FINAL PROJECT

FINANCIAL DATA ANALYTICS — TERM PROJECT

Team: [Your Team Name]

Date: 2026

Research Question:

"For equity indices (NASDAQ, Dow Jones, Russell 2000) in the US

market during 2015-2024, does CPI inflation surprise affect daily

excess returns relative to S&P500, using CPI release date as

event timestamp?"

Workflow Structure:

1. Parameters — Flow variables controlling all key parameters

2. Data Acquisition — 6 CSV files imported via flow variables

3. Preprocessing — Returns and excess returns computed

4. CPI Processing — Inflation surprise variable constructed

5. Quality Checks — 6 quality checks implemented

6. Analysis — 3 OLS regressions, model evaluation, visualizations

Data Sources:

- Price data: investing.com (S&P500, NASDAQ, Dow Jones, IWM)

- CPI data: FRED (CPIAUCSL series)

BIAS & TIMING RISK DOCUMENTATION

1. LOOK-AHEAD BIAS: Avoided — CPI surprise computed

using only data available on release date. No future

data used in any calculation.

2. SURVIVORSHIP BIAS: Minimal — indices (NASDAQ,

Dow Jones, Russell 2000) are broad market indices

not individual stocks. Index composition changes

over time but this is standard practice.

3. SELECTION BIAS: Acknowledged — IWM used as

Russell 2000 proxy. ETF closely tracks index

with minimal tracking error.

4. CALENDAR/TIMING RULE: CPI release date used

as event timestamp. All 117 dates confirmed

as weekdays. Trading day rule applied via

Inner Join — only dates with price data retained.

5. REVISION BIAS: CPIAUCSL from FRED may be

revised after initial release. We use downloaded

data which may reflect revised figures.