RE:CZ

It is now February 10, 2026, in the afternoon.

Today, I shared my latest experimental results with Ryan and QY. I originally intended to invite Mage as well, but Mage was out today and couldn't attend. I didn't want to schedule another separate meeting, so I'll save it for next time when there are more results to share.

Next, I envisioned a Three-Body Dynamics version of a signal gating mechanism. The key lies in estimating the three core market state variables: $\delta$, $\mu$, and $\sigma$. Assuming we have these estimates, we can use a gating mechanism to decide under which market conditions to bet on a momentum strategy and under which conditions to withdraw from it.

Taking the momentum strategy as an example: maintain minimal investment during the overcooled period, increase investment during the momentum-benefiting period, and withdraw from the momentum strategy before the momentum-harming period (i.e., during the value-benefiting period, market-making-benefiting period, and overheated period).

In fact, any strategy X (where X = momentum, value, market-making) follows this pattern:

1. During the overcooled period, maintain minimal investment.
2. During the strategy X benefiting period, increase investment.
3. Withdraw from strategy X before its harming period (during the overheated period or other strategies' benefiting periods).
4. Wait for the market to return to the overcooled period or the strategy X benefiting period, then repeat the above process.

This is a universal strategy gating mechanism.

Of course, the prerequisite is that we can accurately estimate the market state variables $\delta$, $\mu$, and $\sigma$. This requires further research and experimentation.

Intuitive Understanding of Market State Variables

First, let's talk about my intuitive understanding of these three market state variables:

The Core of Premium $\delta$ is the Psychological Anchoring Effect

The premium is the degree of deviation between the current price and the intrinsic value. The core of its study lies in determining the intrinsic value. The determination of intrinsic value should start from the psychological anchoring effect.

Volume Profile: By analyzing the volume distribution across different price ranges, we can infer the psychological anchor points of market participants. Price ranges with concentrated volume typically represent consensus prices among market participants and can serve as reference points for intrinsic value.

It implies one thing: trading volume occurs only when buyers and sellers reach a price consensus; a lack of consensus leads to shrinking volume. Higher trading volume means greater acceptance of that price by both buyers and sellers, and thus a stronger psychological anchoring effect. To find the anchor point for intrinsic value, one must identify price ranges with concentrated volume.

Multi-Anchor Point System

In most cases, there may not be just one psychological anchor point, but multiple anchor points acting together. Each psychological anchor point can influence the market price, forming a complex psychological anchoring network. The interactions between these anchor points may lead to price fluctuations and trend formation.

Lifecycle of Psychological Anchor Points

Psychological anchor points emerge and become invalid during market evolution. For example, after a new All-Time High (ATH) appears, the market may form a new psychological anchor point after a pullback and consolidation. For instance, when the market rapidly breaks through a previous anchor point, it means that anchor point has become invalid and can no longer attract value capital to intervene at that level; value capital will use a lower anchor point to intervene.

Note: Long-term moving averages are generally considered to represent long-term average cost prices. Whether they can serve as a simplified model for the psychological anchoring effect remains unknown. The correlation between cost price and anchor price is somewhat linked to the phenomenon of active trading around cost prices. My intuition is that using long moving averages as a simplified model for anchor pricing has some merit but is still one-sided; the truth may be more complex.

The Core of Momentum $\mu$ is the Rate of Price Change

Momentum is the speed and direction of price change. The core of its study lies in the speed and direction of price change. The strength of momentum can be measured by the rate of price change: faster price changes indicate stronger momentum; slower price changes indicate weaker momentum.

Clearly, it's well-defined. We can use the moving average of the logarithmic returns of prices to measure the strength of momentum.

Other similar momentum indicators include using the difference between fast and slow moving averages of prices, constructing multi-period, multi-moving average systems, or constructing multi-level difference systems to measure momentum strength. These methods are linearly related; they all measure momentum strength on different time scales. A typical example is the MACD indicator, which measures momentum strength through multi-layered moving average differences. Despite multiple rounds of differentiation and smoothing, we still believe MACD can only measure momentum and cannot address premium and volatility issues.

The Core of Volatility $\sigma$ is the Magnitude of Price Change

Volatility is the magnitude of price change. The core of its study lies in the magnitude of price change. The level of volatility can be measured by the standard deviation of logarithmic returns, which is a classic method for measuring volatility.

This method is quite mature and widely used in options pricing. It has a very broad influence. Therefore, I believe this formula is best for volatility. The more people who use it, the more effective it becomes.

There is almost no controversy about this.

How to Judge the Effectiveness of Estimation Methods?

The more effective the estimation, the better the gating effect, and the greater the improvement in the signal strategy's return.

Therefore, we can judge the effectiveness of estimation methods by the quality of the gating effect. We can design an experiment to compare the performance of different estimation methods within the gating mechanism, thereby determining which estimation method is more effective.

One More Thing

To be honest, many signal strategies themselves are estimating these three market state variables: $\delta$, $\mu$, and $\sigma$. For example, momentum strategies inherently estimate the strength of momentum $\mu$, value strategies inherently estimate the degree of premium $\delta$, and market-making strategies inherently estimate the level of volatility $\sigma$.

Some more advanced signal strategies (often what quantitative trading develops) inherently include estimations of other market state variables. For instance, momentum strategies often incorporate premium filtering conditions. Remember the common divergence exit mechanism used in moving average strategies? That's a judgment on premium, a prediction of value capital behavior!

Therefore, the development of signal strategies naturally incorporates the estimation of market state variables. So, an advanced signal strategy already contains the gating system I mentioned. However, the gating systems within signal strategies are often unsystematic; practitioners often know the 'how' but not the 'why'. It's still necessary to understand the Three-Body Dynamics of Capital Markets within the entire systematic framework.

So, if signal gating is decoupled, then $\delta$, $\mu$, and $\sigma$ are essentially just factors for signal strategies. Therefore, the Three-Body Dynamics essentially states that these factors can play a very decisive role.

Then, what remains for signal strategies? Could it be that only the simplest signal strategies are sufficient?

The finest ingredients often require the simplest cooking methods. — A Bite of China