We maintain our previous key judgment that the market has established a "bull market mindset." Once a trend is formed, it is difficult to reverse in the short term. We will not easily use the empirical rules of oscillating or bear markets as signals, and we will stick to industrial mainlines.
**Strategy 1: Calendar Effects of Low-Valuation Blue Chips, But Certain Conditions Must Be Met**
Entering the fourth quarter, there is an important characteristic in market sector structure: Since 2005, pro-cyclical industries have had an upward probability of over 65% in Q4, with over 60% probability of outperforming the CSI 300.
The "calendar effect" of pro-cyclical industries in Q4 is based on expectations of improving macroeconomic fundamentals. Historically, such judgments have come from two scenarios:
Scenario 1: Relying on PPI improvement expectations. Years with PPI improvement can be divided into two types: ① Natural economic cycles, such as Q4 2009/2013/2019 anticipating inventory cycle initiation ② Economic stimulus policies (supply-side or demand-side), such as Q4 2014 shantytown renovation targets, Q4 2015 supply-side reform
Scenario 2: Relying on grand narratives, such as 2014's Belt and Road Initiative and state-owned enterprise reform.
From these two perspectives, can pro-cyclical calendar effects appear in 2025?
First, the transmission path of "inventory cycle → economic cycle → PPI expectations" appears unlikely at present;
Second, according to year-end important meeting timelines, supply-side policies focus on substantial progress in "anti-involution," while demand-side policies track policy details at the end of November;
Finally, regarding grand narratives, attention should be paid to subsequent developments from the Fourth Plenum and the 15th Five-Year Plan.
**Strategy 2: Pre-positioning High-Growth Sectors During Earnings Vacuum Period**
1. In years when macroeconomic fundamentals lack highlights but industrial tracks are active, Q4 stock prices have stronger guidance for the following year. For example, 2013-2015, 2019-2021, and 2024. In these years, Q4 leading sectors have a high probability of continuing through the next year's market.
2. How to screen year-end growth tracks?
(1) "Moving Average Deviation" indicator screening results: Currently, the mainline trends of [Optical Modules, PCB, Innovative Drugs, Sci-Tech Chips, Non-ferrous Metals] mostly remain healthy, with only innovative drugs falling into sideways consolidation. Based on patterns of industrial mainlines since 2012, falling into consolidation does not represent the end of trends. As long as industrial trends persist, mainlines still have opportunities to reach new highs after sideways digestion.
(2) "Call Option Attributes" screening results: Besides TMT sectors that have accumulated significant gains, sectors with high overlap include [Auto Parts/Robotics recent lagging segments, Power Grid Equipment, Consumer Electronics], etc. These sectors with call option attributes have generally lagged recently, with relatively controllable downward adjustment ranges and sensitive upward responses to positive news, making them relatively suitable for medium-term forward positioning.
**Main Report**
**Part I: What Are the Historical Patterns of Q4 Markets? When Do Calendar Effects Fail?**
1. Empirically, stable low-valuation and early-cycle sectors perform better in Q4
Statistical review of markets since 2005 shows that some pro-cyclical sectors exhibit relatively obvious calendar effects in Q4. According to SWICS Level 2 classification, only 12 industries since 2005 have had Q4 upward probability exceeding 65% and probability of outperforming CSI 300 above 60%: Kitchen & Bath Appliances, Rail Transit Equipment, Other Power Equipment, White Goods, Food Processing, General Equipment, Automation Equipment, Agricultural Product Processing, Wind Power Equipment, Construction Machinery, Special Equipment, and Leisure Food. These industries have slightly lower earnings volatility compared to other cyclical sectors but still rely on relatively stable macroeconomic expectations.
2. The necessary condition for empirical patterns is relatively stable macroeconomic expectations
Q4 calendar effects stem from two logics: First, after Q3 reports, markets enter a relatively long earnings vacuum period. Therefore, after Q3 earnings are realized, year-end valuation switching markets more easily tilt toward sectors with relatively stable growth rates and ROE, or low-valuation sectors. Second, year-end intensive economic work meetings boost policy expectations, while early the following year is generally the policy implementation phase. Historical experience shows cross-year liquidity environments are generally ample, all favoring low-valuation, pro-cyclical sectors.
Correspondingly, to form relatively stable or optimistic macroeconomic expectations, there are mainly two situations:
(1) Stable recovery from natural economic cycles, typical years being 2009, 2016, 2019, 2020. From inventory cycle positions, these years all showed different degrees of economic prosperity recovery signs at year-end, mostly in the final stages of passive destocking or just starting active restocking, with upward expectations for industrial product prices and improvement expectations for industrial enterprise profits.
(2) Fundamentals have not yet bottomed out, but there are major macro/policy narratives that can support next year's expectations, typical years being 2013 (demand-side - SOE reform and Belt and Road proposal), 2014 (demand-side - reform acceleration, clear shantytown renovation targets), 2015 (supply-side reform). Such major narratives, whether focused on supply-side or demand-side, ultimately point to PPI increases and corporate profit improvement.
Conversely, if natural economic cycle downward pressure has not been eliminated and there are no sufficiently grand policy narratives, Q4 pro-cyclical sector calendar effects tend to fail, typical years being 2011, 2017-2018, 2023.
3. How to view Q4 2025 calendar effects?
As mentioned above, there are two situations where Q4 low-valuation pro-cyclical calendar effects hold: Scenario 1: Natural economic cycle stabilization; Scenario 2: Natural economic cycle has not stabilized, but there are major macro narratives on demand-side or supply-side sufficient to support next year's expectations.
We examine these two situations separately:
(1) Scenario 1: Is a new round of restocking cycle likely to start?
Looking at the combination of industrial enterprise revenue and finished goods inventory, Chinese industrial enterprises had relatively smooth and elastic inventory cycles before 2021. The last inventory cycle peaked in mid-2021 and declined rapidly with the real estate macro cycle until destocking speed began to slow in mid-2023. Over the past two years, if strictly dividing by changes in industrial enterprise revenue and finished goods inventory growth rates, it can be seen as quickly completing another inventory cycle; but compared to previous rounds, this cycle had very small amplitudes whether restocking or destocking, profit recovery or decline. Looking at extended periods, this looks more like constant alternation between macro prosperity expectation stabilization and disappointment, with traditional cyclical industries far from escaping real estate downward impact.
From PPI performance and market forecast comparisons, actual values have been lower than market consensus forecasts most of the time since 2023, indicating most people underestimated the impact of this cyclical downturn on industrial product prices and profits beforehand. There were also expectations of increased stable growth policies in H2 this year, such as accelerated local bond issuance and expanded old renovation scale, but from actual investment growth rates and terminal demand data, there is currently no obvious improvement.
(2) Scenario 2: Can anti-involution recreate the supply-side reform narrative?
Supply contraction policies had two relatively typical episodes in the past 10 years:
2016-2017 Supply-side Reform: At that time, enterprise overcapacity and high leverage became increasingly severe. Policy first mentioned "supply-side reform" in November 2015, focusing on "capacity reduction" to eliminate backward capacity in 2016, and 2017 was the final year of the Ten-Point Air Plan, using "environmental production restrictions" to replace polluting capacity. The policy cycle spanned 2 years, ending in Q4 2017. During this period, broad fiscal expansion supporting shantytown monetization (PSL) drove demand-side resonance improvement in the real estate chain, with PMI and industrial added value continuously improving since February 2016 - therefore this round was ultimately supply contraction and demand release resonance, finally bringing PPI increases and corporate profit improvement.
2021 Energy consumption dual control and power/production restrictions under "dual carbon" goals: "Dual carbon" goals were one of eight key tasks in 2021. Starting with the steel industry at the beginning of the year, it later expanded to aluminum, cement, yellow phosphorus and other industries in H2, with related product prices rising rapidly until supply stabilization policies were introduced in September, temporarily ending supply contraction regulation. On the demand side, export growth rates declined in 2021, manufacturing PMI fell below the boom-bust line; additionally, implicit debt tightening and active real estate regulation led to overall real estate construction slowdown - therefore this round was supply-side and demand-side dual weakness, ultimately not bringing PPI increases or corporate profit improvement.
Regarding this round of anti-involution, we believe it is currently in a relatively contradictory and deadlocked state. On one hand, this round of "anti-involution" since being proposed at the end of 2023, from initial industry association voices and spontaneous production cuts in subdivided industries, to this year's Central Financial and Economic Commission direction-setting and clear industry policy introduction, the narrative level has obviously elevated, which is unquestionable.
But on the other hand, actual implementation of "anti-involution" faces stable growth and employment pressure, as well as supply-side contraction space issues. Looking at crude steel output, crude steel output in the first 8 months of this year was already the lowest since 2020, meaning there is no great pressure to achieve non-increase in production. From local and industry perspectives, motivation to intensify "anti-involution" may not be very strong.
Therefore, from existing policies and real performance, there may not yet be sufficient evidence to support pro-cyclical sector catch-up in Q4 this year. Of course, given intensive major policies and direction-setting meetings in Q4, the possibility of exceeding expected policy narratives cannot be ruled out. Key attention should be paid to October's Fourth Plenum and November-early December direction-setting for next year's economic work.
**Part II: Strategy 2 - Pre-positioning High-Growth Sectors During Earnings Vacuum Period**
1. In years when macroeconomic fundamentals lack highlights but industrial tracks are active, Q4 stock prices have stronger guidance for the following year
We conducted statistics on the relationship between Q4 leading industries and next year's industry performance:
(1) For Q4 markets since 2010, we separately counted the top 20 Level 2 industries leading in [Oct-Dec], [Nov-Dec], [Oct-Nov].
(2) For industries that led in the previous year's Q4, we separately counted next year's [Q1], [half-year], [full-year] upward probability and probability of outperforming CSI 300.
Statistical results show that Q4 positioning for next year's high growth often occurs in years when market style is track-oriented, such as 2013-2015, 2019-2021, 2024. Q4 leading sectors in these years have a high probability of continuing through the next year's market.
Of course, there are also cases of being proven "wrong" in hindsight. Looking back, the main reason for this situation is year-end expectations of high growth/prosperity reversal, but encountering policies or other force majeure in the following year, causing major turns in earnings expectations (typical examples being games and education that led in late 2019, consumer electronics that led in late 2021).
2. Does this year's Q4 meet the conditions? How to screen year-end growth tracks?
From the above, if there is (1) unclear macro prosperity expectations and insufficient policy narratives to reverse PPI trends; (2) simultaneously active industrial hotspots and higher market risk appetite and fund adequacy - then markets may be more inclined to position for next year's high growth rather than bet on low-valuation cyclical catch-up. We believe from existing information, this year's situation is relatively consistent with this.
Of course, with growth stocks having already risen significantly in Q3, the main task for the next phase is to screen and filter industrial mainlines. We screen from two perspectives: "moving average deviation" and "call option attributes":
(1) "Moving Average Deviation" indicator screening results:
We proposed the moving average deviation indicator in our weekly report to measure the strength of mainline market trends: Moving Average Deviation = ln(Close) - ln(ema20), with values approximately representing deviation percentages.
Reference thresholds for entry and exit: don't chase when deviation is too high (>15%) at entry, chase when deviation is moderate (5%-15%), no need to worry above moving average when stopping losses, suggested to hold firm when just breaking below moving average (deviation at -5%~0%), suggested to exit when significantly breaking below moving average (deviation <-5%).
Currently [Optical Modules, PCB, Innovative Drugs, Sci-Tech Chips, Non-ferrous Metals] have mean deviations of 3.1%, 0.1%, -3.6%, 10.2%, 3.4% respectively. Most mainline trends remain healthy, with only innovative drugs falling into sideways consolidation (10 or more days out of 20 closing below moving average). From patterns of industrial mainlines since 2012, falling into consolidation does not represent trend endings. As long as industrial trends persist, mainlines still have opportunities to reach new highs after sideways digestion.
(2) "Call Option Attributes" screening results:
For incremental funds newly entering the market and investors with limited holdings in mainline sectors, choosing "low-position call options" is also a viable approach, using previously introduced call option strategies for systematic efficient mining. In screening approach, we select 4 core stock selection indicators: ① Capacity indicator: average daily turnover of 3-10 billion in the most recent quarter; ② Odds indicator: adjustment range >20% from recent quarter average trading price to year's highest price; ③ Chip indicator: stock price amplitude <25% in recent quarter; ④ Other correction indicators: including current stock price above monthly line, financial indicators, absolute valuation indicators, etc.
According to the latest screening results, besides TMT sectors that have accumulated significant gains, sectors with high overlap include auto parts/robotics recent lagging segments, power grid equipment, consumer electronics, etc. These sectors with call option attributes have generally lagged recently, with relatively controllable downward adjustment ranges and sensitive upward responses to positive news, making them relatively suitable for medium-term forward positioning.
**Part III: Important Changes This Week**
Unless otherwise specified, all data sources in this chapter are Wind data.
**(1) Mid-stream Industries**
**1. Downstream Demand**
Real Estate: As of September 27, cumulative year-over-year decline in real estate transaction area of 30 major cities was 5.01%, month-over-month increase of 12.00%, year-over-year increase of 11.51%, week-over-week increase of 23.33%. According to National Bureau of Statistics data, Jan-Aug new construction area was 398 million square meters, cumulative year-over-year decline of 19.50%, down 0.10% compared to Jan-Jul growth rate; August single-month new construction area was 46 million square meters, year-over-year decline of 19.84%; Jan-Aug national real estate development investment was 6030.919 billion yuan, nominal year-over-year decline of 12.90%, down 0.90% compared to Jan-Jul growth rate, August single-month new investment nominal year-over-year decline of 19.95%; Jan-Aug national commercial housing sales area was 573.04 million square meters, cumulative year-over-year decline of 4.70%, down 0.70% compared to Jan-Jul growth rate, August single-month new sales area year-over-year decline of 10.98%.
Automobiles: Passenger vehicles: September 1-21, national passenger vehicle market retail 1.191 million units, year-over-year increase of 1% compared to same period last September, 8% increase compared to same period last month, cumulative retail of 15.955 million units since beginning of year, year-over-year increase of 9%; September 1-21, national passenger vehicle manufacturer wholesale 1.307 million units, 0% change compared to same period last September, 16% increase compared to same period last month, cumulative wholesale of 19.349 million units since beginning of year, year-over-year increase of 12%. New Energy: September 1-21, national passenger vehicle new energy market retail 697,000 units, year-over-year increase of 10% compared to same period last September, 11% increase compared to same period last month, national passenger vehicle new energy market retail penetration rate 58.5%, cumulative retail of 8.267 million units since beginning of year, year-over-year increase of 24%; September 1-21, national passenger vehicle manufacturer new energy wholesale 724,000 units, year-over-year increase of 10% compared to same period last September, 19% increase compared to same period last month, national passenger vehicle manufacturer new energy wholesale penetration rate 55.4%, cumulative wholesale of 9.668 million units since beginning of year, year-over-year increase of 31%.
**2. Mid-stream Manufacturing**
Steel: Rebar spot prices rose 1.83% week-over-week to 3,284.00 yuan/ton, stainless steel spot prices fell 0.63% week-over-week to 13,501.00 yuan/ton. As of September 26, rebar futures closing price was 3,114 yuan/ton, down 1.83% from last week. According to steel network data, mid-September, average daily output of key statistical steel enterprises was 2.061 million tons, up 5.37% from early September. August cumulative crude steel output was 671.8057 million tons, year-over-year decline of 2.80%.
Chemicals: As of September 20, styrene prices fell 300.18% from September 10 to 7,066.90 yuan/ton, methanol prices fell 33.10% from September 10 to 2,258.50 yuan/ton, PVC prices fell 152.77% from September 10 to 4,802.00 yuan/ton, polybutadiene rubber prices fell 6.74% from September 10 to 11,571.40 yuan/ton.
**3. Upstream Resources**
International Commodities: WTI rose 4.85% this week to $65.72, Brent rose 4.19% to $68.82, LME metals price index rose 1.00%, CRB commodity index rose 2.02% this week to 305.03, BDI index rose 2.54% last week to 2,259.00.
Iron Ore and Coal: Iron ore inventory rose this week, coal prices increased. Qinhuangdao Shanxi quality mixed 5500 warehouse price rose 2.60% to 694.00 yuan/ton as of September 22, 2025; port iron ore inventory rose 1.44% this week to 140.02 million tons; raw coal August output increased 2.50% to 390.497 million tons.
**(2) Stock Market Characteristics**
Stock Market Performance: Shanghai Composite Index rose 0.21% this week. Top three industry gains were Power Equipment (SWICS) (3.86%), Non-ferrous Metals (SWICS) (3.52%), Electronics (SWICS) (3.51%); top three declines were Commercial Trade (SWICS) (-4.32%), Conglomerates (SWICS) (-4.61%), Social Services (SWICS) (-5.92%).
Dynamic Valuations: A-share overall PE (TTM) rose from 19.41x last week to 19.42x this week, PB (LF) remained at 1.79x; A-share overall excluding financials PE (TTM) rose from 28.80x last week to 28.85x this week, PB (LF) rose from 2.47x to 2.48x. ChiNext PE (TTM) fell from 54.53x last week to 54.52x this week, PB (LF) remained at 4.38x; STAR Market PE (TTM) rose from 101.84x last week to 105.25x this week, PB (LF) rose from 5.32x to 5.50x. CSI 300 PE (TTM) rose from 13.44x last week to 13.51x this week, PB (LF) remained at 1.42x. From industry perspective, industries with largest PE (TTM) percentile expansion this week were Power Equipment, Utilities, Non-ferrous Metals. Industries with smallest PE (TTM) percentile expansion were Pharmaceuticals, Commercial Trade, Social Services. Additionally, from PE perspective, among SWICS Level 1 industries, Petrochemicals, Non-ferrous Metals, Construction & Decoration, Utilities, Transportation, Environmental Protection, Social Services, Household Appliances, Agriculture/Forestry/Animal Husbandry/Fishery, Food & Beverages, Telecommunications, Non-bank Financials have valuations below historical medians. Real Estate, Electronics, Computers have valuations above historical 90th percentiles. From PB perspective, among SWICS Level 1 industries, Petrochemicals, Basic Chemicals, Steel, Building Materials, Construction & Decoration, Utilities, Transportation, Real Estate, Environmental Protection, Beauty Care, Social Services, Household Appliances, Light Manufacturing, Textiles & Apparel, Commercial Trade, Agriculture/Forestry/Animal Husbandry/Fishery, Food & Beverages, Pharmaceuticals, Media, Banks, Non-bank Financials have valuations below historical medians. Electronics has valuations above historical 90th percentiles. Equity risk premium fell from 1.59% last week to 1.58% this week, stock market yield remained at 3.47%.
Margin Trading Balance: As of Thursday September 25, margin trading balance was 2,444.317 billion yuan, up 1.93% from last week.
AH Premium Index: AH stock premium index fell to 119.66 this week, compared to 117.11 last week.
**(3) Liquidity**
From September 21 to September 27, the central bank had 5 reverse repo maturities totaling 1,826.8 billion yuan; 7 reverse repos totaling 2,467.4 billion yuan. Open market operations net withdrew 940.6 billion yuan (including treasury cash).
As of September 26, 2025, R007 rose 28.15BP this week to 1.5538%, SHIBOR overnight rate rose 1.10BP to 1.3210%; term spread rose 5.13BP to 0.2157%; credit spread fell 3.04BP to 0.4846%.
**(4) Overseas**
United States: September 24 released Q2 Real GDP: Seasonally Adjusted: Annualized QoQ (Final) (%) at 3.80, previous -0.6; Q2 Core PCE Price Index: Seasonally Adjusted: Annualized QoQ (Final) (%) at 2.60, previous 3.30.
Eurozone: September 22 released September Eurozone: Consumer Confidence Index: Seasonally Adjusted (Preliminary) at -14.90, previous -15.50.
United Kingdom: September 23 released September CBI Industrial Trends Orders Index (%) at -27%, previous -33%; September 25 released September CBI Retail Sales Balance (%) at -29%, previous -32%.
Japan: September 26 released September Tokyo CPI: YoY (%) at 2.50%, previous 2.50%.
Overseas Stock Markets: S&P 500 fell 0.31% last week to 6,643.70; London FTSE rose 0.74% to 9,284.83; German DAX rose 0.42% to 23,739.47; Nikkei 225 rose 0.69% to 45,354.99; Hang Seng fell 1.57% to 26,128.20.
**(5) Macro**
Industrial Enterprise Profits: China August industrial enterprise profit total cumulative YoY 0.90%, previous -1.70%; monthly YoY 20.40%, previous -1.50%.
**Part IV: Next Week's Data Calendar**
Next week's highlights: China September Official Manufacturing PMI, UK Q2 GDP (Revised): Seasonally Adjusted: YoY (%), US September ISM Manufacturing PMI, September Eurozone: CPI: MoM (%) (Preliminary), US September 27 Initial Jobless Claims: Seasonally Adjusted (persons), August Eurozone: Unemployment Rate: Seasonally Adjusted (%), US September Unemployment Rate: Seasonally Adjusted (%)
September 30 Tuesday: China September Official Manufacturing PMI, UK Q2 GDP (Revised): Seasonally Adjusted: YoY (%)
October 1 Wednesday: US September ISM Manufacturing PMI, September Eurozone: CPI: MoM (%) (Preliminary)
October 2 Thursday: US September 27 Initial Jobless Claims: Seasonally Adjusted (persons), August Eurozone: Unemployment Rate: Seasonally Adjusted (%)
October 3 Friday: US September Unemployment Rate: Seasonally Adjusted (%)
**Part V: Risk Warnings**
Overseas economic data and interest rate cut expectation volatility impact on domestic markets; year-end important meeting policy increments and economic work direction-setting falling short of expectations; repeated China-US trade and financial friction; Middle East geopolitical risks bringing energy cost volatility and global risk appetite decline, etc.