The trading room atop Shanghai Tower felt like a precision organism sensing the global economic pulse. Over the past few weeks, the air of highly restrained defensive tension brought by potential sniping had been replaced by a new, more proactive and sharp energy field. The celestial configuration of the global macroeconomy had shifted; major central banks, fighting persistently high inflationary pressure, had—as if by agreement—unveiled the prelude to a **rate‑hiking cycle**. The compass of monetary policy had decisively pivoted from the extreme easing and liquidity glut of the pandemic era toward tightening and withdrawal.
This macro‑scale hurricane swept through every corner of global capital, bringing a fundamental change in market ecology. The earlier complex landscape—universal asset‑price prosperity driven by abundant funds, rapid sector rotation, frequent switching between "choppiness" and "weak trend"—was being supplanted by a simpler, crueler paradigm: **strong trendiness**. Capital began to flow like seawater after ebb tide, obeying clearer gravitational pulls (interest rates, growth‑expectation differentials), moving massively and persistently. The divergence between safe‑haven assets and risk assets intensified; the strength relationships among currencies grew distinct and enduring, driven by interest‑rate spreads.
For Mozi's "adaptive dual‑core model," this was nothing less than a radical transformation of the battlefield. The "choppiness model," which relied on capturing tiny opportunities within the market's complex noise and microstructure, switching states at high frequency via "verifiable confidence," was rapidly losing its turf. Market "noise" was being drowned by powerful trend "signals"; attempting counter‑trend choppiness operations within a trend was like a mantis trying to stop a chariot.
It was time for the long‑dormant **"trend model"** to reveal its cutting edge.
Mozi calmly initiated the strategy‑switch protocol. The model's meta‑logic core formally transferred the primary capital weighting and decision‑making authority to the "trend model." This meant the model would temporarily abandon its excessive focus on the market's complex microstructure, instead dedicating its main effort to identifying, confirming, and following those price‑movement directions that were driven by macro forces, sustained, and large‑amplitude.
The core of the "trend model" lay in capturing the market's **"momentum"**—the inertial force of price continuing to move in a certain direction. It did not attempt to predict the trend's start or end point; rather, it intervened promptly after a trend had initially formed and demonstrated a certain persistence, hitching a ride on that momentum until signals of momentum exhaustion appeared.
To quantify this momentum, the model relied on a series of time‑tested technical indicators, the most central being the **Relative Strength Index (RSI)**.
Mozi pulled up the RSI algorithm's core for review. The essence of RSI was to measure, within a set period (usually 14 cycles), the relative ratio between the total magnitude of price rises and the total magnitude of price declines, thereby assessing the intrinsic strength of market momentum. Its calculation was rigorous and elegant:
**Compute price changes within the period:** Calculate each cycle's (e.g., daily) closing‑price gain (U) and loss (D). If today's close is higher than yesterday's, U = today's close − yesterday's close, D = 0; vice versa.
2. **Compute average gain and average loss:** Average all U values over the past N cycles (default 14) to obtain the average gain (AU); average all D values to obtain the average loss (AD). Early versions used simple arithmetic averages; modern algorithms mostly employ exponential moving averages (EMA) to give more weight to recent price changes, making the indicator more responsive.
3. **Compute relative strength (RS):** RS = AU / AD. This value represented the relative strength ratio between bullish and bearish forces within that timeframe.
4. **Compute RSI value:** RSI = 100 − [100 / (1 + RS)].
Ultimately, the RSI value was normalized within the 0‑100 range.
How did the model interpret and use this seemingly simple number?
* **Momentum identification:** When RSI persistently operated above the 50 midline and kept rising, it indicated that average gains significantly exceeded average losses—the market was in **upward momentum**. Conversely, when RSI stayed below 50 and kept falling, **downward momentum** dominated. The trend model closely monitored RSI's relationship with the 50 midline and its direction, using it as an important clue for judging trend direction.
* **Overbought and oversold conditions:** This was RSI's most widely known application. Typically, when RSI entered the **above‑70** zone, the market was considered "**overbought**," meaning price had risen too much, too fast within the current period, possibly facing short‑term correction pressure from profit‑taking. Conversely, when RSI entered the **below‑30** zone, the market was considered "**oversold**," meaning downward momentum might be short‑term exhausted, with potential for a technical rebound. The trend model treated these zones as potential trend‑**exhaustion warning signals**, but not as immediate reversal orders—in strong trends, RSI could linger in overbought or oversold territory for extended periods, itself a manifestation of trend strength.
Yet isolated RSI signals could be misleading. To enhance trend‑judgment reliability, the model introduced the **Moving Average Convergence Divergence (MACD)** as a crucial auxiliary and confirmation indicator.
Mozi examined MACD's composition. It comprised three parts:
**DIF line (fast line):** The difference between a short‑term (usually 12‑day) exponential moving average (EMA) and a long‑term (usually 26‑day) EMA. It sensitively reflected changes in short‑term price momentum relative to the long‑term trend.
2. **DEA line (slow line/signal line):** The M‑day (usually 9‑day) exponential moving average of the DIF line. It smoothed DIF, filtering out some noise and generating trading signals.
3. **MACD histogram:** The difference between the DIF line and the DEA line, displayed as bars whose height and sign intuitively reflected the divergence between the two lines, representing the state of momentum acceleration or deceleration.
The model's MACD utilization strategy was as follows:
* **Trend‑direction confirmation:** When both the DIF line and DEA line were above the zero axis and maintained an upward posture, the market was confirmed to be in a **bull trend**. When both were below the zero axis and kept declining, a **bear trend** was confirmed.
* **Trading signals:** When the DIF line crossed above the DEA line (forming a "golden cross"), especially near or above the zero axis, it was regarded as a positive **buy signal**. When the DIF line crossed below the DEA line (forming a "death cross"), especially near or below the zero axis, it was considered a **sell signal**.
* **Momentum divergence:** This was one of MACD's most powerful applications. When price made a new high, but MACD's DIF line or histogram failed to make a new high simultaneously (top divergence), it often foreshadowed exhaustion of upward momentum, serving as a potential trend‑reversal warning. Conversely, price making a new low while MACD did not (bottom divergence) might signal exhaustion of downward momentum.
At this moment, the model's trend‑recognition module was emitting clear signals. Against the backdrop of the global rate‑hiking cycle commencing and geopolitical risks persisting, gold—as a traditional safe‑haven and store‑of‑value instrument—after a period of consolidation, began to display robust upward momentum.
On the screen, the gold‑futures price curve rose steadily. Its corresponding RSI indicator, after briefly touching the 50 midline, climbed all the way up, stabilizing in the strong zone above 60; despite fluctuations, it never immediately reversed after entering overbought territory, revealing healthy upward momentum. Meanwhile, MACD's fast and slow lines, having formed a golden cross above the zero axis, continued diverging upward; the MACD histogram lengthened steadily, exhibiting the classic **bull‑trend acceleration** pattern.
Based on RSI's momentum confirmation and MACD's trend‑reinforcement signals, the "trend model"—in conjunction with its internal position‑management algorithm—began gradually but resolutely increasing gold‑futures long positions. Unlike the choppiness model's frequent in‑and‑out, it acted like a patient hunter: once confirming the prey's direction and trajectory, it steadily shouldered the rifle, following its movement.
As if validating the model's judgment, gold prices, resonating with both macro logic and technicals, launched a smooth, powerful uptrend. Prices consecutively broke through key resistance levels, pullbacks were shallow and fleeting, then attacks upward resumed with even fiercer force. The model's gold‑long position floating profits expanded continuously, like a rolling snowball, pushing the entire fund's equity curve to ascend at a breathtaking steep slope, head high, upward, ever upward!
The strength of capital, within this torrent of trend, grew visibly and rapidly, reaching an unprecedented new height. Mozi's financial empire, in this macro‑bestowed opportunity, became larger, more bloated, and… more conspicuous.
Yet behind this glorious victory, no trace of joy could be seen in Mozi's deep eyes; instead, a flash of vigilance graver than before passed through. He clearly observed that, within gold market's trend, aside from the mainstream capital his model followed, those previously faintly visible, anonymous, coordinated counterparty positions had not been forced back or swallowed by this strong rally; rather, as if using the trend as cover, they continued their layout around his core holdings and related derivative markets—more covertly, more patiently.
The cutting edge of trends had brought him enormous gains, but also, like a searchlight, exposed him and his massive capital bulk more clearly in the crosshairs of hidden snipers. He knew this silent war was far from over; indeed, because of this great triumph, it might even advance prematurely into the next, more brutal phase. He had enjoyed the trend's gift; now he must prepare to meet the even fiercer counter‑attack that would follow. Beneath the feast of capital, the undercurrents grew ever more turbulent.