Equity Premium Prediction: Are Economic and Technical Indicators Unstable?

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may IntroductionThere is a long-standing debate whether the equity premium is predictable or not.Whereas predictability seemed to be largely accepted for some time (e.g., Campbell and Shiller, 1988a,b, Fama and French, 1988, 1989, Cochrane, 2008, Goyal and Welch (2008) present strong evidence challenging the view of predictability. They show that standard economic indicators used for predicting equity returns perform poorly over time which is at least partly due to instability issues. In particular, a large share of the forecasting performance arises from the period up to the early 1970s but there is little evidence of predictability in later decades. Seen from this perspective, many earlier results in favor of predictability may be driven by specific samples but do not suggest systematic return predictability. Thus it is our prime goal to examine the stability of predictive performance that forecasting indicators deliver over time.There are two recent developments which further motivate our analysis. (iv) In order to account for instability in the forecasting process (as indicated by the Goyal and Welch-approach and break tests), we propose to neglect data from the distant past.Instead of using a fixed starting point and enlarging the sample period from there on (see Goyal and Welch, 2003, 2008), we use a fixed end-point and shorten the sample by successively shifting the initial estimation period through time; thereby we examine hundreds of overlapping sub-periods. This procedure introduces the idea of rolling windows but avoids unreliable results from a standard rolling window approach. Results confirm that economic indicators do not generate stable forecasting power (Goyal and Welch, 2008), also not at recent periods. By contrast, we show that technical indicators can forecast the equity premium even until the most recent decades.(v) Finally, we apply this rolling recursive approach to assess the stability of forecasts by utility-based metrics. More specifically, we consider a mean-variance investor who optimizes his risk-return profile depending on the predicted equity premium. Performance is determined by the certainty equivalent return and the Sharpe ratio using various risk-aversion coefficients, transaction costs and constraints on portfolio weights. We find that technical 3 indicators are able to beat alternative investment strategies in almost all relevant cases, in particular during most sub-periods. By c...

Does a lot help a lot? Forecasting stock returns with pooling strategies in a data‐rich environment

“…Although Rapach and Zhou (2013) report evidence in favor of principal component predictive regressions, the same set of predictors performs considerably worse than the historical average benchmark by considering a more recent data sample (see, e.g., Neely, Rapach, Tu, & Zhou, 2014). This behavior also seems to be evident for alternative pooling strategies (Baetje & Menkhoff, 2016). Overall, reported findings confirm results highlighted by Goyal and Welch (2008) that commonly used predictors perform unstably over time.…”

“…Interestingly, not solely the OOS performance vanishes but also in-sample evidence. Additional to forecasting regressions based on individual predictors, a declining forecast performance is even visible for several pooling strategies (Baetje & Menkhoff, 2016). Other studies, such as Bossaerts and Hillion (1999) or Goyal and Welch (2003), claim that stock return predictability was never existent.…”

Investing in the S&P 500 index: Can anything beat the buy‐and‐hold strategy?

“…In order to make our results comparable to various other studies (e.g., Baetje & Menkhoff, ; Neely et al, ), we perform the simulations on an updated data set from the Welch and Goyal () study . This data set comprises monthly data from the S&P 500 (inclusive dividends), the Treasury‐bill rate, as well as all fundamental predictor variables discussed in Section 3.3 .…”

“…Given our available data set, we follow Neely et al () as well as Baetje and Menkhoff () and use the time span from 1950:01 to 1965:12 as our initial estimation period for the fundamental predictive regression models. Our baseline out‐of‐sample period then starts in 1966:01 and ends in 2014:12.…”

Mid-Price Prediction Based on Machine Learning Methods with Technical and Quantitative Indicators

“…The interested reader can also find the implementation of ML methods with technical indicators in [19], [20], [47], [66], [92], and [97]. Technical indicators are also used by [2] for equity premium prediction in the US market, where they have been proved to be efficient in the out-of-sample period (1966 -2014). For the German bond market, authors in [4] extend the judgemental bootstrapping domain for the technical analysts' case.…”