Quantitative Division

Algorithms
Built to Beat
the Market

A team of computer science students and finance analysts collaborating to design, backtest, and deploy systematic trading strategies.

In Coverage
Coverage Universe

The S&P 500,
Read From the Filings

Every fundamental in our pipeline is pulled directly from SEC EDGAR, as originally filed, with no third-party data vendor in between. This is what we actually hold: every company, every field, including the gaps.

Companies
—
S&P 500 constituents
Sectors
—
—
Field Coverage
—
—
Screen-Ready
—
Core financials complete
Coverage by Sector
Share of companies with usable data for each line item. Click a sector to filter the list below.
Not applicable to sector
Company Explorer
5+ yearsUnder 5 years MissingN/A
CoveredAt least five annual 10-K periods for the line item, deep enough to measure trends and growth.
ThinPresent, but under five years of history. Usable for levels, weaker for growth.
MissingNo usable figure found across the tags we map. Flagged rather than filled with guesses.
Not applicableLine items a sector doesn't report by design, such as COGS or current assets for banks and REITs.
Screen-ready means revenue, net income, total assets, equity and cash are all present.
Build Log

What's Built.
What's Next.

Our quantitative platform is built in layers, each depending only on the ones below it. We publish where it stands, including what isn't finished.

LAYER 01Complete
Portfolio Analytics
Client holdings, returns, risk metrics and reporting, on our own market-data store.
  • Market data ingestion and caching in an analytical database
  • Simple, log, cumulative and annualized returns; volatility
  • Sharpe ratio and maximum drawdown
  • Live position tracking and per-client reports
  • Automated test suite
LAYER 02In Progress
Investment Screening
Multi-criteria scoring that combines quantitative fundamentals with our analysts' qualitative theses. The first strategy is value investing.
Done
  • SEC EDGAR filings pulled for the full S&P 500
  • Accounting-tag map that reconciles how companies label the same line item
  • Data-quality audit across 27 line items (the explorer above)
  • Company history that keeps past classifications, so old periods aren't rescored with today's labels
  • Storage for analysts' four-pillar qualitative theses
Now
  • Persisting raw filings and computing derived metrics: margins, gross profitability, Altman components
  • Filing-lag rule, so no figure is used before the date it became public
Next
  • Altman Z-score screen to remove financially distressed companies
  • Sector-neutral percentile ranking
  • Weighted qualitative pillar scores
  • Monte Carlo sensitivity across weights and classifications
LAYER 03Planned
Portfolio Optimization
Constrained allocation across screened names, with an explicit risk model.
LAYER 04Planned
Trading Algorithm
Systematic execution, paper-traded before any capital is committed.
The Division

Human Insight.
Machine Precision.

DC Quant is our algorithmic trading division, a collaborative project between our investment team and a cohort of CS students from top Canadian universities.

We build algorithms that exploit market inefficiencies across timeframes: from intraday momentum to multi-week mean reversion. Every strategy is fully backtested before any real capital is deployed.

Our edge is the combination: domain knowledge from our fundamental analysts informing the signals our engineers code. Neither pure quant nor pure discretionary — but the best of both.

Technology Stack
PythonCore LanguageStrategy dev
pandasData AnalysisSignal processing
NumPyComputationMatrix ops
------
backtraderBacktestingStrategy testing
AlpacaComing soon--
PostgreSQLData StorageHistorical data