The platform
No code required.
The same pipeline, driven from a workspace instead of a terminal: build a universe, download data, construct a portfolio, run a backtest and attribution, each step tracked to completion.
Business line 01
Lab.
Modular tools to develop, test and validate your investment ideas. Trace your strategy from hypothesis to live alpha.
Six modules, three phases
Seek.Find.Answer.
Six tools, three phases. In Mayan, Kaxan means seek and find. Nuuk means answer. Three are ready; the other three are on their way.
Data Curator
Acquire. Validate. Unify.
Data Refinery
Data Analyzer
Portfolio Construction
Backtest Engine
Simulate. Measure. Validate.
Attribution Analysis
Decompose. Explain. Improve.
The platform
Six modules.
One module per stage of the process. Each reads the previous module's output, so you can enter the pipeline wherever your work already is.
No module matches both filters yet.
Seek.
Open source
Data Curator
Acquire. Validate. Unify.
Retrieves, validates and homogenizes financial data from multiple providers into clean, point-in-time datasets ready for research.
- Providers
- Validate
- Homogenize
- Features
- Output
- Input
- LSEG, FMP, Yahoo! Finance
- Config
- Excel or Python, Docker image
- Output
- CSV, Parquet, pandas
Coming soon
Data Refinery
Turns curated data into research-grade features: enriched with sector history, macro and alternative data, built across assets and time.
- Curated inputs
- Enrich
- Cross-section
- Features
- Enrich with
- sector history, macro, alternative data
- Transforms
- rolling, relative, interaction
Find.
Coming soon
Data Analyzer
Understand your data before it reaches your strategies: exploratory analysis, outlier detection and pattern discovery across broad universes.
- Raw + features
- Explore
- Outliers
- Dashboards
- Levels
- feature, ticker, cross-section
- Output
- interactive dashboards, outlier reports
Coming soon
Portfolio Construction
Turns features, alpha signals and risk factors into investable portfolios, with a full history ready to run in the Backtest Engine.
- Universe
- Rank
- Size
- Rebalance
- Portfolio
- Allocation
- ranking, mean-variance, Black-Litterman, HRP
- Timing
- calendar, dynamic, hybrid
Answer.
Lab licence
Backtest Engine
Simulate. Measure. Validate.
Realistic, reproducible portfolio simulations, with commissions, slippage and 30+ performance and risk metrics against your benchmark.
- Data input
- Execution
- Valuation
- Performance
- Engine
- PyArrow, columnar
- Execution price
- VWAP, adjusted close
- Costs
- commission, slippage
Lab licence
Attribution Analysis
Decompose. Explain. Improve.
Explains where your returns came from: a portfolio's active return, decomposed with Brinson-Fachler and a factor model.
- Active return
- Allocation
- Selection
- Factors
- Residual
- Methods
- Brinson-Fachler, factor model
- Groupings
- sectors, industries, factors, custom
- Config
- Excel parameters workbook
The Investment Lab
Two ways in.
Write code if you want to. Click if you don't.
The library
Create ideas with code.
Two of the Lab's tools are open source and public: the Data Curator and the Researcher. An idea you can write as code is an idea you can run, share and test. No account, no licence.
pip install kaxanuk.data_curatorThe Backtest Engine and Attribution Analysis come with the Investment Lab licence.