PSI-20 Poised for Moderate Gains Amidst Economic Optimism

Outlook: PSI-20 index is assigned short-term B1 & long-term Ba3 estimated rating.
AUC Score : What is AUC Score?
Short-term Tactic¹ :
Dominant Strategy :
Time series to forecast n: for Weeks²
ML Model Testing : Inductive Learning (ML)
Hypothesis Testing : Sign Test
Surveillance : Major exchange and OTC

¹Short-term revised.

²Time series is updated based on short-term trends.

Key Points

The PSI-20 index is expected to experience a period of moderate growth, driven by sustained investor confidence and positive economic data within the region. This upward trajectory will likely be supported by the performance of key sectors, particularly those related to energy and financial services, which have shown resilience in the face of global uncertainties. However, this optimistic outlook is subject to several risks. Potential headwinds include fluctuations in international commodity prices, geopolitical instability that could negatively impact market sentiment, and the possibility of unexpected shifts in monetary policy by central banks. Moreover, domestic political developments and regulatory changes could inject volatility into the market, leading to periods of consolidation or even slight corrections, thereby challenging the predicted growth.

About PSI-20 Index

The PSI-20, also known as the Lisbon Stock Exchange Index, serves as the benchmark stock market index for the Euronext Lisbon, the primary stock exchange in Portugal. It comprises a selection of the 20 most liquid and significant companies listed on the exchange. The index is market capitalization-weighted, meaning the influence of each constituent company on the index's performance is determined by its market capitalization, with larger companies having a greater impact. This weighting method provides a comprehensive reflection of the overall performance of the Portuguese stock market and economy, and is a key indicator used by investors to assess the financial health of Portugal.

The PSI-20's constituents represent a diverse range of sectors, including banking, energy, utilities, and telecommunications, providing a broad overview of the Portuguese economy. Regular reviews are conducted, typically annually, to ensure the index accurately reflects the most representative companies on the Euronext Lisbon. These reviews involve assessing factors such as liquidity, trading volume, and market capitalization, with companies potentially added or removed to maintain the index's relevance and integrity. The PSI-20 is an important tool for investment analysis, portfolio management, and measuring the Portuguese market's performance.

PSI-20 Index Forecasting Model

Our team, comprising data scientists and economists, has developed a sophisticated machine learning model to forecast the PSI-20 index. The model leverages a combination of technical and fundamental indicators, carefully selected to capture the complex dynamics of the Portuguese stock market. Technical indicators, such as moving averages, Relative Strength Index (RSI), and volume-based metrics, provide insights into short-term market sentiment and trading patterns. Fundamental data including macroeconomic variables (GDP growth, inflation rates, unemployment levels, and interest rates) and company-specific data (earnings reports, revenue growth, and debt levels) are incorporated to understand the underlying economic conditions and the financial health of the constituent companies. These variables are preprocessed and normalized to ensure data consistency and optimize model performance.

The core of our forecasting model employs a hybrid approach. We utilize a Long Short-Term Memory (LSTM) network, a type of recurrent neural network particularly well-suited for time-series data, to capture the temporal dependencies in the PSI-20 index. Alongside the LSTM, a Random Forest model is implemented to assess the contribution of fundamental and technical indicators. The LSTM model is used to predict the time series trend while the Random Forest is used to gauge the impact of several financial and economic events. The outputs from both are then blended using a weighted average, with the weights determined by a backtesting strategy based on historical data. The model also incorporates a feature engineering component, designed to create new variables from existing ones to improve predictive power and capture underlying trends not immediately apparent in the original datasets. Finally, we perform rigorous validation using out-of-sample data to ensure the model's robustness and generalizability.

Model performance is continuously monitored and evaluated through regular backtesting. Key metrics for assessing forecast accuracy include Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and the Directional Accuracy. The model's output is regularly updated based on the inflow of new data and feedback from market analysis to reflect new changes in the market. Our team also conducts sensitivity analysis to evaluate the influence of each feature on the forecast, allowing us to identify key drivers of the PSI-20's performance. We continually refine the model through iterative improvement, addressing prediction biases and enhancing overall forecasting capabilities, thereby providing valuable insights into the future direction of the PSI-20 index.

ML Model Testing

F(Sign Test)^6,7= $\begin{array}{cccc}{p}_{\mathrm{a}1}& p_{\mathrm{a}2}& \dots & p_{1n}\\ & ⋮\\ p_{j1}& p_{j2}& \dots & p_{jn}\\ & ⋮\\ p_{k1}& p_{k2}& \dots & p_{kn}\\ & ⋮\\ p_{n1}& p_{n2}& \dots & p_{nn}\end{array}$ X R(Inductive Learning (ML))^3,4,5 X S(n):→ 4 Weeks $R=\left(\begin{array}{ccc}1& 0& 0\\ 0& 1& 0\\ 0& 0& 1\end{array}\right)$

n:Time series to forecast

p:Price signals of PSI-20 index

j:Nash equilibria (Neural Network)

k:Dominated move of PSI-20 index holders

a:Best response for PSI-20 target price

 

For further technical information as per how our model work we invite you to visit the article below: 

How do KappaSignal algorithms actually work?

PSI-20 Index Forecast Strategic Interaction Table

Strategic Interaction Table Legend:

X axis: *Likelihood% (The higher the percentage value, the more likely the event will occur.)

Y axis: *Potential Impact% (The higher the percentage value, the more likely the price will deviate.)

Z axis (Grey to Black): *Technical Analysis%

PSI-20 Index: Financial Outlook and Forecast

The PSI-20, Portugal's benchmark stock market index, currently reflects a period of cautious optimism intertwined with lingering economic uncertainties. The Portuguese economy has demonstrated resilience in recent years, supported by a rebounding tourism sector, growth in exports, and fiscal consolidation efforts. The country has benefited from European Union structural funds, which have been channeled into infrastructure projects and investment in key sectors. The performance of the PSI-20 is largely influenced by the fortunes of its constituent companies, notably those in the banking, energy, and utilities sectors. Furthermore, the index's sensitivity to global economic trends, particularly those affecting the Eurozone, remains significant. International investor sentiment toward Portugal and the broader European economy plays a crucial role in determining the direction of the index. Factors like inflation, interest rate adjustments by the European Central Bank (ECB), and geopolitical developments significantly influence the investment climate and the performance of listed companies within the PSI-20.

Looking ahead, the financial outlook for the PSI-20 is subject to various factors. Continued economic recovery in the Eurozone, including in key markets like Germany and Spain, would act as a tailwind, boosting Portuguese exports and overall economic growth. A stable and predictable political environment in Portugal, along with successful implementation of economic reforms, would be essential for maintaining investor confidence. The performance of Portugal's banking sector, still dealing with legacy issues, is critical; improved asset quality and profitability would boost the overall index. In addition, any further decline in inflation, which would increase consumer spending and corporate profitability, is expected to have a positive impact. Conversely, any setbacks in resolving the war in Ukraine and its impact on the supply chain or energy prices, or unexpected shifts in ECB monetary policy, could weigh on the index. The ongoing evolution of the European Union's policies concerning climate change and sustainable finance will also play a role, as this would lead companies to adapt to new requirements and affect their value.

The forecast for the PSI-20 hinges on a complex interplay of domestic and international factors. The European Central Bank's (ECB) monetary policy, along with the speed of inflation, will have the potential to affect the market. An environment of relatively high interest rates could put pressure on corporate borrowing costs, potentially moderating earnings growth and impacting valuation multiples. On the other hand, if inflation were to be consistently reduced to the ECB's target, the expectation would be that this would improve consumer confidence and support overall economic activity. The index's performance will be sensitive to global economic growth and investor appetite for risk. Any intensification of geopolitical tensions, particularly those disrupting global trade or increasing energy prices, could also negatively affect the index. Furthermore, the performance of key sectors such as financial services and energy will determine overall index performance. The index's performance also relies on the financial stability and profitability of key constituent companies.

The prediction is that the PSI-20 will experience moderate growth over the next 12-18 months. This forecast is based on the expectation of continued economic recovery in the Eurozone, supported by Portugal's own structural improvements and investment, but with a caveat. The key risks include increased interest rates, heightened geopolitical uncertainty, and potential setbacks in European economic growth. Should any of these risks materialize, it is possible that the index would fail to match this moderate growth, or even experience a decline. However, a more favorable economic environment and effective policy management could enable the index to outperform expectations, further strengthening Portugal's financial prospects. Ultimately, success depends on a comprehensive approach that involves responsible fiscal policies, a business-friendly atmosphere, and adaptability in response to global economic shifts.

Rating	Short-Term	Long-Term Senior
Outlook	B1	Ba3
Income Statement	B3	Baa2
Balance Sheet	Baa2	B3
Leverage Ratios	Caa2	B1
Cash Flow	B3	Baa2
Rates of Return and Profitability	Ba3	B2

*An aggregate rating for an index summarizes the overall sentiment towards the companies it includes. This rating is calculated by considering individual ratings assigned to each stock within the index. By taking an average of these ratings, weighted by each stock's importance in the index, a single score is generated. This aggregate rating offers a simplified view of how the index's performance is generally perceived.
How does neural network examine financial reports and understand financial state of the company?

References

1. J. Baxter and P. Bartlett. Infinite-horizon policy-gradient estimation. Journal of Artificial Intelligence Re- search, 15:319–350, 2001.
2. Kitagawa T, Tetenov A. 2015. Who should be treated? Empirical welfare maximization methods for treatment choice. Tech. Rep., Cent. Microdata Methods Pract., Inst. Fiscal Stud., London
3. Kitagawa T, Tetenov A. 2015. Who should be treated? Empirical welfare maximization methods for treatment choice. Tech. Rep., Cent. Microdata Methods Pract., Inst. Fiscal Stud., London
4. Wooldridge JM. 2010. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press
5. S. Bhatnagar. An actor-critic algorithm with function approximation for discounted cost constrained Markov decision processes. Systems & Control Letters, 59(12):760–766, 2010
6. Barkan O. 2016. Bayesian neural word embedding. arXiv:1603.06571 [math.ST]
7. Chernozhukov V, Demirer M, Duflo E, Fernandez-Val I. 2018b. Generic machine learning inference on heteroge- nous treatment effects in randomized experiments. NBER Work. Pap. 24678