Institutional research, market analysis, and insights for professional market participants. View our latest commentary and announcements here.
An institutional client portal built for hedge funds, proprietary trading firms, and professional market participants — backed by a C++ engine, ONNX-served pricing models, and Aeron transport.
By Drovix Research Desk
Read more
Most Transaction Cost Analysis reports collect dust. Here is how to rebuild TCA so it changes broker selection, routing logic, and execution governance — not just fills a quarterly slide deck.
Read moreInstitutional-grade analysis workflow covering macro events, cross-asset correlations, and risk sentiment across global sessions.
Analyst Spotlight
12+
Daily Market Brief
24/5
Weekly Newsletter
Weekly
Day-by-day project plan for onboarding a wholesale liquidity provider in six weeks — KYC, contracting, FIX certification, parallel running, throttled cutover, and the 90-day review.
Spread is the headline. The real cost of wholesale liquidity is a stack of seven components — and only one of them shows up on the quote screen. A 2026 breakdown for liquidity buyers.
Read more
A decision framework for brokers and funds deciding between a prime-of-prime aggregator like Drovix and direct tier-1 bank lines — balance sheet, operational load, capacity, and the real cost difference.
Read more
A practical playbook for migrating wholesale FX/CFD liquidity providers — parallel running, dark-launch percentages, fill-quality cohorts, and the eight failure modes nobody warns you about.
Read more
An honest, vendor-neutral guide for FX/CFD broker COOs evaluating Drovix as a wholesale liquidity provider — pricing, coverage, credit, and what to ask in the RFP.
Read more
The headline spread is the cheapest, most-shopped, most-honest number on the desk — and the least predictive of what a real fill will cost. A six-component decomposition.
Read more
Last look is defensible when symmetric. When asymmetric, it is a transfer of expected value from the taker — statistically detectable in three distinct signatures.
Read more
Every order carries information for as long as the market hasn't priced it in. The half-life is empirical, measurable, and governs how an LP's memory of a desk's flow decays.
Read more
A margin model is a forecast. Whether it scales up faster than the underlying risk in a stress window is the question that distinguishes a strategic counterparty from a tactical one.
Read more
The premium for being at the front of the queue is small, real, and realised on a microsecond budget. The router that captures or surrenders it is the one that produces TCA-visible cost differences.
Read more
Every LP quote contains a baseline plus a conditional adjustment plus a counterparty-specific overlay. The desk's only leverage is on the third layer — but it is the layer that compounds.
Read more
Marketing claims about 99.9% uptime describe the steady state. The interesting engineering decision is what happens to in-flight FIX traffic in the microseconds while a primary fails over.
Read more
More LPs is not always better. The capacity curve plateaus around 7-10 LPs on a major pair and turns negative beyond 12 — for measurable, mechanical reasons.
Read more
Diversifying across three execution counterparties is nominal diversification if all three run through the same prime broker, the same LP panel, and the same colocation building.
Read more
FIX 4.4 is flexible. Twelve specific tag-handling decisions at the boundary between the desk's OMS and the counterparty's session account for most of the unexplained variance in TCA.
Read more
Spread is the cheapest part. The expensive parts are timing, signalling, queuing, and adverse selection — and most retail-grade reporting hides three of them.
Read more
Last look isn't going away. The right question is no longer whether a venue uses it but how the hold window is sized, whether rejection is symmetric, and what the venue does with the information it just saw.
Read more