Kamba — Research & Perspectives
Kamba · Research & Perspectives

Research and perspectives on financial data workflows.

How institutional teams move from raw data to governed analysis — and where workflows still break.

We write when we've seen something worth saying. Our focus is the gap between data access and data use — the workflow friction that sits between a financial professional and a finished, governed output. When we publish, it's grounded in our work with hedge funds, asset managers, and the data teams that serve them.
Filter:
1 article  ·  More in progress
Workflow gaps Institutional AI Research note February 2026

The Workflow Gap: Why More Data and More AI Still Isn't Working

Every month a dataset sits outside production is a month your portfolio couldn't use it.

The bottleneck is almost never access to data or access to AI. It is the workflow between having it and using it — the gap between an AI pilot and a governed production system, between the budget line and the portfolio impact. This paper names it, documents it, and puts a number on what leaving it open costs.

Read the paper  → 7 sections  ·  16 cited sources  ·  ~12 min read
The evidence
The execution gap
66% priority / 16% executing
66% of asset managers have made GenAI a strategic priority. Only 16% have a fully defined strategy and are implementing it.
BCG Global Asset Management, May 2024
The daily frustration
79% cite integration as #1
79% of practitioners say combining data is their most frustrating daily challenge — more than cost or coverage gaps.
Exabel / BattleFin, Jan 2025
The spend paradox
Margins down 5 years running
Asset manager margins have declined for five consecutive years despite rising technology spend.
McKinsey Global Asset Management, Jul 2025
What the paper covers  ·  7 sections  ·  16 cited sources
01  ·  The cost model
Alpha opportunity lost, by firm size
Per dataset, per firm, across four AUM tiers ($750m–$30bn+). One auditable formula.
02  ·  Two types of cost
P&L impact vs operational waste
Alpha not earned and vendor overspend — defined separately and modelled independently.
03  ·  The hidden waste
Redundant, decayed, and zombie data
How to find it, quantify it, and remove it safely without breaking downstream pipelines.
04  ·  The market data
Where the buy side stands in 2026
16 cited sources — Coalition Greenwich, BCG, McKinsey, SEC, IOSCO — on adoption, spend, and AI execution rates.
05  ·  AI & agentic workflows
The strategy-to-execution gap
Why 66% have a GenAI strategy and 16% are executing — and what the firms executing have in common.
06  ·  The 2026 benchmark
What good looks like now
A governed end-to-end workflow — from discovery to production — where compliance is built in from the start.
~$7.3m
Estimated gross alpha opportunity lost annually for a $2.5bn fund onboarding six datasets a year at status-quo speed. The paper models this across four AUM tiers with an auditable formula your quant team can stress-test in ten minutes. Illustrative model  ·  Gross, pre-costs  ·  Mid-size tier, 6 datasets/yr, 60 bps good-tier assumption  ·  Results vary materially by strategy and execution
Written from the vantage point of Kamba's work with data-intensive hedge funds, multi-PM platforms, and asset managers.
Read the full paper  ↗
Data validation  ·  Coming soon
The DQR Problem: Why Manual Data Quality Reviews Don't Scale
How the weekly DQR became the slowest step in every data onboarding cycle, and what it costs when it fails.
In progress
Institutional AI  ·  Coming soon
Why Hedge Funds Aren't Getting ROI from AI — Yet
The difference between AI as a capability and AI as a production workflow — and which one moves the Sharpe ratio.
In progress
Research operations  ·  Coming soon
The 70/30 Problem: How Analyst Time Gets Spent vs How It Should
Seventy percent on data preparation. Thirty percent on thinking. The research operations problem quantified.
In progress
From research to practice

The firms solving the workflow gap
are starting with Kamba Analyst.

The patterns documented in this research — delayed datasets, manual DQRs, governance gaps — are the exact workflows Kamba was built to compress. See what your team produces when the workflow is handled.

Request a demo  → See Kamba Analyst