Construct a custom index replicating the Nasdaq-100
Hands-on quantitative datasets preparation
Table of contents
- Disclaimer
- What is Nasdaq-100
- Current constituents weights
- Problem
- Goal
- Initial data requirements
- Selecting the data
- Constructing replicating index
- Comparing with original
- Final thoughts
- Links
1. Disclaimer
All data utilized in this analysis is for educational purposes exclusively.
All information and analysis provided are for educational purposes only and do not constitute financial advice.
2. What is Nasdaq-100
The Nasdaq-100 is a stock market index composed of the 100 largest non-financial companies listed on the Nasdaq exchange. It’s a capitalization-weighted index, meaning larger companies have a greater impact on its value. However, there are rules in place to limit the influence of the biggest companies.
Notably, the Nasdaq-100 exclusively includes non-financial firms, with financial companies represented separately in the Nasdaq Financial-100 index.
3. Current constituents weights
At the moment of writing this article 08/2024, According to this webpage, the largest 20 companies comprise most of the weight in the Nasdaq-100, with Apple (AAPL) having the largest weight at 9.05%.
4. Problem
Historical weightings data for the Nasdaq-100 constituents is not publicly accessible.
5. Goal
Construct a custom index replicating the Nasdaq-100 using solely historical constituent data to accurately determine component weights for any specific date.
6. Initial data requirements
- Nasdaq 100 (NDX) end-of-day (EOD) data.
- List of Nasdaq 100 constituent companies.
- EOD data for each constituent company.
- Market capitalization data for each constituent company.
NDX constituent changes will be disregarded for this analysis. This decision is based on three factors:
- The primary goal is to assess the correlation between our custom index and the original NDX, and significant deviations would necessitate adjustments.
- Focusing on a relatively short timeframe minimizes the impact of constituent turnover.
- Changes typically involve lower-weighted companies, reducing their overall effect on the index.
7. Selecting the data
Original NDX data was sourced from both twelvedata and yfinance.
Comparing EOD data for consistency and accuracy.
Data validation confirms a perfect correlation and consistency.
Next, we will obtain EOD data for each constituent company using financialmodelingprep.
8. Constructing replicating index
Now that we can calculate the weight of each asset for any specific date, we can create a new index based on market capitalization. We’ll then measure the deviation of this new index from the original one.
9. Comparing with original
The initial correlation analysis yielded a coefficient of 0.96 between the NDX and our custom index. While this indicates a strong positive relationship, further refinements, including methodological improvements and the incorporation of constituent changes, can enhance the method’s accuracy.
10. Final thoughts
The ability to construct a closely replicating index of the Nasdaq-100 opens up new avenues for financial strategies. By employing algorithmic rebalancing based on advanced valuation and risk metrics, investors can create custom indexes that offer potentially superior flexibility and performance while preserving the core diversification and stability of the Nasdaq-100.
