What News Sentiment Tells Us About Crypto Returns

Share:
Indexes
  1. Determinants of Token Prices
  2. Measuring Sentiment
  3. Sentiment and Traditional Asset Classes
  4. Sentiment and Token Prices
  5. Bottom Line
Photo by Kenny Eliason on Unsplash

We’re beginning a series that will report the latest data on behavioral economics and the cryptocurrency market. This initial article will be longer as we’re going to introduce some new concepts, namely the topic of “sentiment.”

Lots of people look at the price of a token to determine how well it’s doing, but the reality is that the price masks a lot of phenomenon.

  • What is the demand for the token?
  • What geopolitical and macroeconomic events are going on?
  • What is the supply of the token?
  • What role does marketing play?

To be an informed investor, it’s important to understand the factors behind price, and then make decisions based on the different factors at play.

Determinants of Token Prices

Every token has certain fundamentals, including:

  • The consensus mechanism (i.e., how do transactions get uploaded onto the blockchain) — different mechanisms have their own costs and benefits, which feed into market demand and supply.
  • The pain point — different tokens solve different problems, and these problems vary in their importance.
  • The team — tokens and blockchain ecosystems are ultimately driven forward by specific people, and how well they get along and innovate.

But, these fundamentals often overlook the role of economic sentiment, which refers to peoples’ attitudes and expectations about an asset. My research has shown that economic sentiment plays a major role in accounting for fluctuations in consumption and real economic activity, building on what is already a large academic literature on sentiment and security prices.

It’s not surprise that sentiment influences and/or is informative for understanding the price of a token, too.

Measuring Sentiment

There is not a single definition of sentiment, but it broadly refers to beliefs about an asset. Those beliefs can be about the future state of an asset (or the economy), or they can be about a current state.

My prior research has used data from Gallup to quantify the effects of gas prices on peoples’ attitudes about the economy (sentiment), and similarly how sentiment influences consumption. Not surprisingly, greater optimism about the economy leads to greater consumption!

Setting these background details aside, you’ll see new data presented from Refinitiv, drawing on their MarketPsych model analytics. They produce a measure of sentiment for each day and across asset classes (conventional ones, like the S&P 500, and cryptocurrencies, such as BTC and ETH), which is an index that hovers above and below zero.

The index is constructed by feeding every news article and social media through their natural language processing (NLP) algorithms, tagging words and phrases as high or low sentiment. They also disaggregate across different types of sentiment, including trust, uncertainty, urgency, and more.

Sentiment and Traditional Asset Classes

Economists and social scientists have studied the stock market for a very, very long time. However, the role of sentiment remains an active area of study.

Price can depart from value — sometimes for much longer than anyone could imagine! Below is the time series of the S&P 500 (stock index) and an index of sentiment that combines both formal news outlets and social media.

Let’s take note of two factors.

  • Market sentiment here is extremely volatile. You’ll also see the same sentiment pattern if you go back to 2000 — or even 1990.
  • Security prices are generally increasing, but they do not track sentiment very closely — in fact, there are periods when sentiment is increasing and price could increase, but in fact falls.
Authors calculations through Refinitiv

Sentiment and Token Prices

We’re going to see very different phenomena in the cryptocurrency market. Let’s first consider a comparable time series for Bitcoin.

Unlike traditional asset classes, there is a closer correlation between the price and sentiment of Bitcoin. Most recently, we’ve seen a decline in the price of BTC since November 2021, but it climbed back a little in March.

Interestingly, sentiment has generally been above zero (suggesting positive news), which is an indication that the price might come back up, but we have admittedly seen a dip in the most recent days. One reason we’ve seen positive sentiment could be the rise of inflation that affects traditional asset classes, which we have gone in to detail about here.

Authors calculations through Refinitiv

Next, let’s consider Ethereum. Here, we actually see a bigger contrast between sentiment and price, but that is probably because ETH is a newer asset, relative to BTC, and investors are still learning about it.

As of the start of May, we’ve seen a spike in sentiment, but time will tell if the spike remains persistent. That said, we have seen an overall increase in sentiment despite the volatility — the index has generally been above zero.

Authors calculations through Refinitiv

Bottom Line

Critics of cryptocurrencies would potentially take these data as evidence that they are speculative assets. However, a strong link between sentiment and price is not necessarily bad — or indicative of speculation. That can only be determined when we know the true value of an asset.

Speculation is when the price of an asset gets bid up far above its value, but without knowing that true value, we’re just “speculating.”

Here are three takeaways since late April.

  • Traditional asset classes have declined and are below their pre-Christmas high, and there is significant volatility in their sentiment.
  • There is a stronger connection between sentiment and price for BTC and ETH, which may reflect that cryptocurrency holders are more represented in the market than traditional asset classes.
  • While BTC exhibited a recent swing down in sentiment, ETH is up, which could be due to its recent announcements about ETH 2.0.

This article was written by Christos A. Makridis, the Chief Technology Officer and Head of Research at Living Opera. He is also a research affiliate at Stanford University’s Digital Economy Lab and Columbia Business School’s Chazen Institute, and holds dual doctorates in economics and management science & engineering from Stanford University. Follow us at @living_opera! (This is not financial advice!) Thank you to Refinitiv/MarketPsych for their data!