Cryptocurrencies are volatile. Further, in part because of their decentralisation, it can be very difficult to determine what makes crypto “tick”. One theory has it that this problem could be solved by algorithms applied to social media.
Cryptocurrency as a Speculative Asset
Many of us are waiting for the day when large-scale crypto adoption leads to people using crypto as a currency. Indeed, some people already do. Many people, however, treat crypto as a high-risk, high-reward investment.
The thing is, most investment classes are pretty narrowly defined. Further, because they have been traded for so long, we even know how they impact one another. People who own oil can usually guess that conflict in oil-producing countries will make the price change. People who have stock in companies might know that things like the US-China trade war or Brexit may impact value. Often, when people sell off things like stocks, they reinvest that money in a safe haven asset or “hedge”. One of the best-known examples is gold. When stocks go down, gold usually goes up.
Cryptocurrencies like Bitcoin don’t work in quite the same way. This is true for a number of reasons. Perhaps most notably, they haven’t been around forever and they aren’t based in any one location.
Why Crypto Doesn’t Work Like Oil or Gold
Cryptocurrencies haven’t been around for long enough for us to solidly understand how they work with the stock market.
There’s an increased consensus among experts that cryptocurrency could be emerging as a new hedge. However, the value of crypto doesn’t mirror the value of stocks as closely as gold.
Some experts have suggested that crypto rallied earlier this year because the US-China trade war introduced uncertainty into conventional markets. However, the trade war has lasted almost as long as the Trump presidency. During this time, the price of Bitcoin has fluctuated wildly.
The values of cryptocurrencies are also hard to predict because of their decentralised nature. Massive Bitcoin adoption recently took place in Venezuela and Turkey because Bitcoin is more stable than some local fiats.
However, the value of Bitcoin went down some US$2k after President Trump and Federal Reserve Chair Jerome Powell called for more cryptocurrency legislation. The price fell similarly late last month after the Iranian government confiscated mining rigs en masse.
In some ways, the volatility of most cryptocurrencies is its own problem in terms of prediction. US President Trump and Powell made those comments two weeks ago and Bitcoin is on the rise as of this writing. The seizure of rigs in Iran was on 28 June, and by 1 July the value of Bitcoin had recovered.
So, how do you predict something new and international? Maybe by understanding something new and international.
Algorithms and Social Media
Social media is a few years older than crypto and its every bit as international. As a result, money managers are looking at using it to predict changes in the value of crypto. The idea is called “sentiment analysis”. It involves using social media as a source for inputs in equations solved by sophisticated computers algorithms. These algorithms could then tell traders when to buy and sell.
I’ve been talking about this like its a thing of the future. Really, it’s happening already. However, it isn’t common practice (yet) and developers have a lot of work to do.
Social media is huge. There are dozens of major platforms and each one has millions of active users making multiple posts per day. Any effective algorithm would need to sort through all of those posts to find those pertaining to cryptocurrency. It would then have to sort through all of those posts to separate important info from “noise”. Developing computers that can do that kind of operation takes a lot of time and a lot of money. As a result, many investors are just doing things the old-fashioned way.
The idea of a robot that can tell crypto future is an interesting one. Maybe one day it will happen. However, for now, it seems out of reach. At least, until someone develops a cryptocurrency that rewards miners for doing the transactions necessary to predict cryptocurrencies…