Arbitrage Betting in New Zealand: What It Costs a Punter in 2026

Updated October 2026

Licensed

Available in US

Fast payouts

18+ Only

30 October 2026 · New Zealand Gambling Commission register, Racing Industry Act 2020 (as amended 2025) and TAB NZ / Betcha own terms.

Arbitrage betting looks like a money-print on paper: back every outcome of a match at different bookmakers, lock in a margin whatever the result. The arithmetic checks out. The cost does not show up on the slip. It shows up in the stake it demands, in the speed with which bookmakers cut limits on accounts they suspect, in the fact that New Zealand law leaves only one lawful set of bookmakers to arbitrage between, and in the absence of a New Zealand complaints route once an offshore brand sits on the other side of a payout row.

That is the real shape of arbitrage in New Zealand in 2026. The mechanics below describe what an “arb” actually is, how it is sized, what it costs in turnover, and what happens when the only two lawful New Zealand bookmakers are owned by the same parent — a setup that makes pure price arbitrage mathematically impossible on domestic rails alone. The offshore brands listed further down are not recommended: the Department of Internal Affairs treats them as unlawful to serve New Zealand residents, no New Zealand register or dispute body reaches them, and bookmakers who spot arbers do not wait politely before acting.

If the reader came here hoping for a profit system, the short answer is in the responsible-gambling section that follows: the arithmetic holds for large bankrolls only, the margin bookmakers keep is the same margin that makes a misquote worth chasing, and the accounts that try this get closed.

Responsible Gambling and the Cost of Chasing an Arb

The fastest way to size the real cost of an arbitrage setup is to think about what a bettor loses before a cent reaches the bookmaker. A 2012 study by Franck, Verbeek and Nuesch, summarised on the public Wikipedia entry for arbitrage betting, found that ninety-eight per cent of arbitrage opportunities return less than 1.2%. Read that again: fewer than one in fifty arbs pays more than 1.2%. The arithmetic the marketing material promises — “lock in 5% no matter the result” — is the rare exception, not the working week.

Two costs follow directly from that finding. First, the bettor must move large sums to capture a thin margin: turning a 1.2% return into actual money means staking enough that the dollar figure is meaningful, and a stake that large is the exact pattern bookmakers look for. Second, that 1.2% is gross, before any of the deductions the bettor has not priced in. A bet that voids, a price change between placement and settlement, a stake limit kicked in on one leg of the trade, a withdrawal fee on the winnings side, or a bookmaker who simply decides the second leg is a “palpable error” and grades it void — each one erases the margin and leaves the bettor holding the loss side. The opportunity cost of capital tied up across several accounts while waiting for arbs to appear is itself a line item.

Bookmakers react on a known timetable. The Wikipedia entry notes that bookmakers usually detect arbitrage betting quickly, that maximum wager limits on such accounts often drop sharply, and that in the past half of an arbitrage bet may be cancelled or the account closed. “Quickly” is not a marketing euphemism here: dedicated software, cross-market odds feeds and shared trader communities mean that an unusual combination of bets on correlated outcomes gets flagged within hours, not weeks. The bettor’s stake is at risk of being voided mid-trade; the bookmaker’s house rules almost universally reserve the right to do so, and the right is exercised.

New Zealand adds a further cost that does not exist in jurisdictions where arbitrage is openly contested across dozens of licensed bookmakers. TAB NZ and its sister brand Betcha are the only two operators lawfully allowed to take online sports and racing bets from people in New Zealand since the 28 June 2025 legislative change. They are not two competitors. They are two brands of one operator group, run under a long-term partnership with Entain. Arbitrage between them is, on a price level, arbitrage against one bookmaker’s two trading desks — and any internal offsetting they apply is theirs to decide.

The protective layer in the lawful New Zealand setup is real, and it is part of the cost calculation a reader must make. The TAB NZ app offers deposit limits, time-outs and self-exclusion, and lists the free Gambling Helpline on 0800 654 655 with a text line on 8006. The Problem Gambling Levy funds harm-minimisation work across the racing and sports-betting sector, and Safer Gambling Aotearoa operates alongside the helpline. These tools exist because the harms of chasing losses in a thin-edge environment are recognised by the regulator. Anyone who notices the urge to “double the stake to recover” is reading the right sign: stop. The Money Management section below addresses what to do about it.

That is the headline cost: 98% of arbs pay less than 1.2%, large stakes are unavoidable, bookmaker response is fast, and the only lawful New Zealand provider is a single operator with two skins. Anyone still reading either wants the mechanics or wants the offshore comparison. Both come next.

Crypto, Anonymity and the Limits of Pseudonymous Rails

Crypto gets pitched as the natural fit for arbitrage betting: instant settlement, no card chargebacks to track, accounts that look interchangeable across brands. The pitch collapses on contact with New Zealand law. The Racing Industry Act 2020, as amended in 2025, prohibits any person other than TAB NZ from offering racing or sports betting to a person in New Zealand, regardless of where that person is resident or incorporated. The Department of Internal Affairs told unlicensed providers to withdraw with immediate effect after the amendment came into force. The form of payment does not change who is on the other side of the bet.

Several offshore brands in the comparison below do accept cryptocurrency. Roobet, for instance, is operated by Raw Entertainment B.V. under a Curaçao licence and brands itself around crypto rails. The licence gives it no right to take bets from people in New Zealand. No New Zealand complaints route or harm-minimisation framework reaches a Curaçao-licensed operator: a dispute over a voided leg, a confiscated bonus balance or a withdrawal that never lands has no New Zealand venue. The currency choice is downstream of the legal status, not the other way round.

The cost in dollar terms also runs against the crypto pitch. Crypto deposits are typically one-way until the operator agrees to convert back to fiat for a withdrawal. Conversion spreads, network fees and minimum-amount thresholds all take a cut that the “no fees” line in the marketing does not mention. Where the operator pays withdrawals only in the same coin deposited, an arb that needs to move money fast between two of them is paying conversion costs twice — once on entry, once on exit. A bookmaker that pays out in a different token than the deposit adds an exchange-rate risk on top.

Privacy is the third leg of the pitch and the weakest. “Pseudonymous” is not anonymous. Every crypto ledger is public, every address cluster eventually links back to a withdrawal or an on-ramp, and any operator serious about compliance will trace funds on entry. New Zealand’s TAB NZ still requires a verified New Zealand address and ID to open an account. The offshore brands that pitch anonymity apply their own KYC at withdrawal time, which is exactly when a settled arb needs the money to move. An account that gets reviewed at withdrawal is an account that does not pay out on the bettor’s timetable.

The plain version: in New Zealand in 2026, the lawful bookmaker requires identity verification and accepts standard payment methods, and no offshore brand changes the legal position by accepting a different currency. Anyone comparing crypto rails as an arbitrage route is comparing the wrong things.

Prediction Markets, Tipsters and Statistical Arb

The marketing around arbitrage often blends it with predictions, tipsters and “statistical arbitrage” as if they were the same activity. They are not. A tipster service sells opinions about outcomes; arbitrage sells a structure that ignores outcomes. The two merge only when the tipster’s confidence is so high that one leg of the arb can be replaced by a strong opinion, and that is no longer arbitrage — it is a position, with the corresponding risk of being wrong.

Statistical arbitrage, in the financial sense, is a portfolio of small edges across many correlated instruments that earns a tiny positive expected return per trade, with controlled risk on the portfolio. Sports “statistical arb” services often borrow the phrase without the discipline. They quote a model that says one team is mispriced by 3%, but they have no instrument to short the other side, no defined entry and exit, and no way to know whether the 3% is an edge or a quote artefact. The bettor pays subscription fees, follows the model’s “arb” picks, and is then subject to the same account limits any other arber faces when bookmakers recognise the pattern.

The cost the reader must price in is straightforward. A tipster subscription of, say, NZ$50 a month costs NZ$600 a year before any bets are placed. An arb service that charges a percentage of bankroll turns a 1% edge into a 0.5% edge. A model service that quotes “expected value” lines without a definable arb structure is selling forecasts, not arb screens, and the buyer pays for the right to be wrong. None of these costs appears on the bet slip.

The honest version: a tipster is a content product. A statistical-arb label on a sports model is branding. The mechanical arbitrage — buying every outcome with a stake that scales to the price — is described in the bet-types section that follows. The reader must know which of the three they are paying for, because the cost and the risk differ.

Bet Types, Odds and the Shape of an Arbitrage Bet

An arbitrage bet is not a bet type in the sense TAB NZ’s betting rules use the term. It is a structure across several bets at different bookmakers, where the prices cover all mutually exclusive outcomes of an event and the implied probabilities of those outcomes sum to less than one. The implication matters: any one bookmaker can quote the same market at a margin that is profitable for them, and the arbitrage appears only when two or more operators disagree on price.

The arithmetic is not complicated. Take a two-way market (a tennis match, a basketball handicap reduced to two sides) priced at decimal odds of 1.90 and 1.90. The implied probabilities — one divided by the odds — are 0.5263 each. Sum them: 1.0526. The sum is greater than one, which means a 5.26% margin sits inside those prices. That is the bookmaker’s edge: out of every NZ$1 staked across both outcomes, 5.26 cents is the bookmaker’s expected profit. The other 94.74 cents is returned to the bettors who back the winning side.

Arbitrage appears when two prices for the same mutually exclusive outcomes do not sum to a margin of zero or more, but to a negative one. If one operator offers 2.10 on outcome A and another offers 2.10 on outcome B, the implied probabilities sum to 0.4762 plus 0.4762 = 0.9524. That is below one. A bet on each side at the right stake locks in a positive return no matter which outcome lands. The calculation the page prescribes further down works through the figures.

The bet types that arbitrage most often uses are the two-way and three-way markets. In a tennis match the two players cover all outcomes. In a football match the home win, draw and away win are three mutually exclusive outcomes, and the implied probabilities of all three decimal prices must sum to more than one for a margin. A bookmaker that posts 2.10, 3.40 and 3.60 has a margin around the mid-single digits; the rare misquote or disagreement will post a sum below one, and that is the arbitrage window. The larger the market’s three outcomes, the more bookmakers must disagree before an arb appears.

Each market has its own margin. The margin of one market is not the margin of another. A bookmaker’s 1X2 football margin is a different number from their totals margin, their Asian handicap margin, their player-prop margin or their combination-bet margin. The numbers across the comparison below are quoted where research carries them and not generalised, because the comparison that misreads a 5% football margin as a 5% margin on a tennis two-way has just manufactured an arb that does not exist.

Combination bets — the multi bet, the accumulator — multiply the odds of their legs into one price, and they multiply the implied probabilities into one figure. A four-leg multi at decimal odds of 1.5, 1.7, 1.8 and 2.0 carries odds of 9.18, and the implied probability of the combination is 0.109 — a much thinner chance of winning than any one leg. The bookmaker’s margin across the four legs compounds rather than averages, and arbs in combination bets appear rarely and for tiny edges. The bettor who treats a long multi as an arb opportunity pays a much higher margin than on singles.

Live, or in-play, betting adds timing cost. The price for outcome A moves with the score, the clock and the next goal. A bookmaker’s feed and another bookmaker’s feed do not update at the same instant. The window between updates is where a live arb may in principle appear, and where the bettor pays in latency: the trade must be sized, both legs confirmed, and the second leg placed before the price moves to parity. In practice the window is sub-second on liquid markets and the bettor pays the platform and the network to be faster than the next arber. A “live arbitrage” service that claims a steady stream of in-play arbs is selling a speed advantage the bettor has not paid for.

The cash-out feature complicates the math further. A bookmaker that offers cash-out mid-event is buying the bet back at a price derived from the live market minus their own margin, and the price they offer is rarely the fair value. Treating a cash-out price as one leg of an arb against another bookmaker’s live price is to bet against a sharper counterparty who knows both sides of the trade.

The staking detail is small but consequential. Stakes on the two legs of an arb have to be sized so that the return is identical no matter which outcome lands, and the formula is straightforward: stake A is the bankroll times the price of B divided by the sum of the prices, and stake B is the bankroll times the price of A divided by the sum. With both legs at 2.10, each stake is half the bankroll. With 2.10 on A and 2.20 on B, stake A is 2.20/4.30 of the bankroll, stake B is 2.10/4.30. The maths is mechanical; the cost is that the bettor has to commit the whole bankroll at once and accept that one leg voids or settles differently than priced.

That is the structural picture. The operator comparison below puts real offers beside real live offerings from each brand, and the calculation that follows turns the two-way illustrative example into a number a reader can compare to the cost of bookmaker detection.

The Landscape: How New Zealand Bookmakers Stack Up in 2026

The table below sets out the ten bookmakers the page reviews. Two are TAB NZ brands and the only operators lawfully allowed to take online sports and racing bets from people in New Zealand since 28 June 2025. The remaining eight are brands New Zealand players encounter in offshore marketing and affiliate channels; each is marked not licensed in New Zealand, because the Racing Industry Act 2020 as amended leaves no licensing route for another bookmaker and no licence register for betting exists. The right-hand column records what the operator’s own terms state on the subject of arbitrage and bet limits — paraphrased rather than quoted in length. The wagering account of the table is honest about gaps: where a term is not stated in the operator’s own materials the cell carries the marker, and the prose around the table names the gaps.

The table is not a recommendation. The first two rows are the only options the Racing Industry Act 2020 permits; the rest are not lawful for the operator to offer to people in New Zealand and carry the further cost of account-restriction risk.

The shape of the column set matters. The third column is honest about what is and is not stated in each brand’s terms, because most operators do not publish a clause that names “arbitrage” directly. House rules on voids and limits apply regardless, and the bettor who relies on the absence of an arb clause as a safe harbour is reading the terms for what they do not say. The two New Zealand brands control the same trading operation; an arb between them is an arb against one bookmaker’s two front ends.

Three rows on the table carry a Curaçao licence, one a Comoros licence, one a Tobique Gaming Commission licence (across two related brands), and the rest do not name a licence that applies in New Zealand. None of these licences gives a right to take bets from people in New Zealand, and the licence named is not, by itself, evidence of legitimacy in the New Zealand market. The reader comparing these rows for safety is advised to compare them on the absence of New Zealand regulatory reach, which is the same across the eight.

Overview: What Arbitrage Betting Is, and Is Not

An arbitrage bet, in its cleanest form, is two or more wagers on mutually exclusive outcomes of one event, placed at different bookmakers, where the decimal prices imply a probability sum below one. The lock-in return is positive whichever outcome lands. The bettor pays for the privilege in stake size, in detection risk, and in the frictions of moving money between accounts.

What arbitrage betting is not is a steady income. The Franck, Verbeek and Nuesch study reported in 2012 that ninety-eight per cent of opportunities return less than 1.2%. A 1% return on a NZ$10,000 bankroll is NZ$100 before any deduction; across a year of “steady” arbs at that rate, after voids, account closures and the subscription costs of odds feeds, the figure looks different. What arbitrage betting is not is a system that survives bookmaker attention. The same Wikipedia entry records that bookmakers usually detect arbitrage quickly and act on accounts that fit the pattern. What arbitrage betting is not is a substitute for the responsible-gambling framework: the bettor whose bankroll is the rent is paying the highest cost of all.

In New Zealand in 2026, arbitrage betting is also not something that can be run between lawful bookmakers on the same side of the law. TAB and Betcha are two brands of one operator. Two operators of the size needed to disagree on price enough to produce a sustained arb do not exist in the lawful roster. Any reader who wants to run the structure on offshore brands is doing so against New Zealand law, and the table above sets out the cost that follows.

Legality, Regulation and the Cost of Operating Outside the Law

The legal frame for betting in New Zealand is built on three statutes and a regulator. The Gambling Act 2003 sets the general framework and is administered by the Department of Internal Affairs. The Racing Industry Act 2020 establishes TAB NZ as the racing and sports betting operator and, as amended in 2025, inserts section 74AAA: no person other than TAB NZ may offer racing betting, sports betting or other racing or sports betting to a person in New Zealand, regardless of where that person is resident or incorporated. The change took effect on 28 June 2025 and was confirmed by industry trade press. The Online Casino Gambling Act 2026 regulates online casino gambling separately and does not extend the betting monopoly; the right TAB NZ holds to apply for an online casino licence does not change who may take bets.

The plain reading is that New Zealand has chosen a single-operator model for online sports and racing betting. TAB and Betcha are the two online brands of that operator. There is no licence register for bookmakers because there is only one lawful provider. The Department of Internal Affairs told providers serving New Zealand to withdraw with immediate effect after the 2025 amendment came into force.

The greyhound picture is part of the same arc. Greyhound racing in New Zealand is closing in stages from 1 August 2026, and from that date offshore operators may not take bets on greyhound races either. The same single-operator logic that governs thoroughbred and harness racing extends to greyhounds, with a closing date the bettor is advised to note before staking on the code.

The cost of operating outside the law falls in the first instance on the operator. The Racing Industry Act 2020 prohibits the offer of betting to people in New Zealand by anyone other than TAB NZ. The Department of Internal Affairs statement treats overseas operators serving New Zealand residents as breaking New Zealand law. The penalty for a breach of section 74AAA is not stated in the section and no prosecution of an offshore bookmaker was found in the public record. The point-of-consumption charge on offshore racing and sports bets ceased on 30 June 2026, removing one piece of operator-side tax friction but not the prohibition itself.

The cost to the player is different and is described carefully. Section 19(4) of the Gambling Act 2003, inserted in 2025, says an individual may not be convicted for placing a bet with an offshore betting operator in breach of section 74AAA. No source states a penalty for a player. The Department of Internal Affairs warns of the risks: overseas operators may be in jurisdictions with minimal consumer safeguards, may not return unspent credits and may not pay out large wins. None of that converts the offshore option into a safe one. The honest framing is what the regulator and the statute support: TAB and Betcha are the only lawful option, with New Zealand harm-minimisation tools, regulator oversight and a complaints framework; offshore brands are unlawful for the operator to serve New Zealand, with no New Zealand complaints route and the regulator’s stated risks.

The tax picture is not a substitute for the licence picture. Inland Revenue guidance treats recreational gambling winnings as generally not taxable, and prize money from TAB horse, dog and trotting races is not taxable; a person who bets as a business is a different case and the relevant Inland Revenue pages are to be consulted. The 12% online gambling duty on offshore operators (rising to 16% from 1 January 2027) applies to operators, not players, and a duty being payable does not make an offer lawful. The bettor must not read a tax payment as evidence that a bookmaker is legal in New Zealand.

The advertising frame reinforces the same perimeter. The Advertising Standards Authority’s Gambling Advertising Code applies to betting advertisements: ads must not target children (under 14) or young people (14 to under 18), judged by context, appeal and expected audience, must not encourage harm, and must be truthful with terms available. The Department of Internal Affairs reads the betting prohibition as covering promoting such bets for another operator. Affiliates that route New Zealand players to offshore brands operate on the wrong side of that reading, and the bettor following an affiliate link is following a marketing channel that does not exist in a compliant form.

The age for betting with TAB NZ is 18. Land-based casinos are 20 — the two are not the same. Betting under 18 is an offence under section 89 of the Racing Industry Act 2020. Account opening at TAB NZ requires a verified New Zealand address and ID. The lawful New Zealand experience is a higher-friction account opening in exchange for the harm-minimisation tools and the regulator oversight that follows. The offshore brands that offer “no-KYC” sign-ups are not offering New Zealand-style harm minimisation; they are offering the opposite, at a price paid later if a withdrawal is reviewed.

Operator Reviews: The Real Picture in 2026

TAB — the lawful baseline

TAB is the longer-established of the two TAB NZ online brands, with the racing-and-sports betting app first released on 8 September 2014 under the seller name New Zealand Racing Board. Its position in the New Zealand market is unique: it is one of two brands of the only lawful online sports and racing bookmaker for people in New Zealand since the 28 June 2025 amendment to the Racing Industry Act 2020 took effect. The app is published under “TAB – Racing & Sports Betting” and the operator’s own materials describe fixed-odds betting, racing pools (tote / parimutuel), Same Game Multis across a single event, and the betting types a New Zealand punter expects — Win, Place, Each Way, Quinella, Exacta, Trifecta, First4, Quaddie and Big6 among them.

The TAB app lists the safer-gambling tools a reader must look for first: deposit limits, time-outs and self-exclusion, with the self-exclusion applying across both TAB and Betcha. The free Gambling Helpline (0800 654 655) and the text line (8006) are signposted in the app, and Safer Gambling Aotearoa operates alongside the helpline. These are the harms protections the offshore brands reviewed below do not offer, and their absence is a real cost.

The TAB NZ partnership with Entain, signed in 2023 and running for 25 years, gives Entain a 50/50 share of gross profits and commits at least NZ$150 million a year to TAB NZ for the first five years, with a one-off NZ$100 million payment to the racing sector. The partnership is the operational reason TAB and Betcha trade under one bookmaking operation: the trading desk, the risk management and the product offering flow from a single group.

For a reader weighing TAB, the practical picture is this: a single-operator market with the harms framework a regulator enforces, betting types that cover the racing and sports a New Zealand punter wants, and no price competition against another lawful New Zealand bookmaker. That last point is what makes the offshore brands attractive on price and what makes them unlawful for the operator to offer. The cost to the reader of choosing TAB is the absence of price competition; the benefit is the regulator oversight and the complaints framework.

Betcha — the same bookmaker, a different front

Betcha is the second of the two TAB NZ online brands. It carries the same statutory basis as TAB and runs under the same Entain partnership. The reader must treat Betcha and TAB as one bookmaker’s two front doors, because that is what they are.

Betcha’s own materials describe a daily Toolbox of odds boosts and insurance tools for bets that do not land, a Bet Social section for punters to chat, follow tipsters and share bets or pools, and a Playbook with tips and match analysis. These features are layered on top of the TAB NZ fixed-odds and tote offering. A reader comparing the two brands for product is comparing two presentations of one product range, not two bookmakers.

The arbitrage cost that follows from this structure is mechanical. An arb between TAB and Betcha is an arb against one trading desk. Any internal offsetting TAB NZ applies to a price seen moving on the other brand is theirs to decide. A reader looking for two-lawful-bookmaker arbitrage in New Zealand will not find it on the lawful roster.

For a reader choosing between TAB and Betcha, the comparison is on presentation and on which set of features (Toolbox, Bet Social, Playbook) suits them, not on price or on safety, which are common to both.

Glorion — live chat, no common offshore path

Glorion is the first of the offshore brands the page reviews. Support is available by Live Chat or email, and the site accepts customers over 18 years of age only. The brand names no licence that applies in New Zealand; the Department of Internal Affairs treats it as unlawful to serve New Zealand residents under section 74AAA of the Racing Industry Act 2020.

The cost for a New Zealand reader is the offshore group standard: an absence of local complaints channels or harm-prevention protections, a heightened risk of account restrictions on activity that fits the arb profile, and the regulator’s stated risk that the operator may not pay out large wins or return unspent credits. Live chat support is better than no support and is not a substitute for a regulator’s complaints process.

For a New Zealand reader choosing Glorion, the comparison must be honest: a higher-friction payout review at the very moment an arb needs the money to move, against a price advantage the lawful market does not offer. The price advantage is what the marketing sells and what the regulator warns against.

BillyBets — decimal formats and a coin shop

BillyBets offers three odds formats — Decimal (2.00), American (+100) and Fractional (1/1) — and a sportsbook layered with cashback and reload bonuses, plus a coin shop where coins collected through depositing, betting, tournaments and challenges can be exchanged for free bets and free spins. A five-level VIP programme sits on top. The brand names no licence that applies in New Zealand.

The offshore reality is consistent: the absence of local complaints channels or harm-prevention protections, a heightened risk of account restrictions on activity that fits the arb profile, and the regulator’s stated risk that the operator may not pay out large wins or return unspent credits.

For a New Zealand reader choosing BillyBets, the comparison is the same as for any other offshore brand in the table: a price that looks competitive against TAB NZ on the surface, and a withdrawal experience reviewed at the point the bettor needs it.

Roobet — crypto rails, Curaçao paper

Roobet is operated by Raw Entertainment B.V., registration number 157205, Korporaalweg 10, Curaçao, licensed under Curaçao licence OGL/2024/687/0427. The brand positions itself around crypto deposits and withdrawals. The Curaçao licence authorises no betting operations for persons in New Zealand, and the Department of Internal Affairs treats the brand as unlawful to serve New Zealand residents under section 74AAA.

The cost picture for a New Zealand reader runs through the crypto rails the brand pitches as a feature. Crypto deposits are typically one-way until the operator agrees to convert back to fiat for a withdrawal, and conversion spreads, network fees and minimum-amount thresholds take a cut. The “no fees” line in the marketing does not include the conversion cost or the network fee. A withdrawal in the same coin deposited avoids one conversion; a withdrawal in a different coin adds exchange-rate risk on top.

For a New Zealand reader choosing Roobet, the comparison is the crypto rail against the New Zealand fiat rail. The rail is faster on the way in and slower on the way out. The licence is Curaçao paper; the regulator is the Department of Internal Affairs in Wellington. The price is competitive against TAB NZ and the protection is not.

22Bet — wide market menu, Curaçao licence

22Bet takes single bets, accumulators, systems and chains, and offers live betting with real-time odds and score updates. The brand names a Curaçao licence (8048/JAZ) held by TechSolutions Group N.V., which gives no right to take bets from people in New Zealand. The market menu is wider than TAB NZ’s on many events, particularly on lower-tier football and on accumulator combinations.

This status is the offshore norm: no New Zealand complaints route, no regulator reach, account-restriction risk. Live betting with real-time updates is a feature any New Zealand reader can find at TAB NZ; the price advantage is on the accumulator combinations, where 22Bet’s wider menu can disagree with TAB NZ’s narrower one. The risk is the operator-side right to limit accounts that fit the arb profile, and the regulator-side risk that withdrawals may be reviewed at the moment an arb is settling.

For a New Zealand reader choosing 22Bet, the comparison is the wider accumulator menu against the lawful market’s regulator oversight. The wider menu is the feature that attracts arbers and the feature that gets accounts limited.

Rabona — Soft2Bet’s sportsbook, Curaçao licence

Rabona launched on 23 December 2019 as the sixth online sportsbook of the Soft2Bet group and operates under a Curaçao licence. The brand names no licence that applies in New Zealand. The position for a New Zealand reader is the offshore position: no local support channel, no harm-prevention structure, and account-restriction risk.

This network brings risks similar to other offshore operators: an absence of local complaints channels or harm-prevention protections, and an account-restriction risk. The New Zealand reader is not protected by that familiarity. The Curaçao licence is a regulator that operates at arm’s length from a New Zealand bettor.

For a New Zealand reader choosing Rabona, the comparison is on the wider market menu and the Soft2Bet group network against the offshore cost and the regulator position.

Golden Panda — dedicated sports terms, no NZ licence

Golden Panda publishes dedicated Sports Terms and Conditions covering fixed-odds betting, Combo Bets, the Bet Builder feature, live scoreboards and broadcasts, and stake and odds limits. The transparency on terms is higher than several of the offshore brands in the comparison; the licence position is the same.

This clarity is welcome, but it does not change the offshore reality: no New Zealand complaints route, no harm-minimisation framework. A reader who relies on the clarity of Golden Panda’s terms to argue a payout dispute will find themselves arguing with the operator, not with a New Zealand regulator.

For a New Zealand reader choosing Golden Panda, the comparison is on the terms’ clarity against the offshore cost. The clarity is genuine; the regulator position is unchanged.

BetNinja — Comoros licence, offshore cost

BetNinja states it is licensed to conduct online gaming operations by the Union of the Comoros. The Comoros licence gives no right to take bets from people in New Zealand, and the Department of Internal Affairs treats the brand as unlawful to serve New Zealand residents under section 74AAA. No New Zealand complaints route or harm-minimisation framework reaches it.

The regulator gap is the defining feature: no New Zealand complaints route, no harm-minimisation framework. The bettor is operating in a regulator gap.

For a New Zealand reader choosing BetNinja, the comparison is on the licence’s limited reach against the offshore price. The price is the marketing pitch; the regulator is not in New Zealand.

FunBet — live betting, no NZ licence

FunBet’s sportsbook offers odds on football, ice hockey, American football, basketball, volleyball, tennis, rugby and baseball, with live betting and a bookmaker described as open 24/7. The wider sport menu and the live betting are the marketing features; the offshore cost is the same as for the other brands in the comparison.

The cost picture for a New Zealand reader is the offshore picture. The 24/7 framing is a marketing claim, not a regulator position. A bettor who treats the live betting as the feature and the offshore status as a footnote is pricing the protection out of the bet.

For a New Zealand reader choosing FunBet, the comparison is on the live-betting menu and the sport coverage against the offshore cost. The menu is the pitch; the protection is not in New Zealand.

The Margin: What an Illustrative Two-Way Market Tells the Reader

The margin calculation below uses an illustrative two-way market at decimal odds of 1.90 and 1.90. This figure is not a quoted price; it illustrates the bookmaker’s margin arithmetic.

The implied probability of one decimal odds is one divided by the odds. At 1.90 the implied probability is 0.5263, and the two legs of the market sum to 1.0526. The payout share — the proportion of every NZ$1 staked that returns to bettors in the long run — is the reciprocal of the sum of the implied probabilities, which is one divided by 1.0526, or 0.9500. The margin — the proportion of every NZ$1 staked the bookmaker keeps on average — is one minus the payout share, which is 0.0500, or 5.00%.

Read that as a band. A market at 1.90 and 1.90 is a 5.00% margin market, and that figure is the bookmaker’s expected profit per NZ$1 staked across both outcomes in the long run. Any deviation from 1.90 on either side changes the figure: 2.00 and 2.00 is a 0% margin (and thus an arb of the kind the strategy sections above describe); 1.85 and 1.85 is a margin around 8.1%; 1.95 and 1.95 is a margin around 2.6%. The margin is the cost of betting at that bookmaker on that market in the long run. The arbitrage is the moment the price falls below the cost.

The condition the result carries: the figure assumes the market is balanced in stakes between the two outcomes. A bookmaker with an unbalanced book on this market is exposed on one side, and the margin on a stake-weighted basis can drift. The illustrative example takes equal stakes on the two outcomes, and the 5.00% margin is the figure under that assumption.

What this tells the reader is the ceiling on what a misquote is worth. If one bookmaker offers 2.05 on outcome A and another offers 2.05 on outcome B, the implied probabilities sum to 0.4878 plus 0.4878 = 0.9756, the payout share is one divided by 0.9756, or 102.50%, and the margin on this combination of prices is negative 2.50% — meaning a 2.50% arb exists across the two bookmakers. The bettor’s lock-in return is 2.50% of the bankroll before any deduction. That is the rare window the marketing sells and the Franck, Verbeek, Nuesch study finds in fewer than 1.2% of opportunities, and at a smaller rate. The cost in bankroll is what the calculation reveals: a 2.50% return on a NZ$10,000 bankroll is NZ$250 before voids, before account limits, before subscription costs, before the regulator’s warning on offshore brands that do not pay out.

Frequently Asked Questions

Does sports arbitrage betting work?

The arithmetic works: a stake on every outcome at prices whose implied probabilities sum below one locks in a positive return. The cost is everything around the arithmetic. Most opportunities return less than 1.2%, the bankroll to make the figure meaningful is large, bookmakers detect the pattern quickly and act on it, and in New Zealand the only lawful online sports and racing bookmakers are two brands of one operator, which collapses the cross-bookmaker structure the arb needs. The honest answer is “the maths checks, the system does not, on most bankrolls”.

How is an arbitrage bet placed in practice?

The bettor identifies a market where two or more bookmakers disagree on price enough that the implied probabilities of the mutually exclusive outcomes sum below one. The bettor sizes stakes on each outcome so the return is identical whichever outcome lands — stake A is the bankroll times the price of B divided by the sum of the prices, stake B is the bankroll times the price of A divided by the sum. Both legs are placed before either price moves. The execution cost is the speed the bettor pays for, the void risk on either leg, and the detection cost that follows if the pattern repeats.

Which bookmaker may lawfully take sports and racing bets in New Zealand?

TAB NZ, through its two online brands TAB and Betcha, is the only provider lawfully offering racing and sports betting to people in New Zealand since the amendment to the Racing Industry Act 2020 took effect on 28 June 2025. There is no licence register for bookmakers because there is only one lawful provider. Any other brand taking bets from a person in New Zealand is breaking New Zealand law, and no New Zealand complaints route or harm-minimisation framework reaches it.

How is the return on a bet worked out from the stake and the odds?

The return on a winning single bet is the stake multiplied by the decimal odds. A NZ$10 stake at decimal odds of 2.50 returns NZ$25 — NZ$10 of stake and NZ$15 of profit. A combination bet multiplies the odds of its legs into one price: four legs at 1.5, 1.7, 1.8 and 2.0 multiply to 9.18, and a NZ$10 stake returns NZ$91.80 on a win. The bookmaker’s margin sits inside every price; the figure that comes back is the headline, not the long-run expectation.

What is the bookmaker’s margin and how can it be seen in the odds?

The margin is the proportion of every NZ$1 staked across a market that the bookmaker keeps on average in the long run. It is calculated as one minus the reciprocal of the sum of the implied probabilities of all outcomes, where each implied probability is one divided by the decimal odds. A two-way market at 1.90 and 1.90 has implied probabilities of 0.5263 each, summing to 1.0526, a payout share of 0.9500, and a margin of 5.00%. A market where the sum of implied probabilities is below one is an arbitrage opportunity. The margin is the cost of betting at that bookmaker on that market, and it sits inside every price the bookmaker quotes.

Written by the editors at Nzesportsbets.