Table of Contents
Bankroll management arbitrage is not really a sizing problem. It is a placement problem. A bankroll for arbitrage is a liquidity fund, not a stake fund, and the edge only pays when the money is already sitting in the right bookmaker’s account the moment an arb appears.
People size an arbing bankroll the way they size a normal betting bankroll, against variance and losing streaks. Wrong problem. A 2% arb does not care about variance. It cares whether you have 300 EUR free on one side and 250 EUR free on the other, right now. So bankroll management for arbitrage is mostly about distribution, timing, and keeping fixed costs small. I run about 8,000 EUR split evenly across ten bookmakers, and the split does more work than the total ever did.
If the mechanics are new to you, the arbitrage betting article on Wikipedia covers them in three paragraphs, which is roughly all there is.
Bankroll management arbitrage is about float, not variance
Variance is the wrong villain. An arb locks its result the moment both legs are matched. What actually goes wrong is mechanical: one side wins and the money lands in the book you wanted it to leave, or both legs are placed and one balance is now too thin for the next opportunity. The bankroll is fine. Its location is not.
Two forces drain balances unevenly:
- Arbs stake unevenly. A 60/40 split between the two legs is common, so one side of your estate empties faster than the other. After a few good days the soft books are dry while the sharp ones are full.
- Withdrawals are the one thing soft books notice fast. Pull profit out of a soft account every time it swells and you will meet the risk team far earlier than you would like. I withdraw from soft and easily limiting bookmakers as little as possible for exactly that reason.
Value betting has the same shape with more noise. The edge is real but lumpy, and a balance sitting in the wrong account cannot bet either. The fix is the same: design where the money lives before you worry about how much there is.
The break-even maths: what a bankroll must clear
The first rule of bankroll management arbitrage: start with the costs, because they set the floor. A decent scanner runs 69 to 99 EUR a month on annual billing, depending on the package (see RebelBetting’s pricing for one working example), plus card deposit fees, exchange conversion costs, and the occasional withdrawal charge. Beginners forget this part. They size a bankroll at 300 EUR, run it through a 69 EUR subscription, and read the first month as proof that arbing does not work.
The working formula is short. Monthly profit is roughly: arbs placed, times average total stake, times the arb margin. Call it a 2% margin for pre-match arbs as a planning number; live arbs pay more and disappear faster.
Worked example, not a promise:
| Setup | Bankroll | Books | Typical stake per arb | 100 arbs a month at 2% | Software fee | Left over |
|---|---|---|---|---|---|---|
| Testing the waters | 500 EUR | 3 | 40 to 60 EUR | about 100 EUR | 69 to 99 EUR | roughly nothing |
| Side income | 2,000 EUR | 5 | 150 to 250 EUR | 300 to 500 EUR | 69 to 99 EUR | a few hundred |
| Serious | 8,000 EUR | 8 to 10 | 300 to 600 EUR | 600 to 1,200 EUR | 69 to 99 EUR | the point |
The pattern is blunt. Under about 1,000 EUR, the scanner fee eats most of what the arbs pay. Between 2,000 and 5,000 EUR it becomes a proper side income. Past that, each individual arb stops mattering much on its own, which is the same thing as saying the float stops being the bottleneck. Can you start with 500 EUR? Yes, and many do. Just treat it as tuition and expect the first two months to roughly pay for the software.
How to split a bankroll across bookmakers
In bankroll management arbitrage, splitting is a strategy decision, not an accounting one. You are balancing two goals that pull against each other: keep every account looking like a normal recreational bettor, and have enough cash in enough places that no arb dies for lack of a balance.
- Even is fine to start. My own split is deliberately even across ten books. It is the least clever option and also the hardest to regret.
- Feed the books you want to keep. A soft book with a dead balance starts asking questions when you suddenly reload it months later. A steady trickle reads as normal.
- Let sharp books carry the buffer. Pinnacle and the brokers do not limit winners, so surplus cash parked there is cash you can always move. It is the only place where extra money sits for free.
- Mind the minimums. Every book has deposit and withdrawal minimums, and some charge for both. Ten accounts at 50 EUR each is not a 500 EUR bankroll, it is ten slow leaks.
So how many books? Enough that no single balance has to carry a whole arb, few enough that each account keeps a usable balance. Five to ten is the common working range, and every book you add should earn its deposit minimum.
The float trap, and the three ways out
The trap is simple: the best arb you have seen all week needs 400 EUR on a soft book where 40 EUR is left. It comes up often enough that the fix should be routine, not improvised. Three moves, in order of preference:
- Move from a book that does not punish it. Pinnacle, BetInAsia, and other sharp books or brokers do not limit winners, so withdrawing from them to rebalance costs nothing but the transfer time. This is the first lever, and with a few thousand EUR of buffer it covers most shortfalls.
- Top up temporarily from your bank. Card deposits land in seconds. When an arb is good enough and there is no time to move money between books, deposit the shortfall, take the arb, and unwind it later. Keep this deliberate and occasional, because a pattern of constant small top-ups is its own signal.
- Withdraw from the soft book as the last resort. It works, it is slow, and it is the action most likely to draw attention. With a bigger bankroll I rarely need it, which is one more reason to build the bankroll first.
When to top up, when to withdraw, when to do nothing
Default to doing nothing. Every movement of money costs something, in fees or in attention. The routine that holds up:
- Reinvest by default. Profits stay in the accounts until you stop. Withdrawing on a schedule converts a compounding float into a flag history.
- Top up before it is a problem, not after. The time to send money to a book is when its balance drops below roughly a week of your usual stake volume, not when the balance-too-low alert fires mid-opp.
- Keep a reserve outside the soft books. A few hundred EUR parked at a sharp book, or simply reachable on your card, turns a balance shortfall into a two-minute pause instead of a missed arb.
The quiet drains
None of these lose money dramatically. All of them lose it steadily:
- Mixing living money with arbing money. Arbing is close to guaranteed and EV betting is not, but neither survives a forced withdrawal because rent is due on Friday. Decide what you can lock up before you deposit it.
- Too many accounts at too little each. Minimum deposits, minimum withdrawals, conversion fees, and balances too thin to carry a leg. Fewer books, properly funded, beats more books, thinly funded.
- Forgetting the fixed costs. Software subscription, transfer fees, withdrawal charges, and the card’s FX spread all come off the top. Count them before you count the profit.
- Ugly stakes. A 47.13 EUR leg reads as exactly what it is. Round to sizes a recreational bettor would plausibly choose, and let the stake calculator do the precise maths inside that shape.
The honest summary: most arbers do not fail because the bankroll was too small. They fail because it sat in the wrong accounts, moved too often, or was never big enough for the fixed costs to stop mattering. Size it against the maths, split it on purpose, and then let it sit.
If you want to stress-test a split before funding it, the arbitrage calculator does the stake maths, the software directory shows what subscriptions actually cost, and what arbitrage betting is covers the wider frame.
FAQ
How much money do you need to start arbitrage betting?
Treat 500 EUR as the floor for a test and 2,000 EUR as the floor for a serious attempt. The scanner fee decides it: under about 1,000 EUR total, the fixed costs take most of what the arbs pay.
Should you withdraw your arbing profits?
Rarely. Keep profits in the accounts so the float grows, rebalance from sharp books where withdrawal is free, and keep soft-book withdrawals rare because they are the fastest way to get flagged.
What do you do when the money is in the wrong account?
In order: move it from a sharp book that does not limit, top up from your bank if the arb is good enough, and only then withdraw from a soft book. The first option should cover most cases if you keep a buffer parked at a sharp book.
How many bookmakers should you split a bankroll across?
Good bankroll management arbitrage splits the estate so no single balance has to carry a whole arb, but few enough that each account keeps a usable balance. Five to ten is a common working range, and every book you add should earn its deposit minimum.