On Market Forecasting
Let’s talk about forecasting today. What exactly is forecasting? What kind of trickery does ordinary forecasting amount to? And what does a scientifically rigorous forecast truly entail? A genuine forecast means “forecasting without forecasting.” Of course, this is not the same concept as conventional forecasting. Under the deception and misleading influence of the conventional notion of forecasting, many people can’t help but feel the urge to make forecasts from time to time.
All trends in the market are formed by the combined forces at the current moment. Generally, since the component forces of policies or rules remain constant for at least a certain period, most people tend to forget or overlook their existence. However, whether these component forces are constant or variable (changing with each transaction), the combined force is always formed in the present. A constant component force, denoted as F(t), merely indicates that its value is either a constant or a piecewise constant. For any specific moment t, there is no difference between constant and variable components in terms of the rules of combination or the resulting combined force.
Yet, these constant component forces are not eternally constant; they are often piecewise, with breakpoints in their changes. Many fundamental component forces share this characteristic, and these breakpoints create blind spots in forecasting. Admittedly, conducting fundamental analysis and comprehensively examining macroeconomic factors can help minimize these blind spots, but it is impossible to eliminate them entirely. The existence of such factors already turns all so-called “accurate forecasts” in the ordinary sense into a joke.
More importantly, fundamental factors themselves are the product of combined forces. Isn’t politics, economics, and other fields all the result of combined forces? The current global political and economic structure is the outcome of numerous combined forces, and this is even more true within a single country. Many people think rigidly, always assuming that policies are like a “god” — not shaped by combined forces, free from struggles between various interests, and all outcomes are delivered like products from a pre-set machine. In essence, all theories of “accurate forecasting” in the ordinary sense are based on such rigid thinking.
On a deeper, philosophical level, forecasting itself is a component force. Just as an observer is inherently part of the observation process — all observation results are related to the observer, influenced by the observer, and predicated on the observer — forecasting intervenes in the outcome it intends to predict in the same way. Similar to the uncertainty principle in quantum mechanics, the most fundamental principle of any forecasting theory is that “accurate prediction is impossible.”
Some might argue, “Many people have experiences of making accurate forecasts — why is that?” In fact, this is merely a matter of probability. The possible trends of the market, when categorized by any standard, are limited in number; typically, there are only three or four scenarios. Meanwhile, the number of people who enjoy the “forecasting game” and boast about their “accurate predictions” is larger than the number of people involved in illicit relationships worldwide. Even a blind cat can catch a dead mouse occasionally. Even if someone makes correct predictions consecutively, it still falls within the realm of probability — there is nothing remarkable about it. All those who claim to be “accurate forecasters” are either playing this trick or being fooled by it without realizing it. As for those who hide their wrong predictions and only flaunt the few correct ones, they are even more contemptible.
In reality, forecasting is not mysterious at all. The foundation of all forecasting lies in classification — fully categorizing all possible scenarios.
Some might say, “After classification, we eliminate the impossible ones, and the last remaining result is the accurate forecast.” This is a nonsensical idea from a muddled mind. Every elimination is equivalent to a forecast; each time a category is ruled out, according to the multiplication rule of probability, the so-called “accuracy” of the final result becomes less precise. In the end, one still cannot escape the constraints of probability.
The only correct principle for classifying forecasts is to avoid any elimination; instead, one must clearly define the boundary conditions for each scenario. In fact, any classification is equivalent to a piecewise function — the key is to identify the boundary conditions of this function clearly. Take the following function as an example:
f(X) = -1, when X ∈ (-∞, 0); f(X) = 0, when X = 0; f(X) = 1, when X ∈ (0, ∞)
The crucial part is to clarify the range of X for which f(X) takes a specific value — this range is the boundary condition. In the classification of market trends, the only certainty is that the value cannot be negative, so we classify within the range [0, ∞), dividing this interval into N boundary conditions based on a specific classification principle.
Some might ask, “How can a stock’s value drop to zero?” Is there anything strange about that? What about delisted stocks? Not to mention stocks — even money can become worthless. How much was a gold yuan note (a former Chinese currency) worth in 1950? Of course, if your descendants could keep a gold yuan note until the final moment of the universe’s collapse, its value might reach an astronomically terrifying number N. But then, you’d have to wait that long.
Not only are stocks essentially worthless paper, but currency itself is also worthless paper in nature. Their so-called “value range” is the same as that of stocks, and zero is a possible value. From the perspective of the most precise theory, they can even take negative values. For instance, if a certain dynasty or country’s government decreed the death penalty for anyone hoarding the currency or stocks of a previous dynasty or another country, wouldn’t that currency or stock be “negative in value”? As for specific cases where a stock’s value drops to zero, it happens frequently with warrants.
After defining the segmented boundary conditions, the next step is to determine how to act once a particular scenario occurs — in other words, segmenting the operation process as well. Then, we leave all scenarios to the market itself, allowing the market to make its choice in the present moment. For example, a few days ago, I classified the trend range based on the previous two high points and the 10-day moving average, which naturally divided the trend into two scenarios: breaking below the range or not breaking below it. I then pre-determined what to do in each case — that’s all there is to it. This is the most essential form of forecasting: “forecasting without forecasting,” letting the market make its own choice. In the end, the market chose not to break below the range, so we continued to hold our positions.
Some might say, “What if it rises first and then breaks below later?” This is a typical example of muddled, over-forecasting thinking. Any market operator must not get trapped in such meaningless speculation. The market’s failure to break below the range is an established fact, and one’s operations can only be based on facts that have already occurred. If it does break below later, we will address it when that becomes a fact. Therefore, in the context of a truly accurate forecast, you have already pre-defined what to do if a breakdown occurs. Since that scenario has not become a reality, the other scenario (not breaking below) has — so you simply act accordingly.
Generally speaking, people who enjoy forecasting are usually oversensitive, muddle-headed, incompetent in actual operations, and fond of deception. Those who have been predicting a market peak since the 2000-point level — if they had to cut a piece of flesh for every wrong prediction — would now be nothing but a pile of bones (like fake lamb spine hot pot ingredients). Investing is not just about empty talk; it requires practical operations. All operations are essentially responses to the results of different segmented boundaries — the only difference lies in the segmented boundaries defined by each individual.
Therefore, the key is not to predict anything, but to define the segmented boundary conditions. Once you have these principles for segmentation, you just follow them in your operations. What need is there for forecasting? And what is there to forecast anyway?
The history of global financial markets has consistently proven that truly successful operators never make forecasts. Even if they engage in some rhetoric in the media, it is only to leverage the media for their purposes. Genuine operators all have a set of operational principles, and adhering to these principles is the best form of forecasting.
Providing a piecewise function is equivalent to providing the most accurate forecast. All forecasts are made in the present moment — this is the true essence of forecasting. Such forecasting requires no meaningless probabilistic calculations, nor does it involve the deception or excitement of “successful predictions.” The “success” of this kind of forecasting occurs at every moment; if someone gets excited or boasts about it every single time, their mind must be thoroughly muddled. As the saying goes: “Wild geese passing through a clear sky, the wind swirling over green waters — which of these reflects your true nature?”
No comments:
Post a Comment