INSTITUTIONAL FINANCE · UX RESEARCH ◆ INSTITUTIONAL CONCEPT · PRODUCTION-GROUNDED

TradeX
Institutional Trading Terminal

L2 Order Book, Portfolio Risk Matrix (80+ funds), real-time P&L attribution, keyboard-first execution — designed for 120+ concurrent metrics and sub-second decisions.

Grounded in informal PM conversations and direct production exposure from ACY Connect — FIX 4.4 institutional platform serving hedge funds and prime brokers in production. See also: TradeX Hedge Fund (PM strategy and cognitive workflow layer).

Design Provenance — Derived from FIX Protocol

The interaction rules in this terminal — partial fill state display, session-aware SOR routing, single-instrument depth focus, keyboard-first cancel/modify workflows — are not borrowed from Bloomberg screenshots. They are derived from ACY Connect's FIX 4.4 protocol work (production, 12+ institutional clients): Tag 39 OrdStatus lifecycle, session state machines, and anti-pattern constraints that govern real institutional order flow. Concept work grounded in protocol reality.

Live Interactive Prototype All data simulated · Updates every 350–800ms
AAPL 182.48 ▲0.23(+0.13%) MSFT 415.27 ▲1.18(+0.28%) NVDA 891.43 ▼2.10(-0.23%) SPY 524.18 ▲0.87(+0.17%)
NYSE OPEN VIX 18.43 09:31:04
L3 Order Book · AAPL
ORDSIZEBID
SPD
ASKSIZEORD
AAPL · 1m Candlestick
O: H: L: C:
Active Positions · Real-Time P&L Total: —
Microstructure Kyle's λ: 0.0032 Toxic Flow: 23.4% OFI(5s): +14.2% Spread: 1.8 bps T2Q: 0.0012 F9 Execute  ·  Ctrl+R Sort  ·  Alt+O Order Ladder
SOR Routing

TradeX Institutional Terminal · Interactive Prototype · Simulated Live Data · Exchange-Level Infrastructure Demo

Why TradeX Exists

My production work spans institutional B2B infrastructure (ACY Connect FIX API platform) and retail trading interfaces (LogixTrader, Finlogix, TradingCup). TradeX extends that experience into terminal UX for portfolio managers handling $100M–$50B AUM — applying what I know about institutional order flow to a different scale of problem.

ACY Connect is where I ship institutional infrastructure — FIX Protocol integration for hedge funds and prime brokers. The technical logic mapping from that production work — deconstructing FIX message states like 35=8 (Execution Report) and 35=W (Snapshot) with our Tech Lead and QA — gave me the institutional foundation to design the TradeX terminal.

Three institutional design challenges I explored:

  • Level 2 Order Book: Bid/ask depth, spread dynamics, liquidity heat maps. Inspired by order flow patterns from ACY Connect's FIX documentation work.
  • Portfolio Risk Attribution: 80+ funds with real-time VaR, Sharpe Ratio, Alpha, Beta across sector/regional exposure.
  • Multi-Asset Orchestration: Monitor equities, forex, commodities, derivatives simultaneously with cross-market correlation.
What I Actually Researched (Honest Documentation)

I said "portfolio manager interviews" in the description above. Here is what that actually meant — specificity matters for credibility:

Production Exposure

ACY Connect FIX 4.4 — shipped production platform for hedge funds and prime brokers. Direct exposure to institutional order flow, execution reports, and position management workflows.

Directional Interviews (n=3)

45–60 min conversations with one PM, one quant analyst, one buy-side risk officer. Not structured sessions. Findings informed the cognitive state framework and 4 unmet demand categories.

Bloomberg Analysis

Trial access + systematic review of Bloomberg's command structure and IA. Supplemented by FactSet public documentation and Coalition Greenwich research on terminal adoption.

This is not the depth of primary research I conduct for production ACY projects. TradeX is a rigorous design hypothesis informed by real exposure — not a validated product born from structured user research. The distinction matters.

1. Institutional Architecture: Risk & Compliance Matrices

TradeX's matrix architecture is derived from the same constraint logic as FIX protocol design: risk managers need fund-level volatility without navigating away; compliance officers need filing status inline with trading activity — not in a separate system. The layout reflects how information flows in production institutional workflows, not how Bloomberg chose to display it.

2. Design Decisions — Why This, Not That

Each decision below documents the alternative I rejected and the reasoning behind it — grounded in direct protocol experience from ACY Connect's FIX 4.4 production system, not from reading Bloomberg screenshots.

Decision 1 — Data Density

Single-page 960 data points, not progressive disclosure

Alternative rejected: Drill-down navigation (click fund → detail view). Standard in retail dashboards.

Why: From the PM conversations — portfolio managers monitor 80+ funds simultaneously during trading hours. Context switching kills reaction time. Every navigation click is latency in a position decision. Density isn't a style choice; it's a workflow constraint.

Decision 2 — Order Book Focus

Single-instrument L2 depth, not multi-instrument overview

Alternative rejected: Show bid/ask for multiple instruments simultaneously.

Why (from FIX production): FIX sessions are instrument-specific by protocol. A Tag 150 ExecType state for AAPL has no meaning in the context of MSFT depth. Mixing instruments in a single order book panel introduces cross-contamination risk in position sizing. Keyboard context-switching (Alt+O) solves the speed problem without compromising protocol clarity.

Decision 3 — Compliance Timing

Pre-execution constraint display, not post-submission errors

Alternative rejected: Surface SEC Rule 15c3-1 / FINRA margin violations after order submission (standard retail pattern).

Why: In retail, a rejected order is a minor friction. In institutional execution, a compliance rejection after submission triggers a reporting obligation and a trade reconstruction audit. The cost of the error changes the design requirement — surface the constraint before the trader commits, not after.

2. SEC/FINRA Compliance: Catching Violations BEFORE Legal Review

Why this matters for private bank platforms: UHNW $10M+ AUM clients operate under extreme SEC/FINRA scrutiny. One compliance violation = $1M-$50M+ fine + reputational damage. The workflow below shows how I proactively identify potential violations during design—not after implementation.

While TradeX is a concept project, this compliance review process is exactly how I work with Legal teams at ACY Securities. Internal ACY platform screenshots are available during interviews; the workflow shown here is identical to production.

Legal/Compliance Collaboration Workflow - TradeX case study — how Ed Chen identifies potential SEC/FINRA violations…

🔍 What This Workflow Shows:

Violation 1: Unrealistic Profit Representation

Daily Return figure of $1.9M is extreme for a retail dashboard without "Simulated" labels. Solution: Add mandatory SEC risk disclosures.

Violation 2: Ambiguous Risk Warnings

"Moderate" is vague for maintenance banners. Solution: Use standardized status terminology required for audit trails.

Proactive Compliance Design:
  • Identify violations before Legal review (saving weeks of rework)
  • Translate regulatory requirements into specific design constraints
  • Document compliance rationale for audit trails

Real-World Institutional Compliance: Fixed Income Screener & SEC Filing Review

The interface demonstrates a production-grade institutional compliance workflow. This shows how regulatory review is integrated directly into the trading terminal.

1. SEC Filing Queue: N-PORT, N-MFP, N-CEN filing status tracking with "STAT05" standardized compliance codes.
2. Bond Screener Matrix: Real-time bond screening with CUSIP identifiers, OAS spreads, and SEC-required risk indicators.
3. SEC Audit Trails: Live monitoring of filing health metrics (98.43% data integrity) and violation frequency trends.

3. Portfolio Risk Matrix Design

960 Data Points Without Cognitive Overload

80 funds × 12 metrics = 960 concurrent data points. How do you keep it scannable?

Layer 1: Critical Outliers

VaR and Sharpe Ratio outliers highlighted via font-weight and color-coded severity.

Layer 2: Metadata

Supporting fund IDs and regional tags dimmed at 60% opacity to reduce visual noise.

What This Demonstrates:

  • Extreme Data Density: Managing 900+ points without cognitive collapse.
  • Risk Severity Categorization: Using heat-mapping for VaR limits.
  • Institutional Scannability: Typographic hierarchy optimized for pro trading.

Design Decisions: What I Chose — and What I Explicitly Rejected

Each of the following was a real fork in the design. The rejected paths weren't wrong — they were wrong for this specific user context.

Layout Architecture
✕ Rejected
Tabbed Navigation

Tabs hide context. A portfolio manager switching between Risk Matrix and Order Book loses the cross-correlation signal — the P&L movement that caused them to check the book in the first place.

✓ Chosen
Persistent Panel Grid

All primary views visible simultaneously. Panels are resizable but never hidden behind navigation. Matches the multi-monitor mental model of experienced desk traders.

Order Book Rendering Technology
✕ Rejected
SVG-based Order Book

SVG DOM manipulation at L3 tick frequency (10–50ms updates for a liquid equity) causes visible jank above ~200 rows. DOM diffing overhead is incompatible with the latency expectation.

✓ Chosen
Canvas-rendered L3 Book

Canvas repaints the entire price ladder each tick at a fixed cost. Combined with virtual scrolling for the full order depth, this matches the performance profile of production institutional terminals. Accessibility tradeoff acknowledged — screen-reader fallback requires a separate DOM summary row.

Compliance Violation Surfacing
✕ Rejected
Post-Submission Error Modal

Standard retail pattern: submit order → receive rejection → re-enter. In a professional desk environment this costs seconds and disrupts flow. Legal at ACY confirmed the same problem for their manual compliance review workflow: violations found after the fact require full re-documentation.

✓ Chosen
Pre-Trade Inline Validation

Position limit warnings, wash-trade flags, and concentration limits surface in the order ticket itself as the trader types quantity. This mirrors how Bloomberg's TOMS and ION's trading platforms handle compliance — the constraint is visible before commitment, not after rejection.

Color System
✕ Rejected
Bloomberg Green-on-Black Clone

Pure imitation fails WCAG 2.1 AA contrast for financial data display and creates brand confusion. More critically, green-on-black conflates semantic green (profit / buy) with display green, causing parsing errors under stress.

✓ Chosen
Semantic Color System with Dual-Theme

Red/green reserved strictly for directional price semantics. UI chrome uses neutral blues and ambers. Light mode for compliance/risk roles (printable, auditable). Semantic consistency enforced across both themes so traders switching workspaces don't re-learn color meaning.

Multi-Dimensional Analytics Integration

The right-side panel integrates 5 analytical dimensions simultaneously:

  • Sector Allocation Bar Chart: Portfolio exposure by industry (2015–2021 time series)
  • Stress Test Scenarios: Portfolio performance under Stress/Crisis/Montreal conditions
  • Regional Exposure Heat Map: Geographic concentration risk (China, Spain, European Union, Frontier markets)
  • S&P 500 Correlation: Live chart with 3 moving averages (Morning/Median/Closing) tracking index correlation
  • EUR/USD Time Frame: Forex exposure impact with volume distribution

Design Rationale: Institutional portfolio managers don't look at funds in isolation. They evaluate systematic risk, correlation exposure, and macro sensitivities. This multi-panel layout supports holistic portfolio construction, not individual security selection.

4. Level 2 Order Book: Market Microstructure

The Order Book is the most critical institutional trading interface — where professional traders analyze market depth, liquidity concentration, and optimal execution pricing. The design logic for this component was directly informed by my production work on ACY Connect.

By mapping raw FIX 4.4 protocol logic from our Tech Lead and QA (handling Partial Fills, Iceberg Orders, execution reports, and snapshot updates), I gained the technical intuition needed to design a terminal UI that correctly represents market microstructure and liquidity heat maps at institutional scale.

TradeX Level 2 Order Book with bid/ask depth visualization, liquidity heat map, and one-click execution controls

What is an Order Book (and Why It Matters Institutionally)

In retail trading platforms (my LogixTrader work), users see a single "current price" (last traded price). Institutional traders need to see the entire order book — all pending buy orders (bids) and sell orders (asks) at different price levels. This is called Level 2 market data.

Why this matters: A portfolio manager executing a $10M equity order can't just "market buy" — they need to see if there's enough liquidity at the current price, or if their order will move the market (slippage). The Order Book visualizes this liquidity distribution.

Solution: Bid/Ask Depth Heat Map

TradeX's Order Book uses color-coded heat mapping to visualize liquidity concentration instantly:

Quick Execution Interface

The Order Book integrates one-click execution controls directly alongside market depth visualization — eliminating context switching between analysis and execution:

Execution Controls (Per Symbol Card)
  • Entry Price: Adjustable via +/− buttons (default: current market price)
  • Units: Position sizing with +/− controls (supports fractional shares)
  • Est. T.M. (Total Margin): Real-time margin calculation based on account equity
  • Auto Send Checkbox: Toggle between manual confirmation vs. instant execution
  • Buy Manually / Sell Manually: Large, color-coded action buttons (green buy, red sell)

Design Rationale: Professional traders executing 50-200 trades/day need zero-click workflows. The "Auto Send" mode enables keyboard-only execution: Tab to Units → Enter value → Spacebar to execute. No mouse required.

Trade History Integration

Below the Order Book, TradeX displays a real-time Trade History panel with granular status tracking:

Field Data Type Purpose
Status/Type Buy Limit Order classification (Pending, Filled, Closed, Rejected)
Units 100, 200, 5000 Position size per trade
Price 38.31, 120.77 Execution price (or limit price if pending)
Stop Loss 123 Risk management threshold (auto-close if price hits)
Take Profit 42.13 Profit target (auto-close when reached)

5. Multi-Chart Orchestration

TradeX Multi-Chart Dashboard with 4-panel synchronized charts for Microsoft, Apple, Google, Tesla including watchlist…

Challenge: 4 Simultaneous Charts Without Cognitive Overload

Portfolio managers monitoring multiple positions (Microsoft, Apple, Google, Tesla simultaneously) need synchronized time-frame controls — changing one chart's time range should update all 4 charts. This is a coordination challenge absent from single-chart retail platforms.

Solution: Unified Time-Frame Controller

Bottom-right chart controls: 1D, 1M, 3M, 1Y, 5Y, All buttons globally adjust all 4 charts simultaneously.

UX Rationale: Institutional traders analyze correlation patterns (e.g., "How did Microsoft and Apple move together during last month's Fed announcement?"). Synchronized time frames enable comparative analysis without manual re-adjustment per chart.

Terminal Orchestration: The Multi-Monitor Engine

Standard B2B SaaS design stops at the browser tab. Institutional terminal design starts at the workspace level. Top-tier traders at Goldman or BlackRock use 4–8 monitors; TradeX's architecture was designed to manage this Stateful Orchestration.

Cross-Window Synchronization

Using BroadcastChannel API and SharedWorkers, TradeX keeps multiple windows in sync — clicking a ticker on Monitor 1 updates the order book on Monitor 3 and news sentiment on Monitor 4. Browser-based sync typically achieves 100–200ms latency, which works for portfolio monitoring and position management. True sub-50ms sync (required for high-frequency execution) would need native terminal architecture like Bloomberg's C++ infrastructure.

Contextual Persistence

Workspaces are treated as Versioned Snapshots. If a trader moves from a "Macro View" to an "Earning Season View," all 8 monitors must re-provision their specific layouts and data streams simultaneously, preserving the "Ticker Context" across the entire physical desk.

Principal Design Perspective:

Designing for multi-monitor setups requires moving beyond 12-column grids into Spatial Orchestration. I de-prioritized "visual comfort" in favor of "Peripheral Awareness" — using color-coded luminance shifts on secondary monitors to signal market alerts while the primary monitor remains focused on execution. This is systemic design at the environment level.

Watchlist Integration

Left sidebar displays a live watchlist with 15+ symbols showing Bid, Ask, Spread, and 24hr Change%. This mirrors Bloomberg Terminal's "Monitor" panel — professional traders don't navigate to symbols via search; they maintain persistent watchlists of their coverage universe.

6. Adaptive UI

TradeX Light Mode interface with Fibonacci retracement levels, DeMarker indicator, split-screen chart view, and…

Why Light Mode Matters Institutionally

Most institutional terminals (Bloomberg, FactSet) default to dark mode for 10+ hour trading sessions (reduced eye strain). But compliance officers, risk managers, and regulatory reviewers often work in well-lit office environments where dark mode causes glare on printed materials.

Design Decision: Dual-Theme Architecture

TradeX supports both dark (traders) and light (compliance/risk) themes with preserved data visualization:

  • Dark Mode: Pure black (#000000) with neon accents (Cyan #00D1FF, Lime #4CAF50, Amber #FFAA00)
  • Light Mode: Off-white (#F5F5F5) with high-contrast typography (80%+ black text on light backgrounds)
  • Consistent Across Themes: Red/green price movements, chart candlestick colors, and risk severity indicators remain identical (ensures muscle memory transfer)

Technical Indicator Integration: Fibonacci Retracement

The Microsoft/TSLA charts display Fibonacci retracement levels (0.236, 0.382, 0.5, 0.618, 0.786) — a technical analysis tool used by institutional traders to identify potential support/resistance zones. Additional indicators visible:

  • DeMarker Indicator: Momentum oscillator (bottom panel) — identifies overbought/oversold conditions
  • Volume Bars: Integrated below price charts (green = buying volume, red = selling volume)
  • Multiple Time Frames: 1H (hourly) granularity for intraday trading decisions

Professional-Grade Technical Analysis: Fibonacci & Gann Tools

Institutional traders (particularly in commodities, forex, and equity derivatives) extensively use Fibonacci retracement/extension tools and Gann analysis for identifying key price levels and time-based cycles. Bloomberg Terminal, Thomson Reuters Eikon, and TradingView Pro all provide these tools as standard features for professional market analysis.

What This Demonstrates for Institutional Finance Roles

Including comprehensive technical analysis toolsets proves I understand:

  • Professional Trader Workflows: Institutional platforms must support quantitative analysis (algorithmic trading) AND discretionary technical analysis (human pattern recognition)
  • Multi-Strategy Support: Different trading desks use different methodologies — equities desks favor moving averages, commodity traders rely on Fibonacci, forex teams use Gann cycles
  • Tool Parity with Bloomberg: Any institutional terminal competing with Bloomberg Terminal MUST provide equivalent charting tools (Fibonacci, Gann, Elliott Wave, Ichimoku Cloud)
  • Configurability > Simplicity: Unlike retail platforms that hide complexity, institutional UIs expose ALL tools and let power users customize their workspace

Design Philosophy: Retail platforms optimize for "ease of use" (reducing cognitive load). Institutional platforms optimize for "depth of analysis" (maximizing analytical power). TradeX embodies this institutional mindset by providing professional-grade technical tools without UI hand-holding.

Collaboration Integration: Live Webinar Panel

Bottom-left corner shows an ACY Webinar video feed. This represents a unique institutional requirement: real-time analyst commentary. Bloomberg Terminal offers similar functionality through Bloomberg TV integration — traders watch market analysis while monitoring positions.

Design Insight: Institutional trading isn't solo work. Traders coordinate with research analysts, risk managers, and compliance officers via integrated communication tools (video, chat, voice). TradeX's webinar panel demonstrates understanding of this collaborative workflow.

7. What This Shows

TradeX demonstrates that I understand the architectural differences between retail and institutional finance UX — critical knowledge for designing platforms serving portfolio managers and wealth advisors:

Core Transferable Skills for Institutional Finance Roles

1. Extreme Data Density Design

Institutional platforms like Bloomberg Aladdin display 200+ risk metrics per portfolio. TradeX's 960-datapoint Risk Matrix demonstrates the information architecture approach required for scannable hierarchies at institutional scale.

2. Market Microstructure Understanding

Order Book design demonstrates knowledge of bid/ask spreads, market depth, and liquidity visualization — including the handling of Partial Fills and detection of Iceberg Orders (splitting large orders into smaller chunks to hide size from the market). These are critical for any institutional trading platform serving professional market participants.

3. Multi-Asset Class Orchestration

Professional traders monitor equities, forex, commodities, and derivatives simultaneously. TradeX's multi-chart synchronization mirrors multi-asset institutional terminal architecture.

CONCEPT EVOLUTION

From Institutional Terminal to AI-Native Hedge Fund Dashboard

The Portfolio Risk Matrix and Compliance architecture in this project form the structural foundation for a deeper exploration: what happens when you rebuild this terminal from scratch with intelligence-first architecture — no legacy constraints, no incremental feature additions? TradeX: Designing the Dashboard Bloomberg Can't Build answers that question across 8 views.

This Project Establishes
  • Data density architecture (960 data points)
  • Compliance matrix with SEC audit trails
  • Order Book microstructure visualization
  • Keyboard-first execution design
Hedge Fund Screen Adds
  • Live macro correlation panels (Sector / Regional / S&P)
  • Intraday alpha attribution decomposition
  • AI-driven hedge fund signal intelligence
  • Causal chain mapping (not just price correlation)
See the full intelligence-first evolution
Exchange-Level Layer · 2026

Exchange-Level Market Infrastructure

The gap between "institutional terminal" and "exchange-grade infrastructure" is defined by what lives below the chart. L3 order book depth, venue routing intelligence, microstructure analytics, real-time TCA, and T+0 settlement visibility are the layers that separate Bloomberg from consumer-grade terminals. The interactive prototypes below demonstrate each layer.

L3 Order Book — Full Depth, Individual Order Resolution

L2 aggregates by price level. L3 shows every individual order in the queue — the institutional standard for execution timing and informed flow detection.

AAPL · L3 Order Book · NASDAQ
09:31:04.847 LIVE
ORDERSSIZEBID
SPREAD
ASKSIZEORDERS

Market Microstructure Panel

Toxic flow detection, adverse selection metrics, and order flow quality — unavailable in any retail terminal. This panel answers: are informed traders positioned against me right now?

Kyle's Lambda
0.0032
↑ Info asymmetry elevated
Toxic Flow %
23.4%
⚠ Above threshold (20%)
Trade-to-Quote
0.0012
✓ Normal range
Realized Vol (5m)
18.2%
vs Implied Vol 19.7%
Bid-Ask Spread
1.8 bps
Today avg: 2.1 bps
OFI (5s window)
+14.2%
Bid-side imbalance

Smart Order Router — Venue Allocation

Real-time routing decisions across lit markets, dark pools, and off-exchange venues. Not a black box — institutional desks need to know if they're being front-run by their own router.

VENUEFILL%REBATELATENCYALLOCATION

Real-Time TCA — Transaction Cost Analysis

Every fill measured against arrival price — live, not end-of-day. Alpha attribution is meaningless if you can't distinguish genuine skill from favorable execution conditions.

TIMEARRIVALFILLSLIP(BPS)VENUESTATUS

T+0 Settlement Tracker (2026 Standard)

Post-SEC T+1 mandate with movement toward T+0. Risk management doesn't end at execution — it runs through CCP clearing and DTC confirmation.

Why This Layer Is the Real Institutional Qualifier

Retail terminals show price. Professional terminals show market depth. Exchange-level infrastructure shows why the market is moving — toxic flow, information asymmetry, venue liquidity fragmentation. Kyle's Lambda tells you if someone smarter is trading into your position. The SOR tells you whether your broker is optimizing your fill or their rebate. TCA tells you the real cost of your decision. In 2026, a terminal without these layers is a retail terminal wearing an institutional suit.

React Prototype · Figma Make

Explore the Full Terminal

12-screen interactive terminal built in React + shadcn/ui — Dashboard, Trading, Portfolio, Risk Matrix, Factor Exposure, Execution Analytics, Compliance, and more.

TradeX Institutional Terminal — React Prototype
React · TypeScript · shadcn/ui · recharts · Tailwind v4 · Figma Make · All data simulated
Dashboard Trading Portfolio Risk Matrix Factor Exposure Execution Analytics Compliance Markets Account KYC
Open Prototype

8. Reflection

"Retail platforms design for the 80% happy path. Institutional terminals design for the 1% catastrophic scenario."

The Portfolio Risk Matrix exists for the days when markets crash — when a portfolio manager managing $10B in AUM needs to identify which 5 out of 80 funds are violating VaR limits in under 5 seconds. That's not a retail UX problem. That's an institutional design requirement that demands extreme density, color-coded severity, and keyboard-first workflows.

This concept challenged most of the UX principles I use in LogixTrader and Finlogix. Those platforms benefit from progressive disclosure and cleaner layouts. TradeX needs the opposite: maximum information density and every data point visible at once. Knowing when to use which approach — and why — is the core skill I've been developing across both retail and institutional work.

Project Details

Design Focus Areas

#Order Book #Portfolio Risk #Market Depth #Multi-Chart #Data Density #Level 2 Data #Institutional UX

Related Production Work

Concept Extension

🧠

Why This Research Matters

My production portfolio (LogixTrader, Finlogix, TradingCup) proves I can ship retail/broker fintech products. TradeX demonstrates that I understand institutional-grade requirements — directly applicable to roles where design targets professional portfolio managers, not retail investors.