Nasdaq IXIC and Bitcoin Price Movement Patterns EDA, Cross Correlation

 Introduction

For my personal interest, I've done a quick EDA on the pattern in the relationship between Nasdaq IXIC and Bitcoin price movement.

I collected and looked into one minute granularity price change data for the past 30 days.

This analysis is a part of my journey to find out whether I can make use of the similar price change movement of Nasdaq IXIC and Bitcoin as my Bitcoin scalping trading strategy.

Questions

1. In what condition Nasdaq IXIC and Bitcoin become coupled for a while:

    

    - Does the direction of the price change movement seem to matter for their coupling?

    

    - Does one seem to start 'following' the other's movement when its movement gets too different from the other's?

    

    

2. As a part of the answer to the question, which follows which one's price change:

    - Is there any sign of a moment one's movement is lagged following the other's movement?

Data Collection

I scraped the required data as follows:

1. Nasdaq IXIC index price data : Yahoo Finance API

    -One minute granularity price for the past 5 days (2022-02-24 ~ 2022-03-18)

2. BTC price data : Binance API

    -One minute granularity price for the past 5 days (2022-02-24 ~ 2022-03-18)

(please refer to kaggle for the scraping code, csv files)

Tools Used

Python (pandas, matplotlib, sklearn)

Time Spent for the Project

1 day : 2022-03-20 ~ 2022-03-21

Python Code

Kaggle : https://www.kaggle.com/code/dankdata/nasdaq-ixic-and-bitcoin-coupling-pattern

Results

1. The direction of the price change movement, whether the price goes up or down, does not seem to affect the degree of correlation between Nasdaq IXIC and Bitcoin's price movement.

2. There seems a tendency that the two's movements get more correlated one to three hours after they show a relatively great degree of difference in their movement.

3. Although it's difficult to say that Bitcoin follows Nasdaq IXIC's movement with only one case and a moderate correlation, there is a sign that it may do. I found one case where it follows Nasdaq IXIC's price change with an 8 minute lag.

Summary of Analysis

<Hourly Correlation and Hourly Mean of Percentage Price Change, One Minute Granularity, Nasdaq IXIC and Bitcoin>

It seems that there seems to be a pattern of the two getting similar one to three hours after one deviates far from the other. After a relatively great degree of difference in their movements, the correlation coefficient becomes higher as shown on the above plot.

<Lagged Correlation at 2022-03-14 13:00, Bitcoin Lagged>

I found one case where there is a moderate lagged correlation when Bitcoin's price change is lagged. At 13:00 UTC 2022-03-14, with an 8 minute lag, Bitcoin and Nasdaq IXIC price change movement showed over 0.5 correlation coefficient.

<Price Change of Nasdaq IXIC and Bitcoin, Bitcoin with an 8 minute lag>

The above plot shows how similar their movements were at the significantly lagged-correlated hour when Bitcoin's price change is lagged by 8 minutes (shifted 8 times).