Time and Celestial Cycles

Time Cycles are often considered taboo or pseudoscientific, especially in fields such as finance, statistics, and engineering. I do not promote traditional theories or mystic theories, and take a quantitative approach to both time and celestial cycles. And I don't claim deep expertise in those classical pseudoscientific techniques. Here I bring a skeptical yet curious mindset shaped by years of experience in computer science, algorithmic trading, and data analysis.

Over time, I’ve explored select features inspired by time cycles, celestial movements, and other theories and used them as inputs to probabilistic and machine learning models. Surprisingly, some of these experiments have led to meaningful predictions, not just in market behavior but also aligning with significant real-world events like economic events, geopolitical shifts, wars, or natural disasters.

These patterns were compelling enough to warrant deeper exploration. After all, markets respond to more than just charts—they react to the world.

If you’re here expecting mysticism, you won’t find it.

Any openness to explore the unconventional realm is an hypothesis to be backed by data, not by belief alone. I've also found that these things may not have a deterministic relationship, but a probabilistic relationship. I explore these realms, so long as it's approached with transparency, critical thinking, and a commitment to data. The work here is ongoing and evolving; more evidence will be shared as the journey unfolds.

Early dismissals

I used to dismiss these theories outright, until I stumbled upon a correlation that changed everything. I had to investigate. That experience sent me down a rabbit hole I couldn't ignore. Over time, I also found that forecasts do not always work. These are forecasts and not predictions. The forecasts are probabilistic, not deterministic—just like any model. That variability is part of what makes this exploration so fascinating.

Many people fall into two camps: the scientific and quantitative thinkers, and the mystical and belief-driven ones. I belong to the first group—but I can't ignore that some things from the other side seem to work. Science, at its core, demands open-mindedness and rigor. Yet I often find the so-called "pro-science" crowd becoming just as dogmatic and closed as the belief-driven side. On the other hand, those who rely on mysticism usually have little else to hold on to; therefore, it is pretty challenging to get them to see the scientific perspective when they take invalidation or questioning as a personal attack. People like me—who venture into that space without abandoning logic—walk a fine line. It's easy to get lost in a never-ending abyss.

Correlation is not causation.

Not claiming to know the "why" behind these patterns, nor do I have the resources to investigate their root causes. That's not the goal.

Why call it "Time and Celestial Cycles"?

I initially used the term “quantitative astrology.” However, the word “astrology” often rubs people the wrong way—sometimes even more than the actual ideas behind it. Even worse, it attracts those who seek belief-based validation rather than data-driven inquiry.

I don't mindlessly use conventional mystical methods, but I use a subset of fundamental features. To understand these features, I have to explore classical pseudo-scientific techniques, attempting to understand the underlying "why this feature is important". And later apply math and logic to see if it does appear in the real world. And I am often surprised. These features are mixed, matched, and embedded into probabilistic models and machine learning systems to test for real-world relevance. That's why I prefer the broader and more neutral term "time cycles" or just "cycles"—it reflects the essence without the baggage.

Validate everything

Whether it's a technical setup, such as a 3-bar reversal trade, or a celestial event, both are qualitative ideas until they are put to the test. Here, we treat every pattern—market-based, celestial, time-cycle, or something else—as a hypothesis to be validated with data.