Global Markets Continue to Recover

The global stock indices ended higher despite US-China tension, Job loss claims, etc. But the investor confidence was boosted by the partial lifting of restrictions, gradual reopening of local & global economy and a steady decline in daily new COVID- 19 deaths and infections.

The major US Indices ended positively with an average of around 3%. Financial and industrials sectors outperformed. The European equity indices show a weekly sharp gain of an average of around 5% specifically for France and Germany thanks to a proposed Euro 750 billion recovery fund by the European Commission.

The Japan, Australia, Korea, and India equity indices from the APAC region outperformed US markets with to the lifting of restrictions and the reopening of domestic economies. On the other hand, China and Hong Kong ended positively without much gain due to the dispute on China imposed new security law against Hong Kong and other border-related disputes with the neighbouring countries.

The market is expected to rally on positive sentiment and data such as the gradual reopening of domestic and international global activities in the coming weeks. The below table illustrates the weekly changes for major global indices. However, the outcome of key economic data, including the ISM PMI on Monday, Service PMI and the job rate may reverse the market expectations.

Sentiment Analysis

Sentiment Analysis

The Markets are mostly drive-by emotions which creates cyclical movement in the market. One can make a profit or generate a better Alpha if you can sense cyclical price patterns. There are several tools such as technical analysis, macroeconomic indicators, and fundamental analysis techniques to identify the patterns in the past prices and forecast the future price based on the historical data.

However, these methods are not much effective due to the lack of current day to day events. Adding news-based sentiment into the existing traditional techniques and ML/Deep Learning tools improve the forecasting power of investment model, improve the “Alpha”.

News sentiment data can be approached several ways to make better use of it. I have used daily sentiment data for this analysis which is scaled between the range of 1 to 10. The score above 5 is considered as positive and the score below 5 is considered as the negative sentiment. I have converted them into the range +5 to -5 for more convenience. I consider 14 major global Indices from three regions.

I have computed the weekly average news sentiment by averaging the past 7 days of daily sentiment data for the corresponding indices. During “normal” times, the simple “average sentiment” is a good predictor of markets movements. But in a situation like COVID-19 where the flow of news is continuously negative, we will be looking at the relative reduction or augmentation of the number of negative coverages. A relative decrease in the negative sentiment and of the number of negative news in itself can move the market into positive trends.

So I have decided to find a direction of past 7 days moving average sentiment for each index and categorized as to “positive”, “neutral” and “negative” trends based on their visual appearance and the counts of positive and negative directions of each data point for the past seven days. I am using this sentiment trend as a leading indicator to predict the future direction of each market. I strongly recommend you use this indicator as an additional overlay for your models instead of solely depends on the sentiment trend.

The below outputs showing the sentiment trends for both APAC and US indices will be bullish or remains neutral for next week, whereas the Europe Indices can see the bearish direction.

Buy & Sell Signals

The above buy/sell signals are generated based on InfoTrie’s proprietary Investment/ Trading models. The models are developed with a combination of technical, fundamental, quantitative techniques and primarily the InfoTrie’s news based sentiment score as an additional overlay.

The news sentiment score provides a significant improvement, significant “Alpha” compared to the traditional investment/trading models. For more details on the analysis, accessing complete tearsheets or white paper for the models, Subscription details for News Sentiment data and other consulting works please contact us with the information provided below.

InfoTrie Financial Solutions Pte Ltd, Singapore

Sureshkumar Ramani +65 93376035

Frederic GEORJON +33 (0)6 1304 0600


Leave a Reply

Your email address will not be published. Required fields are marked *