Cryptocurrency has gained an immense amount of popularity in recent years, however there is still plenty of research to be done to understand their markets and all of their intricacies. Bitcoin, arguably the most popular of cryptos, has had their fair share of boom and bust cycles. Given its growing role as a currency and investment, many are seeking to gain an edge with price predictions. The objective of this group is to predict the price movement of bitcoin through its correlation with three other markets/factors- gold prices, GPU prices and inflation rates. The data used comes from four separate datasets- one for bitcoin prices and then one each for the other three factors.
Things I like about the project:
This is an interesting project from a fundamental analysis point of view. I personally believe that using technical analysis only for price prediction in equity markets is incredibly difficult as there are many hidden variables and "past performance does not guarantee future results." However, exploring fundamentals through alternative data is always and interesting option to try and find an edge, and hopefully that applies to bitcoin as well. I don't have much previous knowledge in crypto but will be interested to see if you find meaningful conclusions in your project.
I like how there is a central objective (price prediction) but you are going about it in different ways (i.e. the different markets/factors). This allows you to explore multiple relationships or even use a combination of the variables to drive your predictions.
Like that you used references in your proposal- easy to look up more information.
Suggestions:
Accurate pricing data is crucial to this presentation. Make sure that you can capture data in several different market conditions (i.e. "bear vs bull").
When using NVDA stock price as a factor to look at the relationship between GPU pricing and bitcoin price, make sure that the potential causality is in the direction you are looking for. In other words, check that it is the change in NVDA (or an ETF with holdings in major GPU suppliers) is what then causes a change in bitcoin pricing and not vice versa. During the bubble last year, certain stocks fell because bitcoin was falling, rather than causing bitcoin to fall.
It might be interesting to try and develop a trading strategy using buy/sell indicators that your research provides from the price prediction. Maybe discuss whether or not you think it would be a viable option given the accuracy of your algorithm, or if more data/variables would be needed.
Cryptocurrency has gained an immense amount of popularity in recent years, however there is still plenty of research to be done to understand their markets and all of their intricacies. Bitcoin, arguably the most popular of cryptos, has had their fair share of boom and bust cycles. Given its growing role as a currency and investment, many are seeking to gain an edge with price predictions. The objective of this group is to predict the price movement of bitcoin through its correlation with three other markets/factors- gold prices, GPU prices and inflation rates. The data used comes from four separate datasets- one for bitcoin prices and then one each for the other three factors.
Things I like about the project:
Suggestions: