Routing Beyond the Inside Quote: How Drovix Sends Institutional Flow Outside

A common misreading of an institutional execution platform is that it must be a venue. Our position is the opposite. Drovix is designed to be the system of record for an order, not the destination of every order it sees. When our internal book offers the best price in the size, the order trades there. When it does not, the order is routed externally — to the tier-1 venue or aggregated LP pool where the better price actually lives.

What follows is an honest description of what 'we route' actually means — because, as anyone who has worked behind a smart order router (SOR) knows, the routing layer is the place where unglamorous engineering quietly determines whether a client experiences best execution or theatre. A bad SOR can take a venue's tight headline spread and turn it into a 2-pip realised cost; a good SOR can take a 0.3-pip headline and improve on it.

The router has three jobs

A useful smart order router decides three things in sequence, every single time. Most commodity SORs answer the first; serious SORs answer all three.

1. Where is the best price right now, at this size?

Not the best top-of-book — the best executable price in the actual quantity being sent. A 0.0-pip top quote for one lot is irrelevant if you are sending fifty. The router must consult depth, not just the inside, and it must understand that depth at venue A is fundamentally different from depth at venue B because the venues fill differently under stress.

2. Which venues are healthy?

Quote freshness, last-fill latency, reject rate, and observed adverse selection per venue are continuously scored. A venue with a tighter headline but a 12 % reject rate in the last sixty seconds is not the best venue, regardless of the screen. Worse, it is a venue that will burn through your child orders and announce your intent to the market through the failed fills it generates.

3. What is the maximum amount of information we are willing to leak?

Routing a single large order to one lit venue tells the entire market what you are doing. Splitting it stupidly across venues tells them the same story in stereo. The right answer involves randomisation, ordering, and a per-counterparty leakage budget that is itself a policy parameter — chosen explicitly, not inherited from a default config.

Venues Drovix routes to

Drovix maintains direct connectivity to a curated set of tier-1 institutional FX liquidity providers (bank and non-bank) and to selected lit venues across FX, metals, indices, and crypto-asset derivatives. The list is intentionally not enormous: a router that can speak to a hundred destinations is, in practice, a router that has not measured any of them. Our criterion for adding a venue is that it can be measured, monitored, and disconnected within minutes if its behaviour degrades.

Venue selection is reviewed quarterly against measurable criteria: realised fill rate, post-fill price drift, reject taxonomy distribution, operational uptime, and counterparty-bilateral credit cost. A venue that drops below threshold on any one of those is put on watch; sustained degradation triggers de-listing. We publish the criteria internally as a written policy so the engineering team is not deciding venue health on Slack.

How Drovix decides internal vs. external

Every incoming institutional order is priced against three surfaces in parallel:

• The Drovix internal book at the requested size, including any inventory skew that improves the price for this counterparty.
• The aggregated external book at the requested size, normalised for venue-specific quirks (lot conventions, fee structures, expected slippage from observed reject rates).
• A counterparty-specific overlay that captures known operational preferences (minimum acceptable fill rate, preferred settlement currency, jurisdiction constraints, opt-out venues).

The order goes to whichever surface wins on a net-of-cost basis. If the internal book wins, you trade with Drovix as principal. If an external venue wins, the order is sent there with our credit line in front and our latency budget protecting the path. The fee structure is fully disclosed in the Institutional Services Agreement — we do not earn an undisclosed spread on externally routed flow, and the routing decision itself is journaled and replayable on TCA review.

What Drovix does not do

A list, because lists are clearer than disclaimers.

• We do not show your order to any counterparty that has not been pre-approved on your credit grid.
• We do not 'last-look' rejected orders back into our own book at a worse price.
• We do not internalise externally-routed flow on the way back when a fill report arrives — externally-routed means externally-routed.
• We do not run an iceberg-detection model against the very institutional clients we are routing for.
• We do not sell, lease, or otherwise share your routing intent with third-party analytics platforms.

Algorithm choice and counterparty consent

Routing algorithm is not an afterthought. Drovix exposes a small, deliberately curated set of execution algorithms — TWAP, VWAP-adapted, Implementation Shortfall, and a participation-based algorithm tuned for retail-broker-hedging cadence. Each algorithm has documented parameters, a stated objective, and a TCA benchmark. We do not invent new algorithms on the fly to suit a single counterparty; we extend the catalogue with a written specification and a backtest.

Counterparty consent on routing is explicit. When you join Drovix, your ISA documents which venues you will route to, which you opt out of (typically for credit or compliance reasons), and which algorithms you authorise. Changes require dual sign-off and are journaled.

Why this routing model is good for retail-broker hedging

If you are a regulated retail broker reading the B-Book to Wholesale playbook, routing matters because the externalised tail of your book is precisely the flow that is most price-sensitive. The Drovix router gives you, on every fill, the better of our internal price and an externally routed price — which means the hedging cost of your most painful exposures is structurally bounded by what an aggregated wholesale market would charge you, not by what a single counterparty would prefer to quote.

Why it matters for hedge funds and proprietary trading firms

If you are taking liquidity at size, the cost is no longer the headline spread — it is impact plus information leakage. A smart router that consults Drovix's internal book and an aggregated external view, with venue-health scoring and counterparty firewalling, is the difference between executing a thesis and announcing it.

The portal exposes routing decisions per fill: which surface won, by how many basis points, what the realised slippage was against the displayed price, and which venues were scored but not used. You can argue with the router on Friday afternoon, and it has a journal of evidence to argue back with.

A platform that always trades against you on its own venue has every incentive to be wrong about price. A platform that knows when to step aside has the only incentive that matters: getting you filled at the right number.

Where to read next

→ The Architecture of a Fair Spread — how the prices the router compares against are themselves constructed.

→ Microseconds Matter — the latency budget that keeps the router faster than the news.

→ Risk Without Friction — pre-trade gates and the immutable journal that make routing decisions auditable.