This AI Model Can Tell You When To Own Certain Cryptocurrencies – Benzinga

The rapid influx and equally hasty exodus of massive amounts of capital to and from cryptocurrencies have left the market with whiplash. Thanks to the precipitous drop from the December 2017 highs, many retail traders now find themselves stymied as to how to approach the still nascent asset class.

One of the challenges is that cryptos have simply been too volatile for traders to rely on many traditional technical catalysts. Add to that the sheer amount of coins, their subtle interplay with one another as well as the challenge of identifying which coin has the most upside, and crypto trading can sometimes start to feel like throwing darts at a board.

However, tools do exist for crypto traders to gain some insight into the potential behavior of the array of digital assets. One such tool from investment research platform Quantamize implements AI-machine learning algorithms to offer three-day buy and sell signals on 30 cryptocurrencies, a section of which you can see below.


Image courtesy of Quantamize

This chart of Quantamize’s Crypto Trading Signals tool shows signals updated as of March 20, 2019. The signals, which break down along either a “Buy” or “Do Not Own” binary, are generated by way of the aforementioned machine learning AI algorithms constantly testing a cycle of predictive models for potential future performance against historical and concurrent price action.

Like the currencies it tracks, the Crypto Trading Signals tool attempts to provide an analysis stripped of human influence and bias. In service of this, crypto traders can get an exact picture of how successful each coin’s model is in predicting its three-day trajectory with the accuracy score, listed to the second column from the left. This score is generated from the algorithm’s success rate in anticipating price direction, both historically and at the moment, divided by the total number of predictions made by the algorithm within its 3-day purview.

The models also provide charts of each model’s simulated performance and hypothetical returns against the actual asset. You can see below that buy/sell signals from the Litecoin model have outperformed the coin’s actual performance dating back to 2017.


Image courtesy of Quantamize

From the end of 2017 to mid-March, 2019, the current algorithm’s signals have managed to outperform litecoin’s buy-and-hold price by more than 130 percent, with its hypothetical active return over the litecoin benchmark currently above 30 percent.

Each of the 30 cryptocurrencies Quantamize’s AI Crypto Trading Signals tool provides exactly this kind of insight. So while the mechanisms of cryptocurrency price action may remain somewhat obscure and mystifying for the time being, traders looking to actively trade the asset class can find an edge in alternative tools.

Quantamize is a content partner of Benzinga