Your processor dropped you on a Tuesday, or it's holding 10% of every settlement and you only found out after the first batch cleared. Every ranked list you've opened since has the same 12 logos, the same line about competitive rates for high risk merchants, and no numbers anywhere on the page. This one ranks providers by what they put in writing before you sign, and works the arithmetic on the reserve, because the reserve is where your money goes.
High risk is a number on a bank's spreadsheet, not a label on your store
Nobody at an acquiring bank decides your brand feels risky. They price 3 things: your MCC, your dispute and fraud numbers against the card network thresholds, and how likely you are to stop shipping and leave them holding the refunds.
The thresholds moved this year and most pages ranking for this query quote the old ones. Visa's Acquirer Monitoring Program cut its merchant threshold from 220 basis points to 150 basis points on 1 April 2026, across the US, Canada, Latin America, Europe and Asia Pacific. CEMEA stayed at 220. The VAMP ratio counts TC40 fraud reports plus TC15 disputes over settled transactions, and a merchant comes into scope only after 1,500 combined fraud and dispute items in a month. Mastercard's Excessive Chargeback Program is separate: 100 to 299 chargebacks in a month at 1.5% or higher, and a high excessive tier at 300 or more with 3% or higher.
The number that actually decides your fate is none of those. It's the acquirer's own portfolio band. Visa grades acquirers on the blended ratio of every merchant they board, and there the bands are 50 to 70 basis points for above standard and above 70 for excessive. That is more than twice as tight as the merchant threshold, and it did not move when the merchant number did. Which is why a store sitting well under 150 basis points still gets a 30 day termination notice instead of a phone call. You were never near your own limit. You were pulling up theirs.
If that termination gets filed to MATCH, Mastercard's list of terminated merchants, it sits there for 5 years. Only the acquirer that filed it can remove it early, and only in 2 cases: it was filed in error, or it carries the PCI non compliance reason code and you have since become compliant. Processing cleanly elsewhere does not shorten it.
Most of the names on these lists are not processors
This is the part the comparison pages skip. Durango Merchant Services states on its own site that it is a registered ISO for Fifth Third Bank, N.A. Corepay's footer says it is a registered ISO/MSP of North American Banking Company. Merchant Maverick's review of eMerchantBroker says flatly that the company does not process in house, and names Fiserv, iPayment and North American Bancard among its back end processors. Easy Pay Direct's own FAQ describes working with acquiring banks to provide merchant accounts, naming Chase and Wells Fargo as examples. Host Merchant Services is a registered ISO sponsored by Wells Fargo.
None of that is a scandal. A good ISO earns its cut by knowing which bank has appetite for supplements this quarter. But it changes the questions worth asking. The person quoting you does not set your reserve, does not make the termination decision, and cannot reverse one once it lands. So the useful question is not what rate can you do. It's which bank is this MID on, and what is its appetite for my MCC.
Where the aggregators stop
Stripe, Shopify Payments and PayPal will board a supplement or skincare store in an afternoon, which is exactly the trap. Stripe's published restricted businesses list prohibits pseudo pharmaceuticals and nutraceutical products that are unsafe or that make claims capable of causing harm. Separately, under unfair and deceptive practices, it prohibits negative option marketing, negative option membership clubs, and discounted trial periods with unclear or hidden pricing. A supplement sold on a discounted trial that rebills sits in 2 of those buckets at once, on an account a web form approved in minutes.
Shopify is worth checking properly, because the answer is not where you would look for it. The Shopify Payments terms of service carry no prohibited business list at all. They say the prohibited and restricted categories are set by each payment processor in your country, and point at a separate processor list. The category language you need sits in the help centre eligibility page, which names pseudo pharmaceuticals and defines them as products making health claims not verified by a local or national regulatory body. That is wide enough to catch supplement and skincare copy that read as ordinary marketing to whoever wrote it.
Either way, nothing happens until volume or disputes trip a review, and then everything happens at once. If you're there today, the first 14 days of a Stripe freeze is the sequence to run.
The ranking
Ranked on 4 things: what they commit to in writing, whether the contract lets you leave, whether they board your vertical, and whether the reserve gets named before you process rather than after your first settlement. None of them publish a full high risk rate card, so what separates them is how much they state before the application.
- Durango Merchant Services. First, because it is most specific about the thing that costs you most. It publishes no rate card, and says so, but its own site defines the reserve structures it uses with numbers attached: a rolling reserve typically holding 5% to 10% of each deposit for 6 months, and a capped reserve holding 5% to 15% of each transaction until the balance reaches half to a full month of volume. It names Fifth Third Bank as its sponsor, lists supplements, telemedicine, subscription sites, dating, travel, pharmacies and vape among its verticals, and charges no setup fee. Merchant Maverick reports most merchants get month to month with no termination fee, that some higher risk merchants get a 3 year initial term with 1 year auto renewals and a fee of up to $500 to leave early, and that its BBB profile showed no complaints in 3 years.
- Easy Pay Direct. A gateway with account placement attached, and load balancing is the reason to look at it. You set a monthly dollar volume limit on each merchant account, and the gateway moves to the next account when a limit is hit or one goes on hold. Merchant Maverick lists card present at 1.59% plus $0.17 and ecommerce at 2.44% plus $0.34 with a $99 setup fee, and notes high risk runs above those. It reports contracts usually running 3 years with automatic renewal for successive 1 year periods, termination fees uncommon for standard merchants but common on offshore accounts, so the renewal clause is what to read. Boards firearms, nutraceuticals, CBD, hemp, cannabis and adult.
- Corepay. A registered ISO/MSP of North American Banking Company. It publishes reserve mechanics instead of rates, and documents the real band: 5% to 15% held for 90 to 180 days, longer where there is recurring billing, a complex fulfilment cycle or cross border volume. Named verticals include ecommerce, gaming, med spa, medical, pharmacy and telemedicine. Worth a call when the reserve, not the rate, is what you're negotiating.
- PaymentCloud. The widest published vertical list of this set: adult, alcohol, CBD, continuity and subscription, debt consolidation, dropshipping, firearms, nutraceuticals and supplements, online pharmacy, travel and timeshare. It states no setup fees and claims an approval rate above 98%, but publishes no rates, no reserve terms and no approval timeframe, so every number that matters arrives at quote stage. Hold it to the vertical list and get the reserve in writing.
- Soar Payments. No published pricing, everything quoted after an application. What it does give you is structural: Merchant Maverick reports its rolling reserves carry predetermined caps, so the hold stops growing once it reaches a set figure, and that held funds release on the pre established schedule if you end the contract. The percentage and hold period are not disclosed. High risk accounts carry a $495 termination fee. Industries include firearms, travel, CBD, vape, subscription, credit repair and debt collection.
- Host Merchant Services. The published terms are the best of this set, and the catch is that they may not be the terms you get. Standard accounts run month to month with a cancellation fee waiver attached to the agreement, on interchange plus pricing of roughly 0.20% to 0.25% plus $0.09 to $0.10 card present, an extra 0.35% on ecommerce, and monthly fees from $14.99. But high risk is placed through a partnership with Electronic Merchant Systems, and Merchant Maverick states plainly that those accounts carry different pricing and contract terms than the website's. Also expensive at low volume.
- eMerchantBroker. Last on terms, not on approvals. It boards verticals others decline, which is why it stays on every list. But it does not process in house, it uses tiered pricing, contracts run 1 to 3 years with automatic renewal, and the termination fee runs $295 to $595, lower on 1 year terms and higher on 3 year terms. Tiered pricing hides more than any other model, so ask in writing which of your transaction types land in the qualified, mid qualified and non qualified buckets.
The reserve is the price. The rate is the argument you're allowed to win.
Run the number. Take $300,000 a month in card volume on a 10% rolling reserve with a 180 day hold. Month 1, the acquirer keeps $30,000. So does month 2, and every month after. Nothing comes back until month 7, when month 1 releases. From there the balance sits flat: $180,000 of your money, parked, for as long as the account is open.
Now price the thing the salesperson wanted to talk about. Moving from 3.99% to 3.49% saves $1,500 a month at that volume. You would have to hold the better rate for 10 years to equal the cash already sitting in the reserve. So negotiate in this order: reserve percentage, hold period, whether it's capped, then rate.
Capped matters more than it sounds. A capped reserve stops accruing once the balance hits a set figure, so scaling doesn't increase what's parked. An uncapped one grows in lockstep with volume, so the better the quarter, the more working capital you've handed over to fund it.
Then ask the question nobody volunteers: when does the reserve release if the account closes? Get a date or a formula into the agreement. If nobody will name one, assume the full hold period from your last settled transaction, and price the account as a 6 month hostage going in and a 6 month hostage coming out.
2 MIDs behind the same bank is 1 MID
Splitting volume across accounts is the right instinct and it is usually implemented wrong. If a load balancer is routing between 2 MIDs that were both placed by the same ISO on the same acquiring bank, one underwriting decision closes both on the same morning. Real redundancy means different acquirers, ideally different jurisdictions, with descriptors and support numbers that match the store the customer bought from.
The rest of the job is not sending your ratio up in the first place. Dispute volume trails the ad spend that caused it, so the ratio you're graded on this month is measuring campaigns you ran weeks ago. Why your chargeback rate reports last month's ad spend covers the denominator lag, and explains why a scale up looks clean for 6 weeks and then doesn't.
The other account that goes down in the same week
Here's the trade off, stated plainly. Running on an agency held account means you don't hold the account. You can't sell it, you can't take it with you, and you're inside someone else's portfolio and someone else's rules about what runs on it. That is not free.
What it changes is the exposure. The billing side of an ad account and the merchant account fail from related events, and operators find that out in the same week. A dispute filed against a charge, a card that declines at the billing threshold, a payment method swapped at the wrong moment: the billing signals that fire an unusual activity disable are mostly payment events, not creative ones. When the processor freeze lands, the ad account is often next, and you rebuild both while the offer is still converting.
An account issued off a portfolio spending around $50 million a month carries payment history a fresh account does not have, and it gets replaced when it's restricted instead of appealed for weeks. It's a flat monthly rate with 0% of spend, so unlike a reserve, the cost doesn't climb when the month is good. What a cut of spend does to the bill as you scale runs that arithmetic the way this one ran the reserve, and the tiers and their spend ceilings sit side by side. If a provider has ever taken a top up and gone quiet, how handover actually works is the part to read first.
None of that removes the underwriting problem on the payments side. Get the reserve percentage, the hold period and the release date in writing. Get a second acquirer boarded before you need it. Keep the ratio down. That part stays your job.