Thesis: Price determinants of Bitcoin

[nasdreas]

Hi all,

To finish my master I have to write a master thesis economics. The subject of my thesis is Bitcoin and to be more precise the study attempts to find out which factors significantly influence (determine) the price of Bitcoin. In this post, I will give a brief overview of my research proposal. I would like to start a discussion and find out whether or not I should include other variables that are expected to be connected with the price of Bitcoin.

Regarding prior studies:
Prior studies highlighted several perspectives among which: economical, technical and behavioral related variables. 
- Economic related variables (Interest, Dow Jones industry average, gold price, major indices) resulted in mixed results / no strong influence. Bitcoin economical related variables such as trading volume and trade exchange ratio were positively related with the price. 
- Technical (mining difficulty, hash rate): in general, the technical related variables had a significant positive impact on the price of Bitcoin.
- Behavioral related (Google search queries, Wikipedia, Tweets) very significant and positively related with the price of Bitcoin.

Summing related works, it appears that the price of Bitcoin is mainly driven by behavioral related variables ('investors attractiveness'). Hence, some authors state that Bitcoin is highly speculative.

In order to extent prior work, I aim to contribute in the following way:
- Use a longer time horizon that includes the increase of volume of 2017 and decrease in 2018.
- Use a new behavioral related variable that has not been studied before.
- Perhaps differentiate between time periods

So far, I include the following variables in my research:
- Technical: Hash rate
- Economical: Shanghai Stock Index, S&P500(?), trading volume;
- Behavioral: Google search queries and a new unexplored variable

The decision to include above variables is based on literature review.
Probably a VeC or Var model will be used...tbd

Im currently looking through all the materials that have been published so far. If you think this is incomplete please say so, suggestions are welcome!

Other question: on blockchain there was a metric know as the "Trade Exchange Ratio" (https://www.quandl.com/data/BCHAIN/TVTVR-Bitcoin-Trade-Volume-vs-Transaction-Volume-Ratio). However, the ratio is not being supported anymore. Does anyone have a clue if there is another data source ?

[1713594550]

1713594550

[1713594550]

1713594550

[1713594550]

1713594550

[1713594550]

1713594550

[1713594550]

1713594550

[tophe15]

Nice thesis out there my friend, determining the factor of how bitcoin is influenced by something is really tricky for my opinion. It may be due how people are using it, how big companies are accepting it, how one country accepting it, and on the other how they are prohibiting it. In your variables, I think it's somehow good and I can't suggest any at this time due to my lack of knowledge also of how really bitcoin is being affected by this influences and factors.

[HeRetiK]

Be sure to share your thesis once you've got your master's degree

A few thoughts:

1) Take a close look at timeframes when it comes to behavioural related indicators. It has been brought up before that an increase in google searches leads to a price increase, however it is just as likely that the cause and effect are the other way round -- ie. the price drives the increase in google searches. It would be interesting to see whether either causality can be debunked or confirmed when looking at the concrete time frames more closely.

2) The following economical variables may also be interesting to look at:
 a) Gold
 b) Closely related industry sectors, eg. NASDAQ, finance
 c) alt coins (trading volume, popularity, the rate at which new coins and tokens emerge)

[nasdreas]

Be sure to share your thesis once you've got your master's degree

Of course

A few thoughts:

1) Take a close look at timeframes when it comes to behavioural related indicators. It has been brought up before that an increase in google searches leads to a price increase, however it is just as likely that the cause and effect are the other way round -- ie. the price drives the increase in google searches. It would be interesting to see whether either causality can be debunked or confirmed when looking at the concrete time frames more closely.
yes, i'm aware of the causality issue. Probably a granger causality test among others will be used. Thanks

2) The following economical variables may also be interesting to look at:
 a) Gold
A lot of mixed results regarding Gold. Some studies suggest Bitcoin has properties like gold i.e. being a safe haven. In general, the relationship between gold and Bitcoin is weak

 b) Closely related industry sectors, eg. NASDAQ, finance
Do you mean by closely related, sectors that make use of blockchain technology?
 c) alt coins (trading volume, popularity, the rate at which new coins and tokens emerge)
Will take a look at this

Thank you!!

[squatz1]

Honestly, while it may look simple to comprehend on the top layer, it isn't when it comes down to bitcoin and the power players who push the price up and down. But in reality, there are probably a ton of groups (politicians, bankers, traditional investors, people from the mainstream media, miners from all places, exchange owners, and so on and so forth.

So I mean I really can't wait to see this, it'll be eye-opening for me and many others in cryptos

[nasdreas]

Honestly, while it may look simple to comprehend on the top layer, it isn't when it comes down to bitcoin and the power players who push the price up and down. But in reality, there are probably a ton of groups (politicians, bankers, traditional investors, people from the mainstream media, miners from all places, exchange owners, and so on and so forth.

So I mean I really can't wait to see this, it'll be eye-opening for me and many others in cryptos

Interesting point. Indeed, on the top layer I include trading volume and it doesn't differentiate between groups. It would be very valuable if that information would be available as it indicates groups to follow or have an eye on.

[HeRetiK]

[...]

 b) Closely related industry sectors, eg. NASDAQ, finance
Do you mean by closely related, sectors that make use of blockchain technology?

Thank you!!

No, I actually mean any sector that is related to information technologies, computer hardware and finance. Maybe take a look at startups in those sectors as well, although I'm not sure how much of this information is public.

As far as I know none of the American companies in this space are publicly traded yet. (Ignoring companies that simply put "Blockchain" in their name such as the Long Blockchain Corp FKA Long Island Iced Tea Corp and companies that so far only dabbled with permissioned blockchains such as JP Morgan. Come to think of it, maybe it does make sense to look at companies that only dabbled with blockchains after all.)

Other metrics that just came to my head:

1) Wallet downloads
2) Amount of visitors to Bitcoin related websites and exchanges (eg. using Alexa ranking)
3) Subscriptions to cryptocurrency related subreddits

Just throwing some ideas out there, not sure how much sense it will make for your work in practice.

[coolcoinz]

From what I've observed the main two factors are the politics and rich traders called whales. The politics dictates everything and you can see it by analysing the charts and comparing them with the news. The biggest pumps were the result of political decisions. The Chinese market opening in 2013, the Japanese and Korean markets opening in 2017. We have a similar situation when we compare crashes. Silk Road being taken down by the government in 2013, Chinese banks cutting the flow of money to the exchanges in 2014, China going against mining in 2018 along with Russia. All these form tension in the market that is abused by wealthy traders who either pour money into the market or drain it mercilessly.

[buwaytress]

No, I actually mean any sector that is related to information technologies, computer hardware and finance. Maybe take a look at startups in those sectors as well, although I'm not sure how much of this information is public.

As far as I know none of the American companies in this space are publicly traded yet. (Ignoring companies that simply put "Blockchain" in their name such as the Long Blockchain Corp FKA Long Island Iced Tea Corp and companies that so far only dabbled with permissioned blockchains such as JP Morgan. Come to think of it, maybe it does make sense to look at companies that only dabbled with blockchains after all.)

Other metrics that just came to my head:

1) Wallet downloads
2) Amount of visitors to Bitcoin related websites and exchanges (eg. using Alexa ranking)
3) Subscriptions to cryptocurrency related subreddits

Just throwing some ideas out there, not sure how much sense it will make for your work in practice.

I was just going to comment on this, but Heretik's post probably better explains the other variable I feel you need to take into account, which is "number of market users/players". The timelines you use, if you include 2017/18, will have a distinctly difference in terms of market makers/users. Cambridge is the only academic reference I know of (2017) that tried to put a number to unique Bitcoin users, but I daresay this number dwarfs those of previous periods/timelines. It's an important variable because the price swings in 2011 would be much easier to manipulate or be impacted by a handful of players (or even one major exchange) than it would be today where even the largest exchange can only do so little.

You can then apply the same reasoning to your other variables, if you believe they have influence: hashrate, etc.

And yes, please share. A lot of students come here looking for feedback, I always try to participate, but we seldom get anything back.

[nasdreas]

No, I actually mean any sector that is related to information technologies, computer hardware and finance. Maybe take a look at startups in those sectors as well, although I'm not sure how much of this information is public.

As far as I know none of the American companies in this space are publicly traded yet. (Ignoring companies that simply put "Blockchain" in their name such as the Long Blockchain Corp FKA Long Island Iced Tea Corp and companies that so far only dabbled with permissioned blockchains such as JP Morgan. Come to think of it, maybe it does make sense to look at companies that only dabbled with blockchains after all.)

Other metrics that just came to my head:

1) Wallet downloads
2) Amount of visitors to Bitcoin related websites and exchanges (eg. using Alexa ranking)
3) Subscriptions to cryptocurrency related subreddits

Just throwing some ideas out there, not sure how much sense it will make for your work in practice.

I was just going to comment on this, but Heretik's post probably better explains the other variable I feel you need to take into account, which is "number of market users/players". The timelines you use, if you include 2017/18, will have a distinctly difference in terms of market makers/users. Cambridge is the only academic reference I know of (2017) that tried to put a number to unique Bitcoin users, but I daresay this number dwarfs those of previous periods/timelines. It's an important variable because the price swings in 2011 would be much easier to manipulate or be impacted by a handful of players (or even one major exchange) than it would be today where even the largest exchange can only do so little.

You can then apply the same reasoning to your other variables, if you believe they have influence: hashrate, etc.

And yes, please share. A lot of students come here looking for feedback, I always try to participate, but we seldom get anything back.

Again thanks all. I'm going to dig a little deeper in variables that are 'usage' related variables i.e. wallet downloads or number of unique Bitcoin users (though quite hard to estimate).
haha, i will create an agenda item that reminds me of sharing the information.

[jefbigbob70]

Factor aggravation of the political situation, increasing the likelihood of armed conflict (North Korea) adversely affects the course bitcoin.

Legislative restrictions on turnover bitcoin (China, the US prohibition of ICO).

[Olrac]

Several factors that affect bitcoin's price are:
1. Positive and negative media news
2. Market manipulation by big investors or "whales"
3. Current economic situation
4. Wide acceptance by the public
5. Supply and Demand

[alyssa85]

There is one element that you have missed out - and that is world events.

Bitcoin was launched just after the great financial crash in Jan 2009 - it's likely that if we hadn't had the credit crunch, it would have never been launched.

It then remained a pure nerd currency until 2013, when the EU decided to seize the savings of the people of Cyprus to bail out their banks. People went bankrupt, as did businesses. There were stories of businesses who had their payroll in their current account ready to send it to their employees and the EU seizing all the money to bail out the banks.

That is the real point bitcoin took off - it's based on distrust of corrupt authorities. Without orgs like the EU stealing people's money, bitcoin wouldn't be as popular, it would be just like second life tokens, a curiosity.

[nasdreas]

There is one element that you have missed out - and that is world events.

Bitcoin was launched just after the great financial crash in Jan 2009 - it's likely that if we hadn't had the credit crunch, it would have never been launched.

It then remained a pure nerd currency until 2013, when the EU decided to seize the savings of the people of Cyprus to bail out their banks. People went bankrupt, as did businesses. There were stories of businesses who had their payroll in their current account ready to send it to their employees and the EU seizing all the money to bail out the banks.

That is the real point bitcoin took off - it's based on distrust of corrupt authorities. Without orgs like the EU stealing people's money, bitcoin wouldn't be as popular, it would be just like second life tokens, a curiosity.

Interesting point. I assumed that market events would affect the price. Currently I plan to use a financial market stress indicator (e.g. OFR financial market stress index) that generates an aggregate measure of several underlying indices. See: https://www.financialresearch.gov/financial-stress-index/files/indicators/index.html

I also found a data set on sentiment that I'm planning to use in my research. A sentiment score is based on several scanned documents e.g. social media, blogs, forums etc.

Perhaps I can assume that if there are extreme events in the world, unrelated to Bitcoin, financial stress would go up. Unrelated events: north korea for instance.
Regarding related events, or the overall sentiment and stress regarding Bitcoin, I'm planning to use sentiment data.

Update research so far


Its a lot of work and a struggle to find data. I've two issues at the moment. If anyone could help me, I would appreciate it. I pledge to share the results of the research  
Problem 1: data on ICO volume is limited. I can't find a good source that captures amount of funding raised per time stamp. This variable is really important I believe. According to traditional finance literature, it appears that IPO volume raised goes up when the overall sentiment is positive. IPO volume hit a low around 2008. Since it appears that ICOs go up during a price climb of Bitcoin, I believe that a similar relationship exists in the world of cryptocurrencies. Does someone know a good source about ICO funding volume raised?

Problem 2: Google search queries might be related to the Price as well. According to earlier research, it appears that the relative amount of search queries is a good explanatory factor. The problem: Google trends https://trends.google.nl/trends/    only provides daily  data for a timeframe of 90 days. I really need daily data. Does someone know a way to retrieve daily data for long time frames? (5 years +).


In april i'm going to start with the analysis of the data and start with the first regressions. If someone knows a way to solve of the above mentioned problems, I would really appreciate it!!!!

[kuckong]

I think the most important thing that affects bitcoin is the amount of supply and demand. also the existence of negative or positive news involving bitcoin and cryptocurrency. Example yesterday there was negative news about cryptocurrency that banning ads on facebook and google after this news circulating bitcoin price down.