How to Choose an Outsourced Dealing Desk Provider

Choose an outsourced dealing desk provider the way you would hire a head of dealing: on evidence of operating depth, not on the pitch. The short version: demand proof of real desk experience on your specific platforms, a written risk methodology, daily reporting you can sample before signing, read-only access as a hard requirement, and a contract with a clean exit. Everything else is detail, and this article walks through that detail.

Our questions to ask any provider checklist covers the basics. This article goes deeper: how to evaluate the answers, and what the weak ones sound like.

Dealing experience vs software resellers

The market contains two different animals sold under similar language. The first is a dealing operation: people who have run books, made routing decisions under pressure, and can describe a bad night in specifics. The second is a software reseller with a monitoring dashboard and a support team reading alerts. Both have their place, but they are not the same product, and the second cannot do the first's job.

The test is conversational. Ask who will actually watch your book at 3am, what their background is, and what they did the last time a client's bridge failed during a macro release. Operators answer with stories and specifics. Resellers answer with feature lists.

A second test is what the provider sells when you describe a problem. Describe a coverage gap and a reseller will propose a monitoring licence; an operator will propose a coverage model. Describe a toxic flow problem and a reseller will show you a dashboard filter; an operator will ask about your routing rules, your classification cadence and your bridge configuration, because that is where the problem actually lives.

Platform and bridge depth

Coverage claims are cheap; depth is verifiable. If you run MT5, ask which manager functions the desk uses daily and how it handles group-level exposure across multiple servers. If you run cTrader, Match-Trader, TradeLocker or DXtrade, ask the same questions for those back offices. A provider deep on one platform and shallow on yours will learn at your expense.

The bridge layer is where shallow providers are easiest to catch. Ask how they would detect configuration drift in oneZero, PrimeXM, FXCubic, Tools for Brokers or Centroid, whichever you run. A serious answer involves configuration exports, documented target states and execution baselines. A weak answer involves the word dashboard.

Coverage model and escalation

Ask what coverage actually means in hours. A team that is awake during European session and on call otherwise is not 24/5 coverage; it is daytime coverage with a phone number. You want a roster: named shifts, handover procedures, and a defined escalation matrix that says who gets woken, in what order, for which events.

Then ask the harder question: what happened the last time the escalation chain was used? Every real desk has been tested. A provider that cannot describe an escalation that went well, or one that went badly and changed the procedure, has not been operating long enough to trust.

Risk methodology

A provider should be able to show you its methodology as a document, not a vibe: how exposure limits are set and reviewed, how routing decisions are made and recorded, how toxic flow is scored, what triggers an escalation versus a routine adjustment. The document does not need to be long. It needs to exist, and your playbook needs to plug into it.

Be wary of providers whose methodology is discretionary brilliance. The entire point of an outsourced desk is consistent execution of agreed rules. A desk that improvises is an unaudited risk you cannot see.

Reporting and transparency

Ask for a sample daily report before you sign. It should show exposure snapshots, actions taken, exceptions escalated, and anything the desk noticed but did not act on. If the sample is a screenshot of a dashboard, that is what you will get every day: data without judgement.

Then test the report against reality during the assessment phase. Pick one incident from the sample period and ask the provider to walk you through it: what was seen, what was done, who was told, and what changed afterwards. The walkthrough tells you more about the operating culture than any document, because it reveals whether reporting is a genuine record of decisions or a compliance artifact produced for clients.

Transparency also means access to your own data. You should be able to see everything the desk sees, in near real time, without asking. Providers running their own risk platform, as we do with Verix, can give you a read-only view into the same monitoring the desk uses. Providers who cannot offer this are asking you to trust a weekly PDF with an operationally critical function.

Data access and partitioning

Two non-negotiables. First, access must be read-only. A monitoring and advisory desk does not need write access to your server, and any provider requesting it, or anything near client funds, should be a hard stop. Second, your data must be partitioned from other clients: separate credentials, separate storage, no shared views. Ask how partitioning is implemented technically, not just contractually.

Ask also about the provider's own access hygiene: who inside their team can see your data, whether access is named and logged, and what happens to your data when the engagement ends. A provider that cannot answer these questions quickly has not thought about them, and your client data deserves better than that.

Contract, pricing structure and exit

Pricing should be scope-based and legible: you pay for servers, symbols, groups, bridge connections and coverage hours, and you can see which line item moves the price. Illustratively, a well-scoped retainer commonly runs 20 to 40 percent of the all-in cost of an equivalent in-house desk; if a quote is far below that range, ask what is excluded.

The exit clause matters as much as the price. Look for a defined notice period, a written handover protocol covering credentials, reports and documentation, and no lock-in beyond what onboarding genuinely justifies. A provider confident in its service does not need to trap you.

One more contract point that is easy to miss: liability and error handling. Ask what happens when the desk makes a mistake, because every desk eventually does. A serious provider has a written incident process: immediate disclosure, a post-incident report, and a procedure change where the mistake revealed a gap. A provider whose contract is silent on errors, or whose answer is that errors do not happen, is telling you how the first real incident will go.

Red flags

• Requests for write access to your servers or any proximity to client funds
• No sample daily report available before signing
• Coverage described in adjectives rather than shifts and rosters
• No clarity on the seniority and experience of the people who will actually watch your book
• Methodology that lives in a pitch deck rather than a document
• Pricing that is suspiciously cheap with no scope definition behind it
• No written exit protocol, or visible discomfort when you ask for one
• Guaranteed P&L outcomes: a desk controls process, not market outcomes, and anyone promising the latter is telling you something about their honesty

Frequently asked questions

Deltar's outsourced dealing desk is built on the criteria in this article: read-only access, written playbooks, daily reporting and the Verix platform giving you the same view the desk has. Start with a Health Check to see how we work before any retainer is discussed.