Bet on war? Oil prices, stock market returns and extreme geopolitical events
Knut Nygaard (Oslo Metropolitan U.) and LQ Sørensen (Storebrand Asset Mgt.)
We show that the ability of oil price changes to predict stock returns is largely limited to five extreme geopolitical events: the invasion of Ukraine in 2022, the invasion of Iraq in 2003, the Persian Gulf War in 1990/91, the collapse of OPEC in 1986, and the Arab-Israeli War in 1973. In the counterfactual scenario where these events did not occur , the t-statistics are reduced on average by 75% compared to those reported by Driesprong, Jacobsen and Maat (2008). We also find that a market timing trading strategy based on oil price changes typically generates insignificant abnormal returns, contradicting previously published results. Our results serve as an example of how a significant predictor in a time series forecast regression is not necessarily a useful or profitable market timing signal.
Economic fundamentals and stock market valuation: an approach based on the CAPE
Maria Ludovica Drudi and Federico Nucera (Bank of Italy)
This article estimates a fair value model, based on macroeconomic fundamentals, of Shiller’s cyclically-adjusted price-to-earnings ratio (CAPE). Performing a cross-country analysis, we find that CAPE – a widely used measure for stock market valuations – is, in general, positively related to economic growth and negatively related to the real long-term interest rate and measures of economic volatility calculated using industrial production and inflation data. Empirical evidence from predictive regressions of actual stock market returns indicates that deviations of CAPE from its estimated fair value are negatively related to future stock market returns. A prediction model based on these deviations outperforms, in many cases, a model based on CAPE levels both in-sample and out-of-sample.
Stock market bubbles and the predictability of gold returns (and volatility)
David Gabauer (independent researcher), et al.
First, we use the Multi-Scale LPPLS Confidence Indicator approach to detect positive and negative bubbles in the short, medium and long term for the stock markets of the G7 and BRICS countries. We were able to detect major crashes and rallies in all 12 stock markets during the period from the 1st week of January 1973 to the 2nd week of September 2020. We also observed a similar timing of strong LPPLS indicator values (both positive and negative) in the G7 and BRICS countries, suggesting the interdependence of extreme movements in these stock markets. We have found that our bubble indicators, especially when positive and negative bubbles are considered simultaneously, can accurately predict short to medium term gold returns, as well as time varying estimates of gold return volatility to a lesser extent. .
Predicting Cryptocurrency Returns
Nilanjana Chakraborty (independent researcher)
This article studies two cryptocurrencies and finds that their prices can be estimated or predicted better than their returns because returns, being price ratios, do not always show the economic relationship that may exist between two price series. However, average returns use multiple prices in their ratios that capture the economic behavior of the price series. In addition, the forecasting performance of traditional previous return models is compared to previous average return models and the latter generally perform better in terms of root mean square error (RMSE) and average return on investment (ARoI).
Predicting stock returns
David Rapach (Atlanta Fed) and Guofu Zhou (Washington U. at St. Louis)
We review the literature on stock return forecasting, highlighting challenges faced by forecasters as well as strategies for improving return forecasting. We focus on the predictability of US equity premiums and illustrate key issues through an empirical application based on updated data. Some studies argue that, despite ample in-sample evidence for capital premium predictability, popular predictors in the literature fail to outperform the simple historical mean benchmark prediction in out-of-sample tests. Recent studies, however, provide improved forecasting strategies that offer statistically and economically significant out-of-sample gains over the historical average benchmark. These strategies, including economically motivated model restrictions, forecast combining, diffusion indices, and regime shifts, improve forecasting performance by addressing substantial model uncertainty and parameter instability surrounding the data generation process for stock returns. In addition to the US equity premium, we briefly review the out-of-sample evidence supporting the predictability of US and international stock returns. The significant evidence for the predictability of stock market returns around the world has important implications for the development of asset pricing models and investment management strategies.
Economic cycles and extreme portfolio risk forecasting
Clara Zhou (Macquarie University), et al.
We propose a new 3-step resampling approach to forecasting the tail risk of the portfolio based on the economic situation. The approach first predicts economic states using a set of macroeconomic and financial variables. We then forecast the joint allocation of multiple assets in the portfolio based on expected economic states. Finally, we calculate measures of extreme portfolio risk using the distribution of expected joint returns. This approach takes into account favorably the variation over time of the upper co-moments of the joint distribution of the returns of the assets of the portfolio and applies to large portfolios. In an out-of-sample forecasting analysis, the new approach outperforms a set of existing models widely used for forecasting extreme portfolio risk.
Time-varying risk aversion and international stock returns
Massimo Guidolin (Bocconi University), et al.
We estimate a time-varying aggregate risk aversion function using options, stock returns and macroeconomic data for a sample of 8 countries. We find that in most countries the degree of risk aversion is countercyclical. Furthermore, we show that the estimated risk aversion function predicts the monthly stock index returns up to 12 months ahead. This effect is statistically significant in panel regressions, and it survives the inclusion of additional control variables. Finally, we show that the estimated time-varying risk aversion function provides useful information to an investor aiming to time the market. An investment strategy that uses the estimated measure of time-varying risk aversion to solve a mean-variance asset allocation problem, offers significant returns.
Learn how to use R for portfolio analysis
Quantitative analysis of the investment portfolio in R:
An introduction to R for modeling portfolio risk and return
By James Picerno