a slow leak of self-justification	Employer, clients, a small team	Audit gap, expense pattern, a colleague noticing	Intimacy inside an institution, one person’s lie against process
Investment fraud	Greed, or faith in someone the reader already distrusts	Private savers, pensioners, a whole community	Withdrawals nobody can explain, returns that arrive on time	Wide victim chorus, a slow, patient reveal
Market abuse	Ambition, and the certainty nobody is watching	Retirement savers, a company’s staff	Pattern in trade timing, a colleague’s question, a filing mismatch	Insider access, a lawyer protagonist, moral fog
Insurance fraud	Desperation dressed up as cleverness	Shared premiums, a claim handler	Inconsistent statements, a claim that is almost too tidy	A sympathetic fraudster and a close reader, domestic noir shape
Invoice and procurement fraud	Quiet accumulation, nobody feels it	A supplier, a charity, a small business	Duplicate payments, a vendor that only exists on paper	Systems as villain, easiest to dramatise through a spreadsheet
Romance and investment scams	Connection, then extraction	One person, usually alone	A pattern in the relationship, a request that breaks pattern	Deepest emotional stakes, hardest to make non-repetitive
Identity theft	Reinvention, or plain opportunism	An individual, then a stranger’s life	Applications in two places at once, a credit file nobody expected	Slow-build dread, wrongful suspicion
Community or affinity fraud	Belonging first, greed later	Members who trust the organiser personally	Missing records, a new treasurer, someone who stops attending	Betrayal inside friendship, the worst kind of fallout

2. Build the motive from pressure and contradiction

Build the motive out of two forces pulling in opposite directions: a concrete need and a self-justification. A widowed treasurer chasing returns for her clients is not the same character as a man who wants a boat, and the second one is much harder to write sympathetically.

Give the fraudster a threshold scene: the exact moment the line moves, usually small and usually private. Then let the justification grow one step at a time, the way it does in real life. The first breach is survivable, the second is a pattern, and by the tenth the character has built a whole personality on top of it.

One reader in a comment thread under a long essay on con artists made a distinction worth stealing: some cons are done for greed, others for seduction, self-protection, or performance. A fraudster who is protecting something, or performing a version of themselves they cannot otherwise afford, is a different engine from a simple thief.

3. Decide what the scheme takes from the story

Decide who pays the most, and write that person’s life in detail before you write the fraudster’s spreadsheets. Money in fiction is abstract until it is attached to a specific person losing a specific thing: a second mortgage, a daughter’s tuition, a retirement account that was supposed to be enough.

Map the theft onto something the reader can feel. Money taken equals security lost. Status bought with someone else’s loss equals an identity that cracks under inspection. Security bought that way equals other people now carrying the risk. Trust betrayed means the story has damaged something it will not easily repair, which is exactly where a third act gets its weight.

4. Research institutions, systems, and consequences

Research the process and the aftermath, not the method. What a reader really wants to know is how a suspicion gets raised, who reads it, what makes someone escalate, and what it costs when they do. That sequence, described accurately, will sound like expertise to almost any reader.

Work from public record: regulatory filings and company annual reports, enforcement announcements and settlement orders, indictments and sentencing records, professional guidance, and the trade press a sector actually reads. Then talk to one practitioner if you can. Ask what an alert looks like when it first arrives, who has authority to open an investigation, and what the paperwork means to someone outside the department.

Hold the ethical line you wrote earlier. Keep method detail at the level of what a person needs to know in order to be suspicious, not what a person needs to know in order to do it. One sentence tells you which side of that line you are on.

5. Design the investigation as a sequence of discoveries

Design the investigation as a chain of discoveries, not as a solution arriving at the end. Each chapter should contain one real clue, planted plainly enough to survive rereading, and one wrong explanation that is true but irrelevant. That is the whole difference between a fair twist and a cheap one.

Financial crime gives you an unusually good red herring: a decoy transaction, a genuine discrepancy with an innocent cause, a second person who looked bad for the right reason. Put a real clue next to a fake one and let the reader draw the wrong conclusion themselves. A reader who feels they were outsmarted once will forgive almost anything. A reader who feels cheated will not come back.

6. Structure pressure around deadlines and incomplete information

Build your scenes around decisions that cannot be postponed. An audit that has started. A filing due on Friday. An internal review someone senior has requested. A family obligation. A disclosure nobody can postpone without losing something permanent. That is where jeopardy lives when nobody is holding a weapon.

Vary the texture, though. Two counting-the-money scenes back to back will flatten the reader faster than any plot hole. Alternate a scene of pressure with a scene of quiet dread, let some scenes begin after the decision has already been made, and hold back one piece of information the viewpoint character has but does not yet understand. The gap between knowing and understanding is free tension.

7. Write the revelation as a character event

Attach every financial fact to a human one. The number matters because of what it proves about a marriage, a friendship, a career, a man’s account of himself. If a paragraph of explanation can be deleted and the scene still works, delete it.

Good revelation scenes change the temperature in the room. Somebody denies, somebody forgives too quickly, somebody becomes cruel, somebody asks a practical question that is really an accusation. That is where the plot turns, not the arithmetic.

8. Revise for credibility, fairness, and narrative payoff

Revise in passes rather than all at once. First pass, terminology and process: is every term used correctly, and would a person in that job actually act that way? Second pass, timeline: does the arithmetic of dates, statements, and holidays hold up? Third pass, motive: is the pressure specific enough and dated enough to explain the choice?

Fourth pass, fairness: could a careful reader re-read chapter four and see the key clue? Fifth pass, payoff: does the final choice answer the question the book opened with, or does it just stop? And one last pass for the ethical line you wrote before you started.

Common Mistakes

These seven flaws account for most of the fraud manuscripts I have seen go wrong, and each one has a straightforward fix.

• The villain is a face, not a person. Fix: give your fraudster a competence and a private win. People with a good life to protect are far more frightening than people who simply want more.
• The motive is “only money”. Fix: attach the want to a specific, dated pressure. A mortgage, a diagnosis, a child’s fees, a redundancy notice. Money alone is not a reason; money in a life is.
• Technical overload. Fix: give the reader three facts per scene, one figure, one document, one comparison. Everything else can go into a list the character reads and reacts to.
• The clues are either obvious or invisible. Fix: aim for the narrow band between. A clue spotted in chapter four wastes the book; a clue nobody could re-find is a cheat. Plant it plainly and then distract with something true.
• A consequence-free ending. Fix: decide what justice actually costs. Restitution without admission, an admission without restitution, a plea deal that saves one person and abandons another. Pick one and let it cost.
• Factual overconfidence. Fix: verify every specialised claim against a source, then hedge in the prose where you are still unsure. Being accurate is the goal; being exhaustive is not, and readers are not taking notes.
• Operational detail that helps rather than reveals. Fix: cut any passage that would function as an instruction. If it reads as a method, it is one. Retain the suspicion, the evidence, and the consequences.

Quick Tips Before You Revise

Five checks, in this order, catch most of what matters.

• Read the draft without your research notes open. Anywhere you stumble, the manuscript is doing the author’s work instead of the character’s.
• Verify every specialised claim against a named source, and delete the ones you cannot support.
• Strengthen the emotional consequences. If the money is still a number, the scene is not finished.
• Confirm the clue chain end to end, chapter by chapter, with one real clue per chapter and every red herring still true.
• Cut any passage that is interesting only because it explains how the scheme works. If the scene dies when you remove the explanation, the explanation was the scene.

Frequently Asked Questions

Conclusion

Start with one decision, not a scheme. Write a single sentence naming who wants what, which pressure makes the want unbearable, and which relationship makes the consequences land. Then work outward: the type of fraud, the research you owe yourself, the clue chain, the deadline that will not wait. That is the whole method, and it starts on a blank page in about ten minutes.