The 2020 card ban, priced as a cost line

April 2020 is the line the British market drew under borrowed stakes: since then, an operator holding a Gambling Commission licence may not accept credit cards for gambling, online or offline. The ban is usually described as a protection, and it is one, but this page reads it as a price line. Before it, a stake could ride on a lender’s money as easily as on a debit card. After it, a licensed cashier runs on debit rails, and the decision to borrow for play sits where the money sits, with the reader rather than the cashier.

What the ban did not do is take the cost of borrowed money out of the world. Card rails still run at cashiers outside the British licence, and where they run, the borrowed-money exposure the ban lifted from the licensed market behaves exactly as it did: a stake on a lender’s money, interest gathering from the day it moves, and the game’s own edge stacked on top of the credit line’s. The phrase credit card casino therefore describes a real part of the market, just not the licensed part of it, and the honest response to the phrase is arithmetic rather than alarm.

That arithmetic is the whole job here. A stake costs what it costs either way: paid from money already earned, the bill is the stake itself. Paid from borrowed money, the bill is the stake plus the price of the borrowing, and the second figure is the one no headline prints. The sections below state both lines in the open, run the chain on a worked deposit, and set the card’s borrowing cost against the two nearest alternatives, an arranged overdraft and a planned withdrawal from savings, so the spread between routes is visible before any money moves.

The assumptions behind the sums

The sums on this page run on stated inputs, and the first is the wagering multiple: forty times, taken as the working figure because offers of this kind typically ask it. The worked deposit is £30, the match is 100%, the cap that would trim a larger deposit sits at £150, and the house edge used to price the play is 3%, a working mid-point across a typical slots mix. Every one of those numbers is an assumption in the open rather than a quote from any operator, and each can be swapped in the calculator for a reader’s own.

The borrowing lines carry assumptions of their own, stated just as plainly. The card route is modelled at an all-in cost of 24% a month on a running balance: an assumption, deliberately round and deliberately heavy, defined to cover interest charged from the day the money moves plus a fee layer on top, and it is not a quote of any card issuer. The overdraft route is modelled at about £5 a month per £100 used. The savings route is modelled at about £1, the forgone interest on £100 for a month plus the small friction of moving money that was already the reader’s.

Two rules hold the model together. First, the deposit is committed cash: the £30 leaves the pocket whether the pack clears or not, so it enters the bill in full. Second, the borrowing cost is charged on the borrowed sum for the month it is out, independent of the game’s edge, which is why the two lines add rather than interact. The rate matters at whatever balance a reader actually runs, and the chart later in the page scales the same assumptions up to £100 so the three routes sit beside each other at one size.

The arithmetic of a borrowed deposit

Work the chain on the £30 deposit and every step is a sum a reader can redo. The first step sizes the bonus: a £30 deposit at a 100% match pays £30, and the £150 cap trims nothing. The second turns the bonus into turnover: at forty times, 40 × £30 = £1,200 of required play. The third prices that turnover: 3% × £1,200 = £36 of expected loss spread across the clearing. The fourth adds the deposit to the bill, because the money leaves the pocket whatever the clearing does: £30 + £36 = £66, the expected cash cost of the exercise.

The fifth figure is an exit rather than an entry: the walkaway, the balance at which banking what is left beats playing on. With roughly half the turnover still to run, the edge still to come is about £18, which puts the exit at the deposit back plus whatever edge is left, about £48. A sixth line prices the pace: first-hour cost, the deposit plus the edge on an hour’s play, which at 600 bets of £0.50 is £300 of turnover and £9 of edge, or about £39 in the first hour.

The borrowing line sits on top of all of that and changes none of it. If the £30 is drawn on a card at the modelled rate, carrying the borrowed sum for a month adds 24% × £30 = £7.20, and the same pack priced on borrowed money lands near £73.20 all-in. The game’s edge never noticed the difference; the lender did. That separation is the point: the wagering requirement and the house edge price the play, the credit line prices the money, and a bill that quotes only the first is missing the second.

Change any input and the chain moves with it, proportionally: a 50× label on the same £30 bonus lifts the turnover to £1,500 and the expected edge to £45, a 0.5% table game cuts the edge line to £6, and a £60 deposit under the same £150 cap doubles the base and everything that scales from it. The borrowing line moves only with the borrowed sum and the rate, which is why it is modelled as its own line rather than folded into the multiple.

Model output at the stated assumptions, not a quote of any promotion, and every row reruns from the calculator. Swap the card rate for a reader’s own monthly figure and only the borrowing line moves; the edge, the cash cost and the walkaway stay independent of where the deposit came from.

Scaled to £100 so the routes compare at one size, the model’s three lines are the chart below: about £24 a month on a card at the assumed rate, about £5 a month on an arranged overdraft, about £1 for a planned withdrawal from savings. The spread between the top and bottom bars is this page’s argument in one picture, and the next section reads it bar by bar.

Borrowed stakes at 24% a month

The card bar prices first, because it is the one that explains why the ban was worth drawing. At 24% a month, £100 of borrowed stake money costs about £24 in the model’s first month, and the cost compounds while the balance stays out: held a second month at the same rate, the modelled cost passes £53 on the same hundred. Interest of that kind gathers from the day the money moves, it is indifferent to whether the play won or lost, and it lands monthly whatever the clearing did. A £24 modelled month on £100 is close to the whole stake of a modest session, paid to a lender rather than risked in a game.

The overdraft bar is the middle route, modelled at about £5 a month per £100 used. It is cheaper than the card on these assumptions, but it shares the card’s worst property, which is that the cost runs on time rather than on play: the meter ticks whether the reader plays or not, and money left sitting in an overdraft has a way of becoming standing money. The savings bar is the floor: about £1, the forgone interest on a hundred moved deliberately from a pot that was already the reader’s, with no lender, no fee layer and no clock beyond the plan itself.

Read together, the bars price the difference between owning a stake and renting one. The spread between the card line and the savings line, about £23 a month per £100 on the model, buys nothing inside the game: the house edge is identical whichever rail the money arrived on, so every pound of that spread is pure carrying cost. The honest use of the chart is to price a route before money moves, and the honest reading of it is that the cheapest line on it involves no lender at all. Nothing here recommends borrowing for play; the model’s only job is to make the cost visible before it is paid.

The licence gap that locks the cost in

The offers covered by this site hold offshore licences and hold no UK Gambling Commission licence. That fact is what keeps the card bar on the chart reachable. A British licence has barred credit-card gambling since April 2020, so at a licensed cashier the route the model prices at £24 a month does not exist: the rails are debit cards, bank transfer and the wallets that serve the licensed market. At an offshore cashier the card rail can still be live, and with it the borrowed-money cost line this page spends its arithmetic on.

The gap runs wider than one rail. A licensed operator verifies identity and age before play, and most British gambling starts at 18. It must offer deposit limits, time-outs and reality checks as conditions of its licence, and it must say how customer funds are protected. A dispute that cannot be settled goes, free of charge, to an adjudicator approved by the Commission, IBAS being one of the approved schemes. An offshore licence offers none of that to a British player as of right: no register entry a reader can check in minutes, no national self-exclusion coverage, no free binding adjudication, and no funds rule beyond what the operator volunteers.

This section locks the one before it because the chart’s card bar is only reachable where a card rail exists, and whether a rail exists tracks the licence. Price the gap at the balance a reader would least like to owe a lender while a withdrawal stalls: that is the exposure the two sections price together, borrowing cost on top and a dispute route that does not exist underneath.

Cashing out on borrowed money

Getting money back out is where borrowed stakes hurt most, so the payout window and the reversal question belong in the same breath as the interest clock. Licensed cashiers clear withdrawals within a published window and have largely removed reversal: once a withdrawal is asked for, it proceeds. Offshore cashiers commonly leave a reversal window on, a stretch of time in which a requested withdrawal can be pulled back into playable balance. On money that is already the reader’s, that window is a temptation. On borrowed money it is the sharpest edge on the page, because the interest meter keeps running while the money sits reachable.

The model’s payout range, one to three days, is the category-typical band for the rows in the comparison table, and the window worth pricing is the queue in front of the published figure. A pending withdrawal of borrowed money is a running cost: at the modelled card rate, a £100 balance left waiting adds to the £24-a-month line for as long as it waits. The difference between a cashier that pays in a day and one that pays in three is small against a bonus and large against an interest charge that does not pause for the queue.

Two reads close the section. The rail a deposit travelled on tends to shape the rail a withdrawal can use, and rails keep their own schedules, which is worth knowing before the entry rather than after. And checks at offshore cashiers tend to run at withdrawal rather than before play, when the money is already inside; a licensed cashier verifies before play. Neither read argues for moving money. Both argue for pricing the exit before the entry, which on borrowed money is the difference between a plan and a meter.

The comparison table, read by rails

The comparison table sets the market’s usual ranges beside the British baseline, and on this page the payment rails column does the work the wagering column does elsewhere. A British-licensed cashier has been barred from taking credit cards since April 2020, so the licensed rows show debit cards, bank transfer and regulated wallets, and never a credit card. The offshore rows lean on crypto kinds and e-wallets, and the card rails some of them still run are the reason this page exists: that is where the borrowed-money cost line stays live.

The welcome column reads as kinds, and the ranges on this page anchor on one family: matches from 50% to 200% with ceilings to £250. The cap and the minimum deposit together set the biggest claim available: a pack paying 100% to a £150 ceiling with a £10 floor is a £10 test at the bottom and a £150 claim at the top, and everything between scales linearly. The wagering band to expect runs forty to sixty times, and on this page’s model every extra ten times on a £30 bonus adds £9 of expected edge.

Payout windows in these rows run one to three days at the typical pace, and minimum deposits cluster between £10 and £20, which sets the smallest test a balance can fund. The licence label column of the comparison table is the one that does not age: a British entry takes minutes to confirm on the Commission’s public register, while an offshore label stays a claim until the register it names backs it. Where a row and the register disagree, side with the register. Where a row and the sums disagree, rerun the model with the row’s own numbers.

Comparison rows are category-typical ranges compiled from operator-published terms at the time of writing. UKGC rows are the British baseline shown for contrast; none of them are claimable at the offshore offers.

Frequently asked questions

Six answers cover the questions readers bring to this page, and each reuses the working figures from the sections above.

Can you still gamble on a credit card in Britain?

Not at a British-licensed operator. Since April 2020, licensees have been barred from accepting credit cards for gambling, online and offline, which is why the licensed rows in the comparison table show debit rails and no credit cards. The phrase credit card casino survives as a description of cashiers outside the British licence, where card rails can still run. What follows on this page is a cost model of that fact, not a route map into it.

What does borrowing £100 of stake money actually cost?

On the page’s stated assumptions, about £24 in the first month on a card at the modelled all-in rate, about £5 a month on an arranged overdraft, and about £1 for a planned withdrawal from savings. The card figure is an assumption in the open, not a quote of any issuer, and a reader with a real statement can swap their own rate into the same arithmetic. Held a second month at the same rate, the modelled card cost passes £53 on the same hundred.

Does a wagering requirement cost more when the deposit is borrowed?

Not through the game. The chain is identical either way: a £30 deposit at a 100% match and forty times makes £1,200 of turnover and £36 of expected loss at a 3% edge. Borrowing adds a separate financing line, modelled here at £7.20 for a month on £30, which lifts the all-in figure to about £73.20. The edge and the interest never interact, which is why the model prices them as two lines that add.

Why do the British rows in the comparison table show no credit cards?

Because their licence forbids it. Accepting credit cards for gambling has been off the British rulebook since April 2020, so those rows run on debit cards, bank transfer and regulated wallets. The rails cell is a quick read on which rulebook a row was built for: a credit card there marks a cashier outside the British regime, and with the card comes the borrowed-money cost line the rest of this page prices.

What is a reverse withdrawal window, and why does it matter here?

A stretch of time after a withdrawal is requested in which the money can be pulled back into playable balance. Licensed cashiers have largely removed it; offshore cashiers commonly leave one on. On the reader’s own money the window is a temptation. On borrowed money it is a running cost: the interest meter does not pause while a withdrawal sits reversible, so every extra month a balance stays out adds to the model’s card line.

How solid is the 24% a month card figure?

It is an assumption, stated in the open, and deliberately heavy: an all-in figure defined to cover interest charged from the day the money moves plus a fee layer on top, rather than a quote of any card issuer’s headline rate. Its job is to make the direction and rough size of the cost visible. The chart’s other two bars, the £5 overdraft line and the £1 savings line, are assumptions of the same kind, and swapping in a reader’s own rate moves only the card bar.