Enigma or riddle? Well, there have been plenty of developments in artificial intelligence (AI) and machine learning (ML), but opinions about them have been quite varied. One can marvel at how AI was able to perform traditionally “human” activities such as winning at poker or writing music. And, Google’s predictive search, which can sometimes make outlandish suggestions and be eerily spot on at times, might make one sceptical too.
However, AI can also be scary, as imperfections in the technology could also lead to harm. Certainly though there is always room for improvement with any technology, but what AI and ML have now are already quite beneficial to various efforts.
It all rather reminds me of the film Enigma, an espionage thriller with Dougray Scott playing a lead codebreaker Tom Jericho at Station X (Bletchley Park) during the Second World War, charged with decrypting encrypted German messages with Enigma and Bomba machines. The latter was a special-purpose machine designed in late 1938 by Polish Cipher Bureau cryptologist Marian Rejewski.
Researchers and a number of companies have recently been embracing AI and ML to drive innovation and solve problems. While previously inaccessible due to the costs and expertise barriers, blockchain-based AI and ML platforms are now gaining traction.
TraneAI, for instance, is leveraging blockchain’s distributed computing to speed up the machine-training process without the need for large centralized infrastructure.
Then we can cite DeepBrain Chain and SingularityNET, who also use blockchain’s decentralization to create a low-cost and accessible neural network to power AI efforts. And, Santiment, which is based in Germany, collects and sells live markets data feeds to crypto traders.
The availability of these tools show promise in addressing market-specific issues.
Yet despite the growth and increased adoption of blockchain and distributed ledger technologies (DLT), the cryptocurrency market remains to be quite the riddle for investors. It is a highly speculative market, which contributes to the volatility of cryptocurrencies.
In such cases, AI and ML’s automation can actually help analyze market data quickly and provide investors with timely insights. Projects such as NeuroBot, Senno and ThinkCoin are specifically focusing on using AI and ML to make better sense of the market.
Traders in speculative markets like crypto could clearly benefit from technical analysis, which uses historical quantitative data to perform. And, it is something that AI and ML can perform on large data sets with high efficiency.
But it should be noted here that AI and ML have been used for trading for decades when it comes to executing fiat trades. And, for cryptocurrency it happened almost the second they got to the exchanges.
With this in mind, here are three ways these platforms are trying to solve the crypto market conundrum.
One of the problems with a highly volatile asset like crypto is its lack of liquidity. Many merchants are still on the fence regarding supporting cryptocurrencies as a payment method. Dealing with a volatile asset makes it difficult for them to manage shifting prices and cash flows.
And, this problem carries over to other activities like trading and investing. Liquidity providers, who can earn based on spread and commissions, and exchanges often charge premium prices and fees in order to compensate for liquidity gaps. A glance at broker spreads can indeed be highly illuminating in the FX space.
Siim Õunap, an FX and cryptocurrency trader, commenting on the level of fees charged observed that: “The biggest difference in fees can be seen with bank related transactions. For banks, crypto-to-fiat bank transfers cost them more than fiat-to-fiat transfers. Coinsbank, a crypto exchange, is an example of a major provider of crypto-to-fiat transactions and starts from $50 for a transfer.”
The Estonian added: “Fiat currency exchanges, however, take much higher fees than crypto exchanges for these same services.”
So are there answers out there to address this issue? Well, upcoming ThinkCoin-powered trading platform TradeConnect uses blockchain to facilitate the trade of a variety of assets – both traditional and crypto. In order to address the pricing problem, the platform uses a persona matching protocol to help find the optimal pair of financial products and traders.
In relation to the persona matching protocol, TradeConnect users each receive a digital persona and the system looks for an ideal broker or bank offering a product that fits the persona’s trading activities.
To facilitate this, TradeConnect uses ML to power with what is touted as a unique scoring system, which helps build a trader’s profile based on the person’s activity. Based on this persona, makers and takers will be able to bid for the ideal trader.
By doing so, liquidity providers on the platform will be able to optimize their activities and be matched with their ideal trades. This helps them reduce costs and, in turn, enable them to price competitively.
Another way AI and ML are influencing crypto trading is through sentiments. Sentiment analysis involves crunching large volumes of textual data from sources such as articles, blogs, comments, social media posts and even video transcripts in order to determine a market’s take on a particular topic – whether it be positive, neutral or negative. Such analysis enables marketers, for instance, to see how the market views a particular product.
Blockchain-based platform Senno, for example, aims to make sentiment analysis accessible to a wider audience. Typically sentiment analysis utilizes large amounts of computing resources to perform, but with the use of a distributed network, AI and ML, Senno is able to put up these resources to perform analyses.
The platform also allows developers to build applications, which can access sentiments through a programming interface and a development kit.
Senno’s BI (Business Intelligence) capabilities are claimed to be extremely useful in the financial markets, with part of the core algorithm aimed at making cryptocurrency trading less daunting. And, since the market is highly speculative, clearly knowing what currencies excite or turn off the market could be useful to traders.
Towards fulfilling these endeavors, Senno has partnered with CryptoScanner, an application that scans various cryptocurrency channels and creates reports and alerts, showing how the market’s views individual coins. Traders can then use the insight to guide their strategies.
Forecasts & Predictions
A third way these platforms try to understand the crypto market is through forecasts. Take a pioneering platform like Augur, which has effectively shown that making predictions – available through blockchain – is feasible. Augur’s predictions, however, draws from peer-to-peer contributions.
Other platforms like NeuroBot, on the other hand, use a neural network that makes it speedier for predictions related to the crypto market. NeuroBot is designed to perpetually crunch data of cryptocurrency exchange rates. These are then used to create price forecasts for the following days, which the platform claims to be between “70% to 90%” accurate.
NeuroBot also takes into consideration essential technical analysis like patterns and signal indicators. The project aims to include other aspects of technical analysis such as Fibonacci retracement, a popular tool among technical traders that is based on the key numbers identified by mathematician Leonardo Fibonacci in the 13th century, the Elliott wave principle and even fundamental analysis.
Fibonacci retracement is created by taking two extreme points (usually a major peak and trough) on a stock or equity chart. The Elliot wave analyzes financial market cycles and forecasts market trends by identifying extremes in investor psychology, price highs and lows, and other factors.
Through neural networks, the NeuroBot platform continuously improves in order to provide enhanced accuracy.
Given that it can take time for human investors to learn the more technical side of trading, many new traders can find prediction platforms useful. They are a user-friendly means to gain the benefits of technical analysis without having to fuss over the technical details.
Bound to Get Better?
These platforms offer interesting and exciting developments for the crypto market. And, it is argued by industry protagonists that if these projects are able to minimize or at least lessen crypto’s volatility, then the space would surely benefit from the wider adoption that price stability brings. But there is no 100% guarantee, even if the aim might be grand.
The main reason for the high volatility is market manipulation, which is often caused by AI and made possible due to small trading volumes. The same can be witnessed in the foreign exchange markets between larger and smaller currencies that are not linked to a major currency, such as those in the G8 currencies including the US dollar, Euro, Sterling, Swiss Franc and Yen.
Crypto trader Õunap, who divides his time between London and Tallinn and is also COO of the Savii Digital blockchain agency, added: “AI solutions and ML programs are often used to manipulate the price of currencies, both crypto and fiat, thus leading to higher volatility. The only feasible solution to reducing the high volatility, other than regulating all currencies, is to increase the trading volume to a point where even a $500 million buy-in or sell has little to no effect.”
The upside though from AI and ML, traders now have access to insights that are distilled from vast amounts of information that a single human mind cannot simply process. And, even if they could the chance to act would probably be gone by the time a trading decision had been made.
As such these platforms are enabling even the ordinary investor access to these powerful tools. Also given that AI and ML are continuously improving, it is fair to say that one can only expect these platforms to perform much better in the course of time.
That said, investors should still temper their expectations and not exhibit blind faith in relation to these platforms. Even if they are able to perform with great accuracy, there are still outliers that they might not be able to take into account.
At the end of the day investors have to keep in mind that these are tools to guide decision making. But only time will tell if these platforms will really be able to crack the crypto market and prove a real game changer.