Ain’t No Lie, Buy Buy Buy? A Study on Crowd Behavior

If the end of 2017 taught us anything about crypto, it’s that our friends and family around Christmas dinner, as well as anonymous Twitter avatars telling us that it’s time to buy means the price is probably going to continue up, right? Nope, not at all. In fact, according to Santiment’s groundbreaking Social and Emerging Trends tools, spikes in social media suggesting positive sentiment and declarations that it’s time for everyone to buy has historically meant the exact opposite in the majority of cases.

To prove just how often the crowd gets it wrong, here’s a look at our Crowd Sentiment Index for the Matic Network ($MATIC) over the past two weeks as of the time of this writing. It measures the difference between positive sentiment and negative sentiment related words to show what crowds are thinking about an asset during a period of time. This particular graphic shows positive vs. negative sentiment among Telegram users on topics related to Matic:

We can see a peak in positive sentiment very clearly on December 7th after a long stretch of rising prices beginning on December 2nd. Then as the price dropped off a cliff on December 9th, so did any remaining positivity revolving around MATIC. We all have heard the advice to "buy low and sell high", but this is obviously easier said than done when nobody knows what dictates a low vs. a high, especially in such a young asset class and particularly with even younger altcoins like MATIC. But this is a solid example of what behavioral effects price fluctuations can have on a community.

Let's take another example, this time with Bitcoin and a more zoomed out look on our Social Trends page. Circled are the five largest peaks in mentions of Bitcoin since the summer of 2017, which was the beginning of the most recent bull run. The highest spike was actually two separate peaks as a higher high was reached after the first initial top occured:

We can see peaks in mentions of Bitcoin as prices rise particularly rapidly, and one particular peak when prices fell from a long-running $6,000 support level at the end of 2018. This phenomenon of very little discourse when prices are flat, then an explosion of chatter when prices rise, has existed for the greater part of this decade. To some, this may seem to be common sense. But we actually have the tools to prove just how drastic this is, and how biased this trend tends to be toward the positive end of the spectrum. It's something we can all learn from, and our free and Pro SanFam members can actually benefit and build profitable strategies from.

With apologies to the short list of talented technical analysis traders out there who can actually make accurate short and medium-term price predictions, real successful long-term crypto trading relies heavily on on-chain and social behavior analysis to really make it anything more than a guessing game. Anyone who says they know what the price of Bitcoin, Ethereum, or another asset will be a year from today is lying to you. We promise.

Market conditions constantly change, and we need to be aware that free advice being dished out, telling you precise times to buy or sell any asset needs to be taken with a serious grain of salt. Whales may control the markets, generally speaking. But for the other 99.9% of traders who don’t have enough capital to single-handedly sway the markets, they can choose whether to collectively trail the moves that large cap traders make to cause sudden price shifts. Collectively, small traders can even come together and continue a price movement’s momentum. However, more often than not, small cap traders simply run out of steam to do it effectively and many are left repeatedly “hodling the bags”.

Take this above example. When prices plummeted from $8,200 to $7,500, the community began to panic and discussion increased around whether Bitcoin was worthy of an investment. Much discussion was of the “sell” variety. In the same respect, two days later, there was a much larger spike in discourse when BTC jumped higher than its initial falling point to the high $8000’s. This was a clear FOMO discussion, as impatient traders came out of the woodwork to declare that it was time to “buy”.

With all this being said, it’s time to take a look at how crowd sentiment works. I present to you Santiment’s “Crowd Buy Index”, using the platform’s Social Trends tool. Our Social Trends tool lets you explore the amount of mentions of any keyword or phrase on crypto social media over time. It does this by collecting and parsing messages from 1000+ crypto-related social media channels (Telegram, Reddit, Discord etc), with ~100,000 new messages analyzed each day.

We put together every significant common bullish keyword that typically indicates someone believes that crypto prices are moving up. They include “buy”, “buying”, "long", “moon”, “lambo”, and “bullish”. Albeit, plenty of these mentions on social media are made in jest. But most are instances that won’t skew the vast majority of unironic mentions of these keywords. Here are the combined mentions on crypto social media over the past 3 years:

What a difference a bull run (or what some would consider a bubble) makes. Discussion revolving around longing and buying BTC was relatively reserved to small groups before the summer of 2017 happened. Then, as discussion went parabolic, prices faded rapidly. Though not always the case, there are more often than not instances where the peak of mentions of these positive-sentiment keywords occur within hours of right before or right after a price top hits. It’s very telling that the vast majority of people on Twitter are reacting positively to prices gradually moving up (expecting them to continue this trend), while the large investors (aka whales) silently control market direction and cyclically await the overeager smaller traders to slowly bid up the price for them until it’s time to unload and cash in their profit.

Referring to the previous chart, when BTC’s price suddenly shot above $10,000, social volume inevitably exploded. However, after a large bounce back and a lower high a few days later, the price gradually descended for three weeks back down to the low $8'000’s. This is a case of social volume divergence that our users often look for. Essentially, extreme social volumes - especially during the uptrend - are often indicative of the asset reaching peak hype, with retail/fish money flowing in. If the amount of hype/chatter diminishes in the coming days (as the price tries to chart new highs), the rally is often made unsustainable due to this divergence in price and social trends.

On the other side of the spectrum, what if we take a look at negative sentiment related words, such as: "sell", "selling", "short", "puke", "dead", or "worthless"? This time, we'll use Ethereum as the study:

It's clear from this 6-month sample from the middle to later end of 2019 that mentions and declarations of selling or crypto being dead are likely called with emotion in mind. The red "incorrect call" circles very often occur right after a price dip, and right before the price inevitably recovers. To the community's credit, though, they did have an outlier spike in negative mentions at a high frequency right at the summer top in late June. However, this very well could have correlated with just a heavier increase of discussion on social media overall as cryptocurrency hit its annual top at this time.

Another example of just overall mentions of a token in contrast to its price, is this article about $REN about how social behavior could have been used extremely effectively to catch the top this summer.

It is important to note that nearly all long traders over the past two years have lost money, regardless of what tokens they have held. So when the majority of traders believe an outcome will happen, this is the same majority that are losing trades since 2017. And the people who claim to be up significantly since the beginning of 2018 likely did so by betting against crypto and not longing or hodling assets. Some may have lost far less than others (percentage-wise, regardless of capital) by having the foresight to get out of their holdings at or soon after the crypto bubble burst in January of 2018.

Those who have most often succeed tend to tap into an understanding of how sentiment drives prices in crypto, and they’re very often going against the grain of the crowd. That's why discovering and leveraging effective top indicators like social and emerging trends are crucial for any long-term trader. Finding optimal exit points is a crucial skill in crypto, and betting against the crowd has often been a good place to start. And this often means they stay silent, because why be vocal when you voted for the winning outcome that the majority of vocal people don’t like?

There is strategy to take away from this, and studying social and behavioral trends in mentions of specific tokens and how often the masses recommend buying or selling them is a great way to find ideal spots to enter and exit the markets. Using Santiment's Sonar alerts are a great way to be notified of when social spikes are occuring, and Pro subscribers have access to unlimited amounts of alerts they are able to create. So give us a look and make an account!

Market perception and sentiment related to where prices are going tends to lag fairly consistently with what direction price has most recently moved. One thing that does correlate between traditional markets and crypto is the fact that when the whole market is overjoyed, a drop is usually imminent. On the other side of the spectrum, when everyone is fearful, a rise is often imminent. Warren Buffett is famous for his advice, "Buy when there's blood in the streets, even if the blood is your own." Most people Buffett explains, "...pay a very high price in the stock market for a cheery consensus."