How to Converge AH?

Summary

Core conclusion: Since the beginning of this year, the Hang Seng TECH Index has significantly outperformed the CSI 300, with a high excess quantile, but there may still be excess after excess. The convergence in the medium to long term does not necessarily indicate a short-term convergence. In terms of magnitude, the current excess still has a gap compared to the period from November 2022 to January 2023, and the time frame is also shorter than the three-month period from November 2022 to January 2023 and the six-month period from August 2019 to February 2021. We believe that about 80% of the current difference in the rise between A-shares and H-shares can be explained by the risk-free arbitrage brought by foreign exchange swaps, with the remaining 20% attributed to the presence of many high-quality technology assets in Hong Kong stocks and benefits from DeepSeek catalysts.

The strength of the Hang Seng TECH Index may not necessarily start to converge in the short term: (1) From the perspective of short-term excessive gains, the current performance of the Hang Seng Index and the Hang Seng TECH Index is highly consistent, resembling the situation from November 2022 to January 2023, with the current excessive gain level of the Hang Seng TECH Index still having a 12% space compared to the extreme value of 47% in January 2023. (2) From the medium to long-term trend of narrowing AH premiums, compared to the lowest point in February 2021, there is still some distance for the CSI 300/Hang Seng TECH Index to reach the bottom. Based on these two points, the rise of the Hang Seng TECH Index may still have room to lead A-shares. (3) From the historical perspective of excessive gains returning to normal, the Hang Seng TECH Index has been excessively high for less than two months since the beginning of this year, compared to the three-month period from November 2022 to January 2023 and the six-month period from August 2019 to February 2021, indicating that there is still room for progress. Looking ahead, historically, it generally takes about one month for the relative return of AH to revert to zero from the lowest point over a rolling three-month period.

The essence of foreign exchange swaps is the subsidized yield for investing in RMB assets, which serves as a "subsidy" for delaying the conversion of RMB into foreign exchange (according to interest rate parity theory, foreign exchange swaps are determined by the interest rate differentials between the two countries). Hong Kong, as an offshore RMB forward market, provides a channel for this arbitrage, allowing foreign capital to synthesize ten-year Chinese government bonds denominated in USD in the presence of interest rate differentials between China and the U.S. When foreign capital invests in Hong Kong stocks, they can obtain an additional foreign exchange swap subsidy compared to investing in A-shares due to the existence of foreign exchange swap points. Based on this logic, the specific proportion of this subsidy can be quantified using the swap points (the difference between the forward exchange rate and the spot exchange rate) divided by the spot exchange rate. Under the no-arbitrage assumption, it can be inferred that for any entity that ultimately needs USD (which can be Chinese individuals engaged in trade, foreign capital, etc.), the discount rate for investing in A-shares should be lower by a difference equivalent to the foreign exchange swap.

According to the DDM model, the difference in discount rates can be calculated by deriving the PE and dividend rates of the Hang Seng Index and the CSI 300 Index, and comparing it with the foreign exchange swap subsidy rate to examine the explanatory power of foreign exchange swap points on the recent performance differences between A-shares and H-shares. Empirically, recent changes in foreign exchange swaps can basically explain the changes in the discount rates (valuations) of the CSI 300 and the Hang Seng Index. Since the beginning of this year, the difference in discount rates between the CSI 300 and the Hang Seng Index has changed by 0.65% from its peak, while the subsidy rate brought by foreign exchange swaps has also changed by 0.52%. We believe this indicates that the foreign exchange swap subsidy may explain 80% of the recent performance differences between A-shares and Hong Kong stocks The remaining 20% is attributed to the fact that Hong Kong stocks have more high-quality technology assets and benefit more from factors such as DeepSeek catalysis, which are more related to the "overheating" of trading.

Industry allocation suggestion: The third phase of the Spike 2.0 offensive is underway, and the pulse of social financing has shown a rebound. Further observation of the improvement in fundamentals will take place after the Spring Festival. The Hang Seng has already taken the lead, while A-shares are lagging behind. Based on economic recovery and market liquidity, the investment main line can be reduced to three directions:

1. The breakthrough of Deepseek and the leadership of open-source technology AI+ (see "DeepSeek Series: AI+ Investment Map" (20250206)), 2) The valuation repair of consumer stocks and the gradual recovery of consumption stratification (see "Let Some People Consume First" (20250102)), 3) The continued rise of undervalued dividends. Dividend pullbacks often occur when strong industrial trends emerge, so the height of undervalued dividends depends on the progress of the AI industry trend, which in turn depends on breakthroughs in AI applications and consumption. The core factor for investment in the consumer sector is valuation. Given the current low valuation of the consumer sector, declining interest rates, and the recovery cycle under policy catalysis (even if it is a very weak slope), being overly pessimistic about consumption from a macro narrative perspective is a risk, emphasizing the Hang Seng Internet.

Main Text

1. Hang Seng's Leading Review and Attribution

1.1. The Space and Time for AH Excess Convergence

Reviewing each round of excessive gains in Hong Kong stocks since 2018, when the Hang Seng Technology Index and the Hang Seng Index are both outperforming the CSI 300, the excessive gains of Hong Kong stocks are often greater; conversely, if the two diverge, the excessive gains of Hong Kong stocks quickly revert. Since November 2024, the Hang Seng Technology Index has significantly outperformed the CSI 300. We believe that although the Hang Seng Technology has significantly outperformed the CSI 300 in the short term, the market may not necessarily interpret the convergence of AH in the short term.

(1) From the perspective of short-term excessive gains: The current performance of the Hang Seng Index and the Hang Seng Technology Index is highly consistent, resembling the situation from November 2022 to January 2023. The current level of excessive gains in the Hang Seng Technology still has a 12% space compared to the extreme value of 47% in January 2023.

(2) From the perspective of the medium to long-term trend of narrowing AH premiums: Since last November, the CSI 300 has overall underperformed the Hang Seng Technology Index. This trend is similar to that from August 2019 to February 2021, also benefiting from the market's continued optimism about the technology innovation industry and the improvement of the fundamentals of Chinese internet companies. Compared to the lowest point in February 2021, the CSI 300/Hang Seng Technology still has some distance to reach the bottom.

(3) From the perspective of the time from historical excessive gains to reversion: From the beginning of this year to now, the Hang Seng Technology has exceeded 35% in less than two months. Compared to the three-month period from November 2022 to January 2023 and the six-month period from August 2019 to February 2021, there is still room for progress. Looking back, the minimum point of relative returns from AH rolling over three months to reverting to 0 generally requires about one month

1.2. The AH performance since the beginning of the year can be explained by risk-free arbitrage by about 80%

The essence of foreign exchange swaps is the subsidized yield of investing in RMB assets, which is a "subsidy" for delaying the conversion of RMB into foreign exchange (according to the theory of interest rate parity, foreign exchange swaps are determined by the interest rate differentials between two countries). Foreign exchange swaps allow investors to borrow RMB at RMB interest rates and convert it to USD in the spot market, while simultaneously signing a forward foreign exchange contract to agree to convert back to RMB at a specific exchange rate in the future. As Hong Kong serves as an offshore RMB forward market, it provides a channel for this arbitrage, allowing foreign capital to synthesize USD-denominated ten-year Chinese government bonds in the presence of the China-U.S. interest rate differential. Specifically, if foreign capital buys ten-year Chinese government bonds and enters into a one-year foreign exchange swap:

1. In terms of coupon dimension, it is equivalent to synthesizing U.S. Treasury rates (Chinese government bond rates + RMB forward premium);

2. In terms of capital gains dimension, it is still based on the capital gains of Chinese government bond yields, which are more advantageous compared to U.S. Treasuries.

Therefore, when foreign capital invests in Hong Kong stocks, due to the existence of foreign exchange swap points, it can obtain an additional foreign exchange swap subsidy compared to investing in A-shares. Based on the above logic, the specific proportion of this subsidy can be quantified using the swap points (the difference between the forward exchange rate and the spot exchange rate) / spot exchange rate.

Under the no-arbitrage assumption, it can be inferred that for any entity that ultimately needs USD (which can be Chinese individuals engaged in trade, foreign capital, etc.), the discount rate for investing in A-shares should be lower than that of Hong Kong stocks by a margin equivalent to the foreign exchange swap difference. According to the DDM model, the discount rate = k/PE + g, where k is the dividend rate and g is the EPS growth rate. Since the constituents of the Hang Seng Index are mainly Chinese enterprises, assuming that the growth rate (g) of the CSI 300 and the Hang Seng Index is the same. We can backtrack the difference in discount rates from the PE and dividend rates of the two indices and compare it with the foreign exchange swap subsidy rate to examine the explanatory power of foreign exchange swap points on the recent performance differences between AH.

Recent fluctuations in foreign exchange swaps can basically explain the changes in the discount rates (valuations) of the CSI 300 and the Hang Seng Index. Since the beginning of this year, the difference in discount rates between the CSI 300 and the Hang Seng Index has changed by 0.65% from its peak, while the subsidy rate brought by foreign exchange swaps has also changed by 0.52%. We believe that among the excess gains of Hong Kong stocks relative to A-shares over the past 25 years, foreign exchange swap subsidies may explain nearly 80%, while the remaining 20% may be attributed to the presence of many high-quality technology assets in Hong Kong stocks More benefits from factors such as DeepSeek catalysis.

1. Industry comparison of capital situation

1. Industry profit expectations & valuations

Article authors: Eddie Wu, Sun Ximin, Source: Kaida Strategy Zhixing, Original title: "How to Converge AH? Hang Seng's Preliminary Review and Attribution"

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this is at one's own risk