Skip to content
Genesis Ark

Buyer's guide

OMS vs EMS vs OEMS in crypto: what each one does and which one you need

The three acronyms describe different jobs: keeping the official record of orders and positions, getting orders filled well, and doing both in one system. Here is what each one does in a crypto setting, where the boundaries blur, and how to decide which one your firm needs.

Author
Genesis Ark Engineering
Published
Reading time
7 min read

In traditional markets, the difference between an order management system and an execution management system is well understood, because the two were usually bought from different vendors and run by different people. In crypto, the same acronyms get used loosely. Vendor pages call almost anything an OEMS, and buyers end up comparing products that do different jobs.

This article sets out what each system is responsible for, why crypto pushes the two together, and how to work out what your firm needs.

The order management system: the official record

An order management system (OMS) is the book of record for orders. Its job is to know, at every moment, which orders exist, who placed them, for which account, in what state they are, and what has been filled. Everything downstream depends on that record being right.

A crypto OMS typically handles:

  • The order lifecycle. An order moves through states such as pending new, new, partially filled, filled, pending cancel, canceled, replaced and rejected. The OMS owns that state machine and makes sure every transition is valid and recorded.
  • Accounts and permissions. Which client or fund an order belongs to, which trader may trade which instruments, and which approvals are needed for what.
  • Positions and balances. Each fill updates the position and balance of the account it belongs to. Risk checks, P&L and client statements all read from this ledger.
  • Allocations. A block order placed for several funds is split across them according to defined rules, and the split is recorded.
  • Compliance and pre-trade checks. Limits, restricted lists and credit checks applied before an order is released for execution.
  • The audit trail. A timestamped record of every event, so you can reconstruct what happened to any order.

The OMS is usually what the portfolio manager, the operations team and the compliance officer care about most. It answers the question: what is our position, and how did we get here?

The execution management system: getting the order filled well

An execution management system (EMS) is about the act of trading. Once an order exists and has been released, the EMS decides where, when and how to execute it.

In crypto, an EMS typically provides:

  • Venue connectivity. Adapters for each exchange, OTC desk and liquidity provider, translating a common order format into each venue's API and translating fills back.
  • Market data. Normalized prices, trades and order book depth across venues, so the trader and the router see one consolidated picture.
  • Smart order routing. Splitting an order across venues to get the best achievable price after fees, given the depth available at each.
  • Execution algorithms. TWAP, VWAP, percentage of volume, iceberg and similar strategies that work a large order over time to reduce market impact.
  • RFQ. Requesting quotes from several dealers at once for block trades.
  • Transaction cost analysis. Measuring how well each order was executed against benchmarks such as arrival price and interval VWAP.

The EMS is what the trader cares about most. It answers the question: did we get a good fill, and can we prove it?

Why crypto pushes the two together

In equities, a buy-side firm might run an OMS from one vendor and an EMS from another, connected by FIX. That works because the market structure is stable: a limited number of regulated venues, a central clearing system, and brokers that handle much of the execution.

Crypto is different in four ways that matter for system design.

1. You hold positions at the venue. In most crypto trading there is no central clearing house. Your balances sit on each exchange or with a custodian, and each venue is effectively a separate pot of money. The system that executes the order is also the system that sees the balance change at the venue. Splitting the record from the execution means two systems trying to agree on where your money is.

2. Venue connections are fragile. Exchange APIs change, rate limits differ, WebSocket connections drop and order acknowledgements arrive late or out of order. When a connection fails mid-order, someone has to reconcile what the venue thinks happened with what your records say. If the OMS and EMS are different systems, that reconciliation crosses a system boundary at exactly the moment you can least afford ambiguity.

3. Pre-trade checks need live positions. A credit check is only meaningful if it knows what the account already holds at each venue, including fills that arrived a second ago. That requires the risk check and the execution layer to share one position ledger.

4. Trading never stops. Crypto markets run every hour of every day. There is no overnight batch in which an OMS and an EMS can be reconciled at leisure.

The result is that most institutional crypto platforms combine order and execution management into a single system, an OEMS. One order model, one position ledger, one audit trail, with execution tools built on top.

What an OEMS should include

A useful way to judge an OEMS is to separate the parts that must be one product from the parts that can be added on.

The parts that must ship together:

  • Order management and the order state machine
  • The execution gateway and venue adapters
  • Accounts, users and permissions
  • The position and balance ledger
  • Pre-trade risk
  • The audit trail and event log

These parts read and write each other's state constantly. Buying them from different vendors means rebuilding the integration between them yourself, and that integration is where orders get lost.

The parts that sit on top:

  • Smart order routing
  • Execution algorithms
  • RFQ and OTC workflows
  • Client pricing and auto-hedging
  • Treasury, settlement and custody connections
  • TCA and reporting

These depend on the core but can be licensed and adopted separately, according to how your business trades.

Layer 4
Add-ons

needs Core

Layer 3
Core OEMS

ships as one product

Layer 1
Shared foundation
  • Venue adapters
  • Common data model
  • Authentication

under every module

Fig. 1. A core that ships as one product, with standalone modules and add-ons layered on top. Illustrative composition with sample data.

Where the labels mislead

Three patterns are worth watching for when you compare vendors.

An EMS described as an OEMS. Some products are excellent at routing and algorithms but keep only a light record of orders and no real position ledger. That is fine if you already have a strong OMS, and a problem if you expected the vendor to be your book of record.

An OMS with a thin execution layer. Some products manage orders and positions well but route to only a few venues, or offer no algorithms or routing logic beyond sending the whole order to one place.

A platform that is really a service. Some providers execute trades themselves, as a broker or a crypto-as-a-service provider, and give you an interface to their desk. That can be the right choice, but it is not the same as owning your trading system. Your orders are their orders, and your execution quality is their decision.

Ask each vendor three direct questions

Where is the official record of my positions held, what happens to an order if a venue connection drops after the order is sent, and who is the counterparty on my trades?

Which one do you need?

The answer depends less on your size than on what you already run and what you are responsible for.

You probably need an EMS if you already have an OMS that you trust as your book of record, and your problem is execution quality: thin liquidity on one venue, market impact on large orders, or no way to prove best execution. Many asset managers coming from traditional markets fit here.

You probably need an OMS if you trade through a single prime broker or dealer that handles execution, and your problem is the record: allocations across funds, compliance checks, position keeping and audit.

You probably need an OEMS if you trade on several venues directly, hold balances at those venues, and are responsible for your own positions and risk. Most crypto brokers, OTC desks, prop firms, exchanges hedging their own flow, and fintechs launching a brokerage fit here.

You may not need all of it at once. Many firms begin with the parts that read data rather than place orders: a portfolio and risk view across venues, reconciliation of fills against internal records, TCA on existing executions, or reporting. These modules use read-only access and carry no execution risk. They also build the venue connections and data model that the full OEMS will use later.

Questions to ask before you choose

  1. Which component is the book of record for orders and positions, and can it rebuild its state after a restart?
  2. How are duplicate and missing fills detected and handled?
  3. Is a retried order request guaranteed not to create a second order?
  4. Are pre-trade checks run against live positions at each venue, or against a periodic snapshot?
  5. Can you see the routing decision for any order, including the prices and fees the router saw?
  6. Who is the counterparty to your trades, and where are your assets held?
  7. Can you run the system on your own infrastructure if your regulator or policy requires it?
  8. Can you start with one part of the system and add the rest later without a migration?

The acronyms matter less than the answers. A system that keeps your record correct through every disconnect, and can show how every order was executed, is doing the job, whatever its vendor calls it.

See it on your own workflow.

A scoping call with an engineer, then a demo on sample data that mirrors your venues, desks and limits.