Sentiment Trading

In trading, a common strategy is to identify market sentiment and use it to your advantage. Sentiment is the tone of the market, i.e. bullishness or bearishness of participants.

Sometimes tradeable sentiment is very informal, and can be as simple as getting questions from personal friends concerned that their portfolios are down. Other times, and more usefully, it can be formal data provided by trading institutions on long versus short positioning.

The traditional market uses COT data (Commitment of Traders) for many trading products. This data shows the exact positioning of the institution’s trading participants, which can then be used in analysis by traders. The cryptocurrency ecosystem offers similar tools which we’ll describe in a bit.

Longs and shorts are made up from margin trading participants, meaning people who are borrowing funds to further leverage their positioning. Usually spot positioning (non-leveraged holdings) are not included in such analysis.

If someone is long or short, and the trade goest against them, then they will either hit their stop and exit the trade, or eventually be forced to exit by liquidation or margin call. Since traders are borrowing funds, they have a limited tolerance for downside before they run into the maximum borrowing amount and can’t hold the position any longer.

Squeezing leveraged traders

One extremely common analysis is to identify an overweight position in the market, and expect there to be some form of squeeze of the heavily weighted side.

If 80% of participants are short, then if the price goes up, a huge percentage of participants are now sweating their positions and must close (by buying), which drives the price up higher, initiating a cascade effect that puts even more pressure on those who are still short. The squeeze continues until enough people have been forced to buy (exiting their trades and driving up the price), and the short term trend exhausts itself.

A squeeze can often happen as a counter-trend to the prevailing one. For example, if the trend is down, but the trade gets “too crowded” with shorts, there may be a forceful counter-trend rally that is a squeeze of those shorts, only to continue down more once the “late shorts” are squeezed out.

Timing a squeeze can be hard, but being able to identify when you’re at risk of being squeezed, or protecting profits in a winning trade before a squeeze, can be really beneficial to your bottom line. Therefore it’s nice to be able to identify market sentiment so you can be cautious of such a scenario.

Counter-sentiment trading: the contrarian trader

Sentiment based trading doesn’t just have to be for the purpose of squeezing overweight positioning of leveraged traders. It is also part of what makes the market so fascinating from a psychological perspective.

The common narrative — and popular sentiment — is often wrong, or at minimum it’s likely late in the trade if it’s still working while the entire world is excited about it. So you know once you’re on the same side of the trade as the rest of the market, the prevailing news coverage, and popular narrative, there is a good chance the profitable part of the trade is nearing its end. It would look a lot like Bitcoin in the second half of 2017, though that was a particularly strong example.

So while “the trend is your friend”, it’s important to not get stuck in too crowded of a trade, whether from an actual positioning perspective (as shown in the squeeze section), or in the “late” part of the trade where it’s the popular trade to be in. A crowded trade will often show itself on the chart anyway, through exuberant pressure in one direction, or price being extremely far from the “mean”, destined to come back to reality.

Contrarian traders often look for what will be the popular trade or in the news next, not what’s the popular trade now. Being a contrarian trader can be a slow and arduous process, that requires great patience, and a likelihood of being too early; but it can also be a very successful methodology. Contrarian traders don’t find their big wins by following along with the popular sentiment.

Tracking sentiment in crypto

In crypto, we have a few ways to measure similar data and use it to our advantage:

Longs vs Shorts on Bitfinex

Bitfinex publishes longs and shorts via their API, and that information is available within TradingView for analysis.

We can get see the number of longs and shorts active in a give timeframe, and also figure from that data what percentage of traders are underwater or in profit from their position, based on calculating the difference of longs or shorts versus the price when they were added.

Longs vs shorts at Bitfinex shown on bottom.

The bottom part of the chart above shows the Long/Short ratio on Bitfinex, with longs in green and shorts in red. You can see where I’ve put the red vertical line, price was testing the bottom range and consolidating, and shorts were building to what would become an all time high level.

This is a ripe scenario for a squeeze. Sentiment was very bearish, testing the level that had previously caused a sharp “V bottom” reversal. This time price was grinding at the lows, and looked ripe to go down. Shorts stacked up in anticipation, and the opposite happened.

Watching price alone, the bearish outlook isn’t a bad one. But seeing all time high levels of shorts is a warning sign for the trader seeking an edge with sentiment analysis.

Everyone expected down, shorts piled on, and therefore when it went the opposite way, it was fast and vicious, triggering stops, liquidation, and forced buying to propel price further, punishing everyone who was short. A classic short squeeze occurred.

Underwater shorts shown in the box

By the time the price reaches what would become the high, short positioning finally gets back to levels not seen since the prior “double top” breakdown. You can visualize with the box how all the added shorts after that time were underwater, and the market didn’t reverse until roughly that same number capitulated. While the exact levels are arbitrary, you can see in that image how to measure underwater positioning.

Simultaneously, it’s worth noting that longs got squeezed before the red long on a particularly sharp down day, and then once the market moved up, longs were quickly taking profit as it went, rather than adding to their positioning.

If we move to the right on the same chart, we can see two examples where the market dumped once shorts were at significant lows, again acting directly counter to speculative positioning.

Shorts at the lows, and the market finally dumps

The first line is similar to our prior example. but in reverse. Longs were fairly high but not a clear peak, but shorts were extremely low relative to the rest of the year, and that’s when the bulk of the drop happened. Then, while shorts rapidly piled in, the market grinded higher, slowly putting those positions in trouble until many capitulated.

The second green line is one of pure exhaustion. Longs were starting to trend higher but weren’t high by any real measure. Shorts simply gave up. The low volatility market created a lack of interest, and that’s when the market tanked and made the “big” move down many were seeking through speculative short positions the prior trips to this level.

Readers who were part of Learn at this time may also remember our notes expecting high volatility incoming, because current volatility was so low. The very low number of shorts and longs wasn’t particularly predictive but it was another good sign that the big move was near, in hindsight. Nevertheless, a keen eyed trader will be on the lookout for severe short/long levels in trading.

Perpetual Swap funding

Bitmex and Deribit use a funding mechanism on their perpetual swaps products. A perpetual swap is a leveraged trading product that has no expiry, meaning the person holding it can do so indefinitely, unlike a futures contract which has an expiry date (typically quarterly).

The price for the perpetual swap is designed to stay close to the “spot price” — or price on regular exchanges. In order to maintain the price consistency, the derivative product (the swap) uses a clever funding mechanism to incentivize trading participants to keep the price close to spot.

If the price is above spot, it means there is more long demand and those who are long must pay those who are short a fee relative to how out of whack the price is. Funding is a percentage of overall trade size, such as 0.1% per period. These funding fees can add up significantly over time.

Whether current funding is paying longs or shorts is an immediate and simple signal for trading sentiment, and can be used similarly and in concert with the Bitfinex data.

Deribit is a much smaller exchange but has similar funding, but it is instantaneous rather than every eight hours. Trends can still be analyzed using this information but it’s not the same type of event as Bitmex’s funding times, which are often gamed and points of short-term volatility.

Funding times on a 30 minute chart

If a trader is short, and also other traders are short, they are forced to pay longs for the privilege of being short, so it is doubly concerning to have the price go against you. Often times severe short term price adjustments align with the funding interval, especially when it’s a big shift in funding, as traders make adjustments to either circumvent or take advantage of funding for the next round.

Similarly to the Bitfinex data (they often align), if funding gets extremely one sided, it can be a good sign that sentiment is too severe in one direction and ripe for a squeeze.

Tracking funding is not our first concern as traders but it is a useful nugget of information to make us more informed.

Non-technical sentiment tools

The crypto landscape has plenty of opportunities to measure sentiment in less technical ways. Some that I either track myself or just generally try to stay apprised of:

  • Twitter trader sentiment (how bearish or bullish is crypto Twitter)
  • News headlines (they are almost always late to the trend)
  • Messages or conversations with non-trading friends (asking how to buy when Bitcoin is expensive, asking if I’m okay when it is down)
  • Polls people put on Twitter or elsewhere (countertrading the popular answer some say is effective)
  • Specific contra-indicators (this one is a bit evil… but many of us know a few folks who are simply psychological gifts, often saying stuff right at the brink of trend reversal)

These are all obviously non-scientific, but it’s human psychology nonetheless, and a chart is nothing but psychology visualized. I’m sure I’m missing plenty of good non-technical indicators. I often tweet either “bottom things” or “top things” when I’m commentating myself on good sentiment indicators for reversal — typically purely based on the sentiment of whatever I’m sharing.

By now you should have a pretty good understanding of how you can start using sentiment analysis in your own trading and market analysis.

This is not a replacement for other analysis. It is supplemental. The crowd can be right, but usually not for too long. However, sentiment analysis would have likely told you to stay away from buying bitcoin from about $4,000 on and yet it went to nearly $20,000 in 2017. So the popular sentiment can be “irrational” for a while. Be cautious when using it, and use all the tools in your toolbox.

Resource Hub »