文/ 万点研究，作者 / 风言
从18C到AI指数，港交所把科技资产这条链补上了。
10月23日，香港交易所宣布正式推出香港交易所人工智能指数及香港交易所半导体指数。人工智能指数追踪40家与AI产业链高度相关的港股公司，覆盖芯片设计、数据中心、AI基础设施、模型及软件与实体应用五层架构；半导体指数追踪30家最大型港股半导体相关公司，涵盖制造、设备、元件、算力硬件及储存设备。
港交所同时宣布与南方基金签订授权协议，后者将在内地发行追踪AI指数的ETF。具体来看，两大指数均纳入中芯国际、联想集团、小米集团、英诺赛科、中际旭创、壁仞科技、天数智芯等细分龙头。
相关一级市场投资者表示，两大指数为港股AI与半导体板块提供了新的定价与配置基准；有望进一步引导资金关注相关龙头，提升板块的代表性与流动性。指数落地叠加ETF授权，港股AI与半导体板块迎来新的配置工具与资金入口。
1 港交所落子主题指数：AI与半导体迎来新定价基准
港交所此次推出两大主题指数，表面是为市场增添参考基准，实质是在科技资产的产品化链条上落子。
客观而言，长期以来，港股科技主题指数的话语权主要由恒生指数公司掌握。据21世纪经济报道消息，截至2026年6月底，追踪恒生科技指数的被动式产品资产管理总值已达404亿美元（约2725亿元人民币），较2020年推出初期的15亿美元增长近26倍。
另据恒生指数公司披露，截至上半年，追踪恒生指数系列的ETP产品管理总值已超过1300亿美元，约为香港本地ETP市场规模的1.5倍。
港交所自主编制AI与半导体指数，意味着它不再满足于仅提供交易场所，而是试图将“上市地优势”向上游的指数编制与下游的产品授权延伸。
南方基金获得的AI指数ETF授权，正是这一链条的第一个落点。但它的起点，是一个已被数百亿美元被动资金深度锚定的市场。
编制细节值得细看。AI指数并非简单按市值排序选股，而是设有明确的基本面门槛：成分股须满足近两年研发费用、资本开支占销售额比例均不低于3%，销售额增速不低于5%。权重设计上，单一成分股不超过10%，应用环节个股不超过3%，按季度调样。
这一套规则传递的信息是：港交所想要的是一个“有实质AI投入”的基准，而非追逐概念热度的篮子。半导体指数则以“30家最大型”为选样口径，侧重规模代表性，与AI指数的基本面倾斜形成方法学上的分工。
但编制规则本身也隐含张力。AI指数10%的权重上限意味着中芯国际、联想、小米等大市值标的无法主导指数走向，40只成分股的“平均发挥”将决定指数表现。在AI板块内部剧烈分化的环境中，均衡配置未必跑赢集中押注少数赢家的主动策略。
半导体指数以规模选样，本质上是在追认当前的市场格局而非预判未来的产业赢家。当规模最大的成分股恰好面临资金撤离时，指数的短期表现将不可避免地受到拖累。这不是编制方法的缺陷，而是主题指数与生俱来的顺周期属性。
对投资者而言，在ETF产品正式落地并运行一段时间之前，新指数更多是观察港股科技结构的参照系，而非即刻可用的配置工具。南方基金的授权是起点，但产品募集时机、后续规模与跟踪误差，才是检验指数实用性的真正标尺。
2 从18C到主题指数：港交所科技资产服务链条渐完整
将视线拉远，港交所此举嵌在两层结构之中。
第一层是港股市场底色的持续位移。科技股在港股总成交中的占比已从十年前的15%升至44%，成交额增长超过7倍；2026年上半年，科创企业占港股IPO发行规模的63%，去年同期仅为16%。
港交所数据显示，2025年12月至2026年5月，AI价值链公司新股集资总额达979亿港元，占同期港股新股集资额的55%，另有70家AI价值链公司在排队。
上市规则第18C章是香港联合交易所于2022年10月发布咨询文件、2023年3月正式推出的专为特专科技公司设计的创新上市制度。
港交所数据显示，18C机制下，截至2026年8月31日，23家特专科技企业再融资总额594.26亿港元，已超过IPO合计募资的403.58亿港元，持续融资通道已经跑通。
港交所此时以官方身份编制AI与半导体指数，是对这一变化的制度性确认。它与18C形成配套：18C解决“让公司进来”，主题指数解决“进来之后如何被定价和配置”。
第二层背景更为复杂。中美科技竞争持续、半导体出口管制未见松动、AI算力自主成为政策优先级，香港被赋予“中国科技资产离岸定价中心”的角色期待。
Wind数据显示，2025年，南向资金全年净买入超1.4万亿港元，创历史新高。2026年7月以来，南向资金净买入规模进一步扩大，7月以来累计净买入超千亿港元。在市场下行阶段起到稳定器作用。
相关证券内部人士对万点研究表示，对新指数的评估应当区分三个时间尺度：制度层面，港交所填补了自主硬科技基准的空白，方向正确，且与18C形成配套逻辑。产品层面，AI指数有了第一个落点，半导体指数尚待观察，两者的节奏差异本身就在传递信息。市场层面，近期回调反而为新指数挂钩产品提供了低位建仓的窗口。对以长期配置为目标的投资者而言，这反而构成了分批低吸、等待挂钩产品落地后逐步建仓的有利时点。
制度红利是长期的，产品落地是中期的，市场情绪是短期的。将三者混为一谈，容易在制度意义的光环下忽略短期的现实阻力。更务实的姿态是：把新指数当作观察港股科技结构的长期参照系，等待挂钩产品实际运行、规模与跟踪数据可查之后，再判断它是否值得成为配置工具。
3 长期制度红利之外，新指数仍需跨过这道坎
不过，长期制度红利落地之前，成分股层面的短期压力已经先行显现。
新指数纳入的“龙头”，恰好分布在当前资金撤离与估值收缩最集中的环节，这为指数的短期前景增添了现实约束。
同样以壁仞科技为例。Wind数据显示，公司股价在6月触及70.75港元高点后持续下行，截至10月26日，公司港股价格报收23.70港元，距离高点累计回撤超65%。
股价承压的同时，公司上市不到一年内两次折价配售，7月以46.20港元配售募资约70亿港元，10月又以31.08港元配售1.3亿股，两轮累计从市场拿走超过170亿港元。
值得注意的是，以壁仞科技股份解禁与配售形成了某种“默契”。7月1日基石解禁，公司三天后宣布配售；10月6日配售股份禁售期届满，两天后再度配售。
以至于市场对公司形成了“解禁即抽水”的预期，2027年1月还有占总股本33.47%的创始人及IPO前股东解禁窗口在前方等待。
壁仞的困境不是孤例，它折射的是国产AI芯片公司普遍的悖论：技术追赶需要持续巨额投入，而资本市场的耐心在股价下行时会迅速消耗。
中际旭创代表另一类压力。作为光模块龙头，它是AI算力基建链条上最直接的受益标的之一，但股价同样承受剧烈波动。
Wind数据显示，中际旭创10月22日单日下跌近8.7%，年内涨幅尽数抹去，股价创4月以来新低。融资余额从7月中旬的315.77亿港元峰值降至162亿附近，近三个月规模缩水接近五成。
融资盘的持续撤出，反映的是市场对AI投资热情的重新审视。光模块需求与北美云厂商资本开支高度绑定，一旦全球AI基建节奏出现任何放缓信号，这类“卖铲子”的标的往往最先感受估值收缩。
外围环境同样不容乐观，高盛Prime经纪业务报告显示，10月8日至14日，对冲基金连续五个交易日净卖出全球AI股票，自2025年初以来的累计净买入在一周内被完全抹去。纳斯达克100指数创下2025年4月以来最大单周跌幅，芯片股连续八个交易日收跌。
市场担忧的焦点从“AI能做什么”转向“投入何时回本”：超大规模云厂商的资本开支承诺规模空前，但变现路径仍不清晰。对此，博通CEO警告，资金问题可能成为AI扩张面临的更大挑战。
这一背景对港交所新指数的含义是直接的：成分股中相当比例的公司估值建立在AI需求持续扩张的假设之上，若全球AI交易进入较长的冷却期，挂钩产品即便推出，也可能面临资金流入缓慢的现实。
4 一张有分量的牌，南方基金能不能打成王炸？
总结而言，南方基金的AI ETF授权是一个积极信号，但一个指数产品的生命力，最终取决于上市后的规模增长与二级市场流动性。
港交所将AI指数的内地ETF授权交给了南方基金，而非其他头部机构，这一选择本身值得解读。公司总经理杨小松将此次合作定位为“践行国际化战略、服务投资者多元资产配置需求的重要实践”。
从战略层面看，这确实是南方基金在跨境指数产品线上的一次差异化占位：港交所AI指数覆盖40只成分股，涵盖芯片设计、AI基础设施、模型与应用全栈，与境内现有的AI主题ETF在成分构成和叙事框架上形成区分。拿到独家授权的先发优势，意味着南方基金在“港股AI”这一细分标签上暂时没有直接竞争对手。
但机遇的另一面是执行层面的多重约束。首当其冲的是发行窗口。据界面新闻消息，2026年三季度，公募新发基金数量创2022年以来新高，但平均募集规模仅约4.77亿元，为2005年以来最低；股票ETF的“踩线”成立比例超过70%。进入10月以来成立的15只非发起式基金中，10只规模不足3亿元，占比67%。
南方基金自身的产品线中，10月12日成立的南方创业板软件ETF募集规模也仅约2.2亿元，踩在2亿元门槛线上。在这一发行环境下，AI指数ETF即便启动募集，首募规模能否达到有意义的量级，是南方基金需要面对的第一个现实问题。
在万点研究看来，一个规模勉强过线的ETF，即便指数编制再精细，也难以在二级市场形成有效的流动性。而港交所AI指数挂钩产品目前仅南方基金一家，还需要更多发行人共同“做大蛋糕”。
一个主题ETF的流动性生态，往往需要多家机构同时布局才能形成，独家授权在竞争壁垒上是优势，在生态培育上却可能是劣势。
更为重要的是，从节奏时间上看，港交所AI指数恰好在科技板块经历一轮显著调整的时点推出，壁仞科技、中际旭创等成分股正处于股价回撤与资金流出的压力之中。
南方基金若在此时以较大的营销力度推动AI指数ETF发行，需要在“把握战略机遇”与“避免高位发行”之间找到平衡点。这不是说产品不该发，而是发行的节奏、规模预期与投资者沟通方式，都需要比市场热度期更加审慎。
从更长远的角度看，南方基金此次授权的真正价值，或许不在于首只ETF的募集规模，而在于能否在跨境指数产品线上建立可持续的运营能力。港交所AI指数按季度调样，设有研发投入与资本开支的基本面门槛，成分股涵盖A股与港股两地上市的AI相关公司。
这意味着产品的日常管理涉及跨市场成分股调整、汇率因素，以及两地交易节奏差异带来的跟踪误差控制。南方基金在跨境ETF领域已有一定积累，但一只全新标的的运营磨合期不可避免。
首只产品的跟踪误差数据、调样执行效率，以及上市后的持续营销能力，将决定这次授权是否成为南方基金跨境产品线的一个长期支点，还是仅仅多了一只规模有限的主题ETF。
综合来看，南方基金在这笔授权中拿到的是一张有分量的牌，但牌局的走向取决于三个变量：发行窗口的选择是否足够耐心，产品定位是否经得起行业导向的审视，以及运营能力能否支撑一只跨市场主题ETF的长期运转。
港交所的指数编制为产品提供了基础，但产品能否从“获得授权”走到“被投资者真正使用”，答案不在授权协议里，而在南方基金接下来的执行节奏中。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
Text/ Point Study, Author / Word
From the 18C to the AI index, HKCs have built up the chain of technological assets.
On 23 October, the Hong Kong Exchange announced the official launch of the Hong Kong Exchange's artificial intelligence index and the Hong Kong Exchange's semiconductor index. The artificial intelligence index tracks 40 port stock companies that are highly relevant to the AI industrial chain, covering the five-tiered architecture of chip design, data centres, AI infrastructure, models and software and physical applications; and the semiconductor index tracks 30 of the largest port equity and semiconductor-related companies, covering manufacturing, equipment, components, computing hardware and storage equipment.
At the same time, it announced an authorization agreement with the South Fund, which will issue an ETF tracking the AI index in the interior. In particular, the two indicators are integrated into the sub-heads of CGI, IRG, Minami Group, Innosaic, Middle Wrestling, Wallstone Technology, Day Whip, etc.
Market investors at the relevant level indicated that the two major indices provide new pricing and allocation benchmarks for the AI and semiconductor segments of the Port Unit; it is expected that the funding will be further directed to the relevant lead points and improve the representation and liquidity of the blocks. The index drop-down superimposes the ETF authorization, and the Hong Kong Unit's AI and semiconductor plates receive new configuration tools and financial entry.
1 Portfolio sub-theme index: AI and semiconductor welcome new pricing benchmarks
This time, the Hong Kong Institute has launched two major thematic indicators, ostensibly to provide market reference points, essentially on the chain of production of science and technology assets.
Objectively, Hong Kong's S&T indicators have long been dominated by the S&T Index. According to economic reports for the twenty-first century, by the end of June 2026, the total value of passive product asset management for tracking the heinogenous technology index had reached $40.4 billion (approximately RMB 272.5 billion), nearly 26 times greater than the $1.5 billion initially introduced in 2020.
Heinrich Index also disclosed that, as of the first half of the year, the total management value of ETP products tracking the Heinogeneity Index series had exceeded $130 billion, approximately 1.5 times the size of the local ETP market in Hong Kong.
The autonomous production of the AI and Semiconductor Index by the Hong Kong Institute means that it is no longer satisfied with the mere provision of a place of dealing, but rather seeks to extend the “market edge” to upstream indexing and downstream product authorization.
The AI Index ETF mandate obtained by the South Fund was the first point of departure in this chain. But its starting point is a market that has been anchored in tens of billions of dollars of passive capital.
The details are worth looking at. The AI index does not simply rank the shares by market value, but has a clear fundamental threshold: the component shares have to cover R & D costs for the last two years, capital expenditures are not less than 3 per cent of sales, and sales are increasing at less than 5 per cent. In the design of weights, the single component unit shall not exceed 10 per cent and the application chain unit shall not exceed 3 per cent, which shall be adjusted quarterly.
The message of this set of rules is that Hong Kong House wants a “substantive AI input” benchmark, rather than a basket to chase the concept heat. The semiconductor index is based on a sample calibre of “30 largest”, with a focus on representation of size and a methodological division of labour tilted towards the basic face of the AI index.
But the rules themselves imply tension. The maximum weight of 10 per cent of the AI index means an unruly trend of the large market value indicators, such as China-Chip International, Consortium, and millet, and the “average utilization” of 40 components will determine the performance of the index. In a highly divided environment within AI, balanced configuration does not necessarily win the proactive strategy of focusing on a few winners.
The semiconductor index, which is sampled on a scale, is essentially about following current market patterns rather than prejudging future industrial winners. The short-term performance of the index will inevitably be delayed when the largest component shares coincide with the withdrawal of funds. This is not a flaw in the methodology, but rather the inherent procyclical properties of the thematic index.
For investors, the new index is more of a reference system for observation of the technological structure of the Port stock than a readily available configuration tool, some time before the ETF product is officially landed and operational. The mandate of the South Fund was the starting point, but the timing of the collection of products, the scale of follow-up and the error in tracking were the real yardstick for testing the usefulness of the index.
2 From 18C to thematic index: the Hong Kong Institute ' s scientific and technical asset service chain is becoming complete
Widening the horizon, the Hong Kong Tsang is embedded in two layers.
The first level is the continuous movement of the bottom colour of the market for shares in ports. The share of the Science and Technology Unit in the total portfolio of the Port Unit has increased from 15 per cent a decade ago to 44 per cent, more than sevenfold; in the first half of 2026, Science and Technology accounted for 63 per cent of the distribution of IPOs in the Port Unit, compared to 16 per cent in the same period last year.
Between December 2025 and May 2026, HK$ 97.9 billion was raised in new shares of AI Value Chains, accounting for 55 per cent of the new shares of Hong Kong Stock during the same period, and 70 other AI Value Chain Companies were in line.
Chapter 18C of the Listing Rules is an innovative listing system designed for specialized technology companies, officially launched in March 2023 by the Hong Kong Joint Exchange in October 2022.
According to data from the Hong Kong Institute of Technology, under the 18C mechanism, as at 31 August 2026, the refinancing of 23 specialized science and technology enterprises totalled HK$59,426 million, exceeding HK$40,358 million in total IPO fund-raising, and the continuous financing corridor had run out.
The development of the AI and Semiconductor Index at this time in the official capacity is an institutional confirmation of this change. It goes hand in hand with 18C: 18C addresses “involving companies”, and the thematic index addresses “how to be priced and configured after coming in”.
The second layer is more complex. China and the United States continue to compete in science and technology, export controls for semiconductors have not been relaxed, AI has become a policy priority, and Hong Kong has been given the role of “the offshore pricing centre for Chinese science and technology assets”.
In 2025, Wind data show that South-to-South net purchases of more than HK$1.4 trillion were recorded throughout the year. Since July 2026, net South-to-South purchases have expanded further, with cumulative net purchases of over HK$ 100 billion since July. Act as a stabilizer at the down-market stage.
According to the 10,000-point study, the assessment of the new index should be based on three timescales: at the institutional level, the Hong Kong firm fills the gaps in the autonomous hard-tech benchmark, is in the right direction and follows the logic of 18C. At the product level, the AI index has its first drop point, and the semiconductor index is yet to be observed, and the rhythm differences between the two are in themselves transmitting messages. At the market level, the recent reversal has instead provided a low-level window for new index-linked products. For investors targeting long-term configurations, this constitutes a favourable time point for low-snaping groups, waiting for tied products to land.
The institutional dividend is long-term, the product is medium-term and the market sentiment is short-term. Combining the three with one another makes it easier to ignore short-term practical resistance in the light of systemic significance. A more pragmatic gesture is to use the new index as a long-term frame of reference for observation of the scientific and technological structure of the Port Unit, pending the availability of data on the actual operation, size and tracking of the linked product before determining whether it is worth becoming a configuration tool.
3 In addition to the long-term institutional dividend, the new index still needs to cross this path.
However, short-term pressures at the component unit level were already visible before the long-term institutional dividends were realized.
The inclusion of “leads” in the new index, which coincided with the most concentrated segment of the current withdrawal and valuation contraction, added a realistic constraint to the short-term prospects of the index.
Also, for example, the technology of walls. Wind data show that company stock prices continued to decline after reaching a high of HK$ 70.75 in June, and as at 26 October, the company ' s stock prices had been reported at HK$ 23.70, which represented a cumulative withdrawal of more than 65 per cent from high points.
At the same time as the stock price was under pressure, the company was on the market twice in less than one year at a discount, raising approximately HK$ 7 billion in July at HK$ 46.20, and in October at HK$ 31.08 million, taking more than HK$ 17 billion from the market in two rounds.
It is worth noting that the release and distribution of the wallen technology shares has created some kind of “conciliation”. On 1 July, the cornerstone was released and the company announced a distribution three days later; on 6 October, the share was closed and re-assigned two days later.
To the extent that the market expects the company to be “free or pumped”, in January 2027 there were also 33.47 per cent of the total equity of the founders and former shareholders of the IPO that were ahead of the embargo window.
The dilemma of walls is not an isolated one; it mirrors the general paradox of national AI chip companies: technology catching up requires huge and sustained investment, and the patience of capital markets is quickly consumed when stock prices go down.
It's a different kind of pressure. As the head of the optical module, it is one of the most direct beneficiaries of the AI base construction chain, but shares are equally subject to sharp fluctuations.
Wind data show that the single-day middle-class boom fell by almost 8.7 per cent on 22 October, with the year-on-year increases wiped out and stock prices at a new low since April. The financing balance fell from a peak of HK$31,577 million in mid-July to close to $16.2 billion, with a nearly 50 per cent reduction in size over the past three months.
The continued withdrawal of the fund pool reflects a re-examination of the market ' s enthusiasm for AI investment. The light module demand is highly tied to the capital expenditure of North American cloud manufacturers, and the object of “sold shovels” of this type tends to be the first to feel the decline in valuation once there is any slowing signal at the global AI capital rhythm.
The peripheral environment is no less optimistic, and the Goldman Sachs Prime Brokering Report shows that hedge funds sold global AI stocks on five consecutive trading days from 8 to 14 October, and that cumulative net purchases since the beginning of 2025 were completely wiped out within a week. The NASDAQ 100 index reached its largest single-week decline since April 2025, with the chip unit falling eight consecutive trading days.
The focus of market concern has shifted from “what AI can do” to “when it comes back”: capital spending commitments by super-large cloud producers are unprecedented in scale, but the path to realization remains unclear. In response, BOTO warned that the issue of finance could become a greater challenge to AI ' s expansion.
This background is direct to the implications of the new Hong Kong law enforcement index: A significant proportion of corporate valuations in the composition stock are based on the assumption that AI demand continues to expand, and if global AI transactions enter a longer cooling period, the linkage product, if introduced, may face the reality of slow financial inflows.
Can the South Fund be the king's bomb?
In conclusion, the AI ETF mandate of the South Fund is a positive signal, but the viability of an index product ultimately depends on post-market scale growth and secondary market liquidity.
The decision of HKC to delegate the authorization of the ETF for the Mainland of the AI Index to the South Fund, rather than to other head offices, is per se a worthy reading. The general manager of the company, Yang Xiaoxong, positioned this collaboration as “a key practice in implementing the strategy of internationalization and in servicing the multiple asset allocation needs of investors”.
At the strategic level, this is indeed the case for the South Fund ' s one-time differential on the cross-border index product line: The HKAI index covers 40 component units, covering chip design, AI infrastructure, models and applications, distinguishing between the composition and narrative framework of the existing AI theme ETF in the country. The pre-emptive advantage of obtaining exclusive authorization means that the South Fund temporarily has no direct competitors on the sub-label “Hong Kong Unit AI”.
But the other side of the opportunity is the multiple constraints at the implementation level. First and foremost is the distribution window. According to the interface news, the number of public fund-raising and development funds had risen since 2022 in the third quarter of 2026, but the average size of fund-raising was only about $477 million, the lowest since 2005; the share of ETF “stepping” was set up by more than 70 per cent. Of the 15 non-initiative funds established since October, 10, or 67 per cent, were on a scale of less than $300 million.
Of the South Fund ' s own product lines, the Southern entrepreneurship software ETF, which was established on 12 October, also raised just about 220 million yuan on the $200 million threshold. In this distribution environment, the first real problem facing the South Fund is whether the AI Index ETF, even if launched, can reach a meaningful level of initial funding.
From the perspective of the 10,000-point study, it is difficult to create effective liquidity in secondary markets for an ETF of marginal size, even if the index is refined. In contrast, Hong Kong's AI index-linked product is currently limited to the South Fund, and more publishers are needed to co-make the cake.
The mobile ecology of an ETF theme often requires a mix of institutions to develop, and exclusive authorization may be an advantage in competitive barriers but a disadvantage in ecological nurturing.
More importantly, in terms of tempo time, the HKAI index was launched at a time when the technology plate was undergoing a significant round of adjustments, and the construction of such components as wall tectonic technology, meso-breeding, etc., were under pressure from stock price reversals and capital outflows.
If the South Fund at this time promotes the release of the AI Index ETF with greater marketing effort, it needs to find a balance between “taking strategic opportunities” and “avoiding high-level distribution”. This is not to say that the product should not be distributed, but rather that the pace of issuance, the expected scale, and the way in which investors are communicated need to be more prudent than the heat of the market.
In the longer term, the true value of the mandate of the South Fund may lie not in the size of the first ETF but in the ability to build sustainable operational capacity on the cross-border index product line. The HKI Index, which is adapted quarterly, has a basic threshold for R & D inputs and capital expenditure, and the component shares cover AI-related companies listed in both shares A and Port.
This means that the day-to-day management of the product involves cross-market component unit adjustments, exchange rate factors, and tracking error controls resulting from differences in the pace of trading between locations. The South Fund has accumulated a certain amount in the cross-border ETF area, but a completely new target for operational consolidation is inevitable.
The tracking error data for the first product, the streamlining of the efficiency of implementation, and the continued marketing capacity after the listing will determine whether the mandate will become a long-term feeder of the South Fund ' s cross-border product line, or whether it will simply add a limited theme, ETF.
Taken together, the South Fund has received a merit card in this mandate, but the direction of the game depends on three variables: whether the distribution window is chosen with enough patience, whether product positioning is able to withstand industry-oriented scrutiny, and whether operational capacity supports the long-term operation of an inter-market theme ETF.
The indexing of port houses provides the basis for the product, but whether the product can move from “authorization” to “real use by investors”, the answer is not in the authorization agreement, but in the next implementation tempo of the South Fund.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.