Interactive daily chart of the Hong Kong Hang Seng Composite stock market index back to 1986. Each data point represents the closing value for that trading day and is denominated in hong kong dollars (HKD). The current price is updated on an hourly basis with today's latest value.

The statistic shows the annual development of the Hang Seng index from 1986 to 2024. The Hang Seng index reflects the performance of the largest stocks traded on the Hong Kong Stock Exchange. The year value of the Hang Seng index amounted to 20,059.95 by the end of 2023.

As of April 2025, the Hang Seng Index at the Hong Kong Exchange amounted to ********* points. After the outbreak of COVID-19, the index dropped as part of a broader Pan-Asian trend. However, by the end of 2020, when the pandemic situation stabilized in many countries and news about a vaccine rollout came out, the Hang Seng Index recovered and recorded significant increases every month. Index composition The Hang Seng Index is the most prominent indicator of stock performance on the Hong Kong Exchange. By including the 50 largest companies, the index represents the market movements of more than half of the bourse’s market capitalization. In addition to that, the Hang Seng Index has numerous smaller indices which mirror smaller industries or market sections. The Hang Seng Composite Index One example of a sub-index is the Hang Seng Composite Index. It reflects the performance of the top 95 percentile of the total market capitalization. The financial industry accounted for the largest share of companies included in the index, followed by the information technology sector. Prominent companies represented in the index are Tencent, AIA, and Meituan.

Hong Kong's main stock market index, the HK50, fell to 25524 points on July 29, 2025, losing 0.15% from the previous session. Over the past month, the index has climbed 6.03% and is up 50.12% compared to the same time last year, according to trading on a contract for difference (CFD) that tracks this benchmark index from Hong Kong. Hong Kong Stock Market Index (HK50) - values, historical data, forecasts and news - updated on July of 2025.

Key information about Hong Kong SAR (China) P/E ratio

• Hong Kong SAR (China) Hang Seng recorded a daily P/E ratio of 11.950 on 26 Mar 2025, compared with 12.200 from the previous day.
• Hong Kong SAR (China) Hang Seng P/E ratio is updated daily, with historical data available from Apr 2000 to Mar 2025.
• The P/E ratio reached an all-time high of 18.400 in Nov 2010 and a record low of 7.450 in Feb 2016.
• Hang Seng Indexes Company Limited provides daily P/E Ratio.

总回报指数：恒生指数在11-01-2018达74,068.82002Jan1990=2838.07，相较于10-01-2018的69,722.85002Jan1990=2838.07有所增长。总回报指数：恒生指数数据按月更新，01-01-1990至11-01-2018期间平均值为23,979.93002Jan1990=2838.07，共347份观测结果。该数据的历史最高值出现于01-01-2018，达88,755.56002Jan1990=2838.07，而历史最低值则出现于01-01-1990，为2,753.51002Jan1990=2838.07。CEIC提供的总回报指数：恒生指数数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

总回报指数：恒生复核材料业：服务业在11-01-2018达6,327.1702000年1月3日=2000，相较于10-01-2018的5,789.5302000年1月3日=2000有所增长。总回报指数：恒生复核材料业：服务业数据按月更新，01-01-2000至11-01-2018期间平均值为4,763.9552000年1月3日=2000，共227份观测结果。该数据的历史最高值出现于02-01-2014，达8,743.9202000年1月3日=2000，而历史最低值则出现于09-01-2001，为1,297.4502000年1月3日=2000。CEIC提供的总回报指数：恒生复核材料业：服务业数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

The rows are the responses and the columns are the hypotheses tested in the Granger causality tests stated in Eq (9) and Eq (10).

总回报指数：恒生中国企业指数在10-01-2018达19,683.9802000年1月3日=2000，相较于09-01-2018的21,386.5302000年1月3日=2000有所下降。总回报指数：恒生中国企业指数数据按月更新，01-01-2000至10-01-2018期间平均值为15,411.2302000年1月3日=2000，共226份观测结果。该数据的历史最高值出现于10-01-2007，达26,791.2012000年1月3日=2000，而历史最低值则出现于02-01-2000，为1,454.3802000年1月3日=2000。CEIC提供的总回报指数：恒生中国企业指数数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

While the global coronavirus (COVID-19) pandemic caused all major stock market indices to fall sharply in March 2020, both the extent of the decline at this time, and the shape of the subsequent recovery, have varied greatly. For example, on March 15, 2020, major European markets and traditional stocks in the United States had shed around ** percent of their value compared to January *, 2020. However, Asian markets and the NASDAQ Composite Index only shed around ** to ** percent of their value. A similar story can be seen with the post-coronavirus recovery. As of November 14, 2021 the NASDAQ composite index value was around ** percent higher than in January 2020, while most other markets were only between ** and ** percent higher. Why did the NASDAQ recover the quickest? Based in New York City, the NASDAQ is famously considered a proxy for the technology industry as many of the world’s largest technology industries choose to list there. And it just so happens that technology was the sector to perform the best during the coronavirus pandemic. Accordingly, many of the largest companies who benefitted the most from the pandemic such as Amazon, PayPal and Netflix, are listed on the NADSAQ, helping it to recover the fastest of the major stock exchanges worldwide. Which markets suffered the most? The energy sector was the worst hit by the global COVID-19 pandemic. In particular, oil companies share prices suffered large declines over 2020 as demand for oil plummeted while workers found themselves no longer needing to commute, and the tourism industry ground to a halt. In addition, overall share prices in two major stock exchanges – the London Stock Exchange (as represented by the FTSE 100 index) and Hong Kong (as represented by the Hang Seng index) – have notably recovered slower than other major exchanges. However, in both these, the underlying issue behind the slower recovery likely has more to do with political events unrelated to the coronavirus than it does with the pandemic – namely Brexit and general political unrest, respectively.

Context

Time series modelling for the prediction of stocks prices is a challenging task. Political events, market expectations and economic factors are just a few known factors that can impact financial market behaviour. The financial market is a complex, noisy, evolutionary and chaotic field of study that attracts many enthusiasts and researches — the first, usually driven by the economic benefit of it, the latter, inspired by the challenge of handling such complex data.

This project aims to predict Facebook (FB) next day stock price direction with machine learning algorithms. Technical indicators and global market indexes are used, and their influence on the forecast accuracy is analysed.

Content

Daily values were retrieved (volume, open, close, low and high prices) from Yahoo! Finance website. For Facebook (FB), July 2012 was the earliest data available. The date range is July 2012 to November 2018.

The closing price of current day C(t) and closing price from the previous day C(t-1) are compared to build the initial dataset. The objective is to define if the price trend is going up or down by analysing these two values. For each instance, a comparison was made and recorded. If the price is going up, C(t) > C(t-1), class “1” is assigned. Class “0” is assigned for the opposite case.

• ID: Sample ID
• Close: Closing value of previous day
• Low: Lowest value of previous day
• High: Highest value of previous day
• Volume: Volume value of previous day

Research was initiated to understand which features could help the model to forecast the stock direction. Three main routes were found: Lag features, Technical Indicators and Global Market Indexes. Below is an explanation of each group of features.

Lag features are features that contain the closing price and direction of previous days and it is a common strategy for Time Series models. The following features were added:

• C(t-5): Closing price of 5 days before
• C(t-4): Closing price of 4 days before
• C(t-3): Closing price of 3 days before
• C(t-2): Closing price of 2 days before
• C_up_4: Output 1 if closing price went up 4 days ago
• C_up_3: Output 1 if closing price went up 3 days ago
• C_up_2: Output 1 if closing price went up 2 days ago
• C_up_1: Output 1 if closing price went up 1 day ago

Technical indicators are used by researches and financial market analysts to support stock market trend forecasting. Common indicators retrieved from the literature were selected and calculated for Facebook stock. Techical Indicators added:

• MA-10: Moving Average considering previous 10 days
• MA-5: Moving Average considering previous 5 days
• WMA-10: Weighted Moving Average considering previous 10 days
• SO: Stochastic Oscillator
• M: Momentum as the difference in closing price in a 10 days interval
• SSO: Slow Stochastic Oscillator
• EMA: Exponential Moving Average for a 10 day period
• MACD_Sline_9: MACD Signal Line for a 9 day period
• RSI: Relative Strength Index
• CCI: Commodity Channel Index
• ADO: Accumulation Distribution Oscillator

Technical indicators provide a suggestion of the stock price movement. Additional features were created for each technical indicator by analysing its daily value and assigning a class according to their meaning. Class “1” is given if the indicator numerical value suggests upper trend, class “0” for a downtrend. In other words, financial market analysis is performed at a simplistic level, in the attempt to translate what the continuous value means.

• MA-10>C: If MA-10 is higher than Closing price output 1
• MA-5>C: If MA-5 is higher than Closing price output 1
• WMA-10>C: If WMA-10 is higher than Closing price output 1
• SO>SOt-1: Output is 1 if SO current value is higher than previous day
• M>0: A positive momentum outputs 1
• SSO>SSOt-1: SSO current value is higher than previous day
• EMA>C: If EMA is higher than Closing price output 1
• MACD>MACDt-1: If MACD current value is higher than previous day output 1
• RSI70-30: If RSI is above 70, output 0. Values below 30 output is one. For values within this range it compares to previous day and outputs 1 if value has increased
• CCI200-200: Similar to RSI, but if threshold set for 200 and -200.
• ADO>ADOt-1: Output is 1 if ADO current value is higher than previous day

For a given country or region, the stock market index characterises the performance of its financial market and the overall local economy. For this reason, the same day performance of these markets could contribute to the machine learning model predictions. Six global indexes were added as features, with their closing direction as up or down, class “1” or “0”, respectively. Data for these indexes (Nikkei, Hang Seng, All Ordinaries, Euronext 100, SSE and DAX) were also retrieved from Yahoo! Finance.

总回报指数：恒生大陆25在10-01-2018达15,795.7102000年1月3日=2000，相较于09-01-2018的17,411.5702000年1月3日=2000有所下降。总回报指数：恒生大陆25数据按月更新，01-01-2000至10-01-2018期间平均值为9,701.6042000年1月3日=2000，共226份观测结果。该数据的历史最高值出现于01-01-2018，达20,571.9902000年1月3日=2000，而历史最低值则出现于08-01-2001，为1,432.1302000年1月3日=2000。CEIC提供的总回报指数：恒生大陆25数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

总回报指数：恒生香港小型股指数在10-01-2018达3,400.9702000年1月3日=2000，相较于09-01-2018的3,773.3002000年1月3日=2000有所下降。总回报指数：恒生香港小型股指数数据按月更新，01-01-2000至10-01-2018期间平均值为3,130.6652000年1月3日=2000，共226份观测结果。该数据的历史最高值出现于05-01-2015，达5,128.5802000年1月3日=2000，而历史最低值则出现于09-01-2001，为956.8802000年1月3日=2000。CEIC提供的总回报指数：恒生香港小型股指数数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于Global Database的香港 – 表 HK.Z001：主板：股票市场指数。

总回报指数：恒生工商部在11-01-2018达36,195.65002Jan1990=2581.28，相较于10-01-2018的33,773.37002Jan1990=2581.28有所增长。总回报指数：恒生工商部数据按月更新，01-01-1990至11-01-2018期间平均值为12,351.99502Jan1990=2581.28，共347份观测结果。该数据的历史最高值出现于01-01-2018，达45,123.52002Jan1990=2581.28，而历史最低值则出现于01-01-1990，为2,454.38002Jan1990=2581.28。CEIC提供的总回报指数：恒生工商部数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

香港证券交易所：指数：总回报：Hang Seng China (Hong Kong Listed) 25 Index在04-01-2025达19,406.950NA，相较于03-01-2025的20,408.380NA有所下降。香港证券交易所：指数：总回报：Hang Seng China (Hong Kong Listed) 25 Index数据按月更新，06-01-2013至04-01-2025期间平均值为14,802.300NA，共143份观测结果。该数据的历史最高值出现于01-01-2018，达20,571.990NA，而历史最低值则出现于10-01-2022，为10,167.910NA。CEIC提供的香港证券交易所：指数：总回报：Hang Seng China (Hong Kong Listed) 25 Index数据处于定期更新的状态，数据来源于Exchange Data International Limited，数据归类于全球数据库的中国香港特别行政区 – Table HK.EDI.SE: Hong Kong Stock Exchange: Monthly。

总回报指数：恒生公用事业在10-01-2018达184,690.19002Jan1990=3393.54，相较于09-01-2018的192,392.90002Jan1990=3393.54有所下降。总回报指数：恒生公用事业数据按月更新，01-01-1990至10-01-2018期间平均值为52,341.78002Jan1990=3393.54，共346份观测结果。该数据的历史最高值出现于08-01-2018，达193,594.99002Jan1990=3393.54，而历史最低值则出现于01-01-1990，为3,382.71002Jan1990=3393.54。CEIC提供的总回报指数：恒生公用事业数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

总回报指数：恒生中国H股金融行业指数在11-01-2018达24,007.9302004年3月5日=5033.14，相较于10-01-2018的23,060.9702004年3月5日=5033.14有所增长。总回报指数：恒生中国H股金融行业指数数据按月更新，03-01-2004至11-01-2018期间平均值为17,452.5862004年3月5日=5033.14，共177份观测结果。该数据的历史最高值出现于01-01-2018，达31,324.3402004年3月5日=5033.14，而历史最低值则出现于04-01-2004，为3,837.9702004年3月5日=5033.14。CEIC提供的总回报指数：恒生中国H股金融行业指数数据处于定期更新的状态，数据来源于香港交易及結算所有限公司 (香港交易所)，数据归类于全球数据库的香港 – 表 HK.Z001：主板：股票市场指数。

香港证券交易所：指数：总回报：恒生指数在04-01-2025达77,013.580NA，相较于03-01-2025的80,237.160NA有所下降。香港证券交易所：指数：总回报：恒生指数数据按月更新，06-01-2013至04-01-2025期间平均值为66,336.910NA，共143份观测结果。该数据的历史最高值出现于01-01-2018，达88,755.560NA，而历史最低值则出现于10-01-2022，为46,543.700NA。CEIC提供的香港证券交易所：指数：总回报：恒生指数数据处于定期更新的状态，数据来源于Exchange Data International Limited，数据归类于全球数据库的中国香港特别行政区 – Table HK.EDI.SE: Hong Kong Stock Exchange: Monthly。

香港证券交易所：指数：总回报：恒生复核材料业：公用事业在04-01-2025达14,287.180NA，相较于03-01-2025的13,813.320NA有所增长。香港证券交易所：指数：总回报：恒生复核材料业：公用事业数据按月更新，06-01-2013至04-01-2025期间平均值为13,669.470NA，共143份观测结果。该数据的历史最高值出现于12-01-2021，达18,086.860NA，而历史最低值则出现于10-01-2022，为10,607.190NA。CEIC提供的香港证券交易所：指数：总回报：恒生复核材料业：公用事业数据处于定期更新的状态，数据来源于Exchange Data International Limited，数据归类于全球数据库的中国香港特别行政区 – Table HK.EDI.SE: Hong Kong Stock Exchange: Monthly。