Data Methodology

How we collect, process, and deliver financial market data

Data Pipeline Architecture

1

Data Collection

Real-time feeds from exchanges & APIs

2

Processing

Validation, normalization & enrichment

3

Delivery

Low-latency distribution via CDN

Collection Sources

  • • Direct exchange connections
  • • Licensed data vendors
  • • Public APIs & feeds
  • • Web scraping (where permitted)

Quality Checks

  • • Outlier detection
  • • Cross-source verification
  • • Corporate action adjustments
  • • Automated anomaly alerts

Performance

  • • < 100ms latency (US equities)
  • • 99.9% uptime
  • • Global CDN distribution
  • • Auto-failover systems

Corporate Actions Adjustments

Stock Splits

All historical prices adjusted proportionally

Example: 2-for-1 split
Before: $100 → After: $50
Volume: 1,000 → 2,000

Dividends

Adjusted close prices account for dividend payments

Adjusted Close = Close × (1 - Dividend/Close)
Maintains accurate return calculations

Mergers & Acquisitions

Historical data preserved with appropriate conversion ratios

Price Types Explained

Open

First traded price of the session

High / Low

Highest and lowest prices during the session

Close

Last traded price at session end (unadjusted)

Adjusted Close

Close price adjusted for all corporate actions. Use this for return calculations!

Volume

Total shares traded. Adjusted for splits.

Calculation Methodologies

Daily Return Calculation

Simple Return = (Pt - Pt-1) / Pt-1 × 100%
Log Return = ln(Pt / Pt-1) × 100%

Note: Always use Adjusted Close prices for accurate returns

Volatility (Standard Deviation)

σ = √(Σ(Ri - μ)² / (n-1))
Annualized: σannual = σdaily × √252

252 = typical trading days per year

Moving Averages

SMA: (P1 + P2 + ... + Pn) / n
EMA: EMAt = Pt × k + EMAt-1 × (1-k)
where k = 2/(n+1)

RSI (Relative Strength Index)

RSI = 100 - (100 / (1 + RS))
RS = Average Gain / Average Loss (14 periods)

RSI > 70: Overbought | RSI < 30: Oversold

Index Calculation Methods

Price-Weighted

DJIA
Index = Σ(Stock Prices) / Divisor

Higher-priced stocks have more influence

Example: Dow Jones Industrial Average

Market Cap-Weighted

S&P 500
Index = Σ(Price × Shares) / Divisor

Larger companies have more weight

Example: S&P 500, FTSE 100, HSI

Equal-Weighted

Custom
Each stock = 1/n weight

All stocks equally important

Rebalanced periodically

Data Quality Standards

99.9%
Data Accuracy
< 1s
Update Latency (US)
24/7
Monitoring
20+
Years History

Quality Assurance Process

Automated validation against multiple sources
Real-time anomaly detection algorithms
Manual review of flagged data points
Audit trails for all data modifications

Important Disclaimers

  • • Data is provided for informational purposes only and should not be construed as investment advice
  • • While we strive for accuracy, data may contain errors or delays. Always verify critical information
  • • Past performance does not guarantee future results
  • • Real-time data may be subject to exchange licensing fees and usage restrictions
  • • Some markets have mandatory delays (typically 15 minutes) for non-professional users