data jungle adventures

I am under the impression half the Medium, Reddit and Twitter crowd is asking this? Also, many software, ML-driven trading bots ask for fees that fall within 5% of the client’s portfolio.

I will not go into the projects and compare outcomes and costs/fees here but just clarify a few relevant basics for decision-making.

• Fundamentals: The technology behind a project/startup or coin is more than relevant to its success. Also, we have seen many non-innovative meme coins ‘mooning’. Knowledge of the use case, the technology used, and token economics (i.e. coin releases, burning, total supply, airdrop practices, …) is essential to evaluate a project’s ‘hard’ value, as every investor should do. Reading up on tech will also remove some emotions from a purchasing decision. This is more relevant for sustainable long-term choices.
• Technical Analysis: Different predictors (stats) on the never halting 24/7 price chart. You can apply and draw whatever you want in Tradingview or elsewhere. There is some consensus on the indexes used (Moving Average Convergence Divergence (MACD), Relative Strength Index (RSI), Bollinger Bands, and On Balance Volume (OBV), and sound training material on the underlying stats, caveats and inferences are available on various platforms. While this knowledge is transferable to stock and other time series data sets (climate, Twitter, etc.), its predictive success appears higher for short time intervals.
• Macroeconomy and current affairs: The economics and political decisions of large populations influence the perception of the current economic model, economic and banking mechanisms (interest rates, commodity scarcity, human stressors, …) and will influence emotional and rational decisions (the great hash rate migration, Brexit, Trump entering office, …).
• Community: I want the community and emotional element as a fourth item. As a decent section of our lives have moved online, and we continue to express ourselves rationally and emotionally in virtual environments, considering collective, virtually expressed emotions are also a valuable data point (#cryptotwitter). New #emojargon has been created as abbreviations to describe collective mindsets (#FOMO, #FUD).

Are AI-based TA trading decisions successful?

I have played around with free and paid solutions that fit models to price time series on multiple coins (#MACD, #BSI) and used deep learning to predict paper and actual trading (ouch). Of course, some configurations are more successful than others. Still, until you incorporate all four topics in your decisions, I cannot see how you can predict price movement reliably short or long term.

I still have a #Gekko bot paper trader publishing the results of a #neuralnet to Twitter.

I am only aware of a single project that connects TA with human-sourced price prediction in an AI environment. I have not used their hybrid-intelligence service (https://cindicator.com/). It attempts to include macroeconomic data as the survey questions answered by humans include price predictions and information on purchasing and employment situations. Suppose such an approach receives more attention (human predictions), connects to major news outlets, and scans social media data for macroeconomic, political, and emotional statements. In that case, price predictions will obtain more substantial support.