Crypto-ML has rolled out updates to the machine-learning models across the board. Over the past month, we have seen a new behavior in price, characterized by channeling and consolidation. This was followed by a sharp drop on October 10. Based on this, several changes in how the models interpret the data and behave are now in place. This post covers these changes and a look at price behavior.
The Larger Cryptocurrency Trend Picture
The challenge with recent price behavior is that cryptocurrency has been historically characterized by extremely strong movements. Capitalizing on large upward movements has been the winning way to profit over the long term. Specifically, this means anticipating bullish breakouts.
However, when price is channeling, the strategy of looking for breakouts may backfire. Crypto-ML has been looking for price to extend above a range. Instead, price has been bouncing within a range. During these conditions, different strategies perform better. The opportunity the Crypto-ML team has been exploring is how to perform within ranges while also being able to capture breakouts. This may seem like a subtle difference, but the technical drivers behind these scenarios are highly complex.
Some recent posts have discussed these market conditions and how the machine learning models are enhanced and updated:
- [Video] Bitcoin Auto-ML Machine-Learning Model Enhancement Live
- How Does Crypto-ML Handle Narrow Trading Ranges?
- Machine Learning Enhancements for Ethereum and Bitcoin Cash Now Live
Cryptocurrency Trading Model Enhancements
Across Bitcoin, Ethereum, Bitcoin Cash, and Litecoin, new learnings from recent data patterns have been incorporated into the machine learning trading models utilized by Crypto-ML. These changes are live now.
Market Conditions and Caution
It is important to note that 2018 is characterized by a bear market and we are still in an overall bearish environment. This means long trades are swimming against the current.
As previously shown, traditional technical analysis shows us that on the daily charts, we are in a bearish triangle formation. Earlier today, it appeared a bullish breakout above the triangle was occurring, but prices have since retreated back to within the larger standing pattern.
Auto-ML saw this as a breakout opportunity on very large volume and opened a trade. It is yet to be seen how price will play out for the remainder of the day. But the signal did the right thing in terms of attempting to grab a breakout. As with other types of signals, it is expected you will see numerous attempts to grab a big movement. While these may result in small losses, the expectation is the larger gains offset them by a considerable amount.
Of additional note, the area just above $6,100 seems to be a strongly established floor going back to July. Once the triangle shape narrows to a point, we will really see it tested and find out if price will settle to a new bullish or bearish pattern.
Regardless, we strongly urge caution. As we have been anticipating, price is at a very pivotal point and will likely react strongly one way or the other.
Today’s Large Movement
To further elucidate the fast changes in price and provide some potential context behind the spike from earlier today, there was an enormous transaction that moved 12,219 BTC (around $80M USD).
That transaction can be seen on Blockchain.com here.
Reactions to this move may be why we are seeing such different prices across various exchanges. Our team has seen some quotes around $6,900 while other exchanges were quoting closer to $6,400. That is a much larger arbitrage spread than would be typically found.
Crypto-ML Standard vs Auto-ML
If you haven’t already, we recommend you review Auto-ML to determine if it is right for you. During the recent channeling, it has made numerous attempts at capturing a breakout, hinting something is coming. However, those breakouts did not occur. That said, it is quick to close out ahead of sharp drops and will likely be the first to signal an upward trend. You can read about the various Bitcoin Trading Models here.