Investors in financial markets (including stock markets) often rely on forecasts by banks and brokerage houses based on a comprehensive analysis of individual businesses. In addition to this, investors are often guided by consensus forecasts based on computed averages of the calculations and recommendations made by major professional stock market participants on a variety of securities, from high-liquidity 'blue chips' to second and third-rate stock.
When a private investor is not capable of making their own predictions on certain aspects of the economy or a specific issuer, they turn to financial analysts who review the market in an attempt to assess the potential future value of securities and advise traders on whether they should buy, hold or sell certain stock. By producing forecasts and recommendations, reputable investment houses effectively set trends and help shape investor outlook. However, their forecasts do not always hold true.
A review of Russian financial analysts' forecasts made in 2012 showed a maximum accuracy rate of just 56% in predicting the direction of stock price change.
Andrei Bronevich and Alexander Lepskiy, Professors of the HSE Faculty of Economic Sciences, Henry Penikas, Assistant Professor of the same Faculty, and student Ekaterina Kosiuk examined the quality of forecasts and recommendations made by financial analysts at Russian investment banks. The study's authors applied quantitative analysis to available data on the quality of Russian investment bankers' recommendations and forecasts concerning the prices of Russian companies' stock, with a particular focus on the accuracy of recommendations and resulting yield, and consistency (or conflict) among them.
The authors used mathematical statistics to analyse the forecast accuracy and resulting yield, and the theory of belief function to assess consistency, i. e. consensus vs. conflict across forecasters. The study was based on approximately 5,000 recommendations and forecasts made by investment banks regarding 40 securities traded in the Russian stock market between 2012 and 2014.
The findings were presented in the paper 'Analysis of Conflicts and Determinants of Accuracy of Forecasts in Russian Financial Analysts’ Recommendations'.
According to the study's authors, certain banks in Russia provide consistently accurate forecasts; these are Metropol, Renaissance Capital, Goldman Sachs, J.P. Morgan and Morgan Stanley. Over the past three years, these five banks have shown more than 50% accuracy in their forecasts, exceeding 60% in some years. Investment portfolios built according to recommendations by Renaissance Capital, Goldman Sachs and Morgan Stanley have consistently brought positive yields.
Other banks in Russia provided forecasts of varying accuracy, rating between 30% and 80% at different times, yet none of the banks in the sample performed below 50% accuracy all the time.
In addition to this, the researchers observed that the forecasting accuracy was higher for the consumer market, consistently exceeding 50%, while the yield was variable across all sectors, except the energy sector where it was consistently negative.
The study found that most Russian investment bankers produce their forecasts and recommendations independently, based on their own analyses and computations. Over the entire study period, forecasting conflicts, i.e. inconsistencies, stood at 78% across all recommendations and at 62% within individual sectors.
"Independence combined with accuracy of recommendations is an important characteristic of high-quality forecasts," according to the study's authors. A high degree of both independence and accuracy can be attributed to the use of exclusive forecasting methods or perhaps to access to insider information. According to Bronevich, Lepskiy, Kosiuk and Penikas, banks whose independent forecasts produce consistently high yields include Credit Suisse, Barclays and Finam.
In addition to this, the economists have identified four sectors – metallurgy, machinery, transportation and construction – where conflicting forecasts are associated with higher returns on investment. According to the authors, "forecasting conflicts can also indicate higher investment risk and uncertainty about the sector's prospects," yet a smart investment strategy should include such sectors, as they are more likely to be reviewed by banks producing high-conflict yet accurate forecasts.
According to the authors, Sberbank CIB showed the lowest accuracy and yield on its recommendations, as well as a fairly low degree of conflict (i.e. independence) of its forecasts.