Pockets of Predictability

Farmer, Leland; Schmidt, Lawrence; Timmermann, Allan

doi:10.2139/ssrn.3152386

Cited by 34 publications

(39 citation statements)

References 93 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results support, among others, the results of Paye and Timmermann (2006), Lettau andNieuwerburgh (2008), andFarmer et al (2018), where stability of (13) is also rejected. The rejection of the model with constant parameters is of particular relevance to studies such as Dangl and Halling (2012), who allow the parameters of (13) to vary over time.…”

Section: Return Predictionsupporting

confidence: 91%

Time-Varying Parameters: New Test Tailored to Applications in Finance and Macroeconomics

Davidson,

Groenborg

2018

SSRN Journal

View full text Add to dashboard Cite

Many economic theories imply a linear relationship with constant parameters between financial or macroeconomic variables. While the linear model with constant parameters is often disputed in the literature, this model specification is rarely tested. This paper proposes a new and intuitively appealing test for model specification tailored for applications in finance and macroeconomics. Importantly, the test allows for autocorrelation, which is often present in these applications. We demonstrate impressive properties of the test in a realistic simulation study and obtain important insights from empirical applications.

show abstract

Section: Return Predictionsupporting

confidence: 91%

Time-Varying Parameters: New Test Tailored to Applications in Finance and Macroeconomics

Davidson,

Groenborg

2018

SSRN Journal

View full text Add to dashboard Cite

show abstract

“…where d is the current trading day, T is the look-back length, and 1 P is the indicator function that converts any logical proposition P into a number that is 1 if the proposition is satisfied, and 0 otherwise, sgn(•) is the sign function. In other words, M DA si,T,d compares the forecasted direction (upward or downward) with the realized direction of the return and yields the probability that the forecasting model can detect the correct direction of returns for a stock s i on a given timespan T prior to the day d. We use such a component on a limited look-back period as a consequence of proven studies [51] that stocks exhibit behavior with short periods of significant returns predictability ('pockets'). These periods are interspersed with long periods with little or no evidence of return predictability.…”

Section: Ranking and Dynamic Asset Selectionmentioning

confidence: 99%

Ensembling and Dynamic Asset Selection for Risk-Controlled Statistical Arbitrage

et al. 2021

View full text Add to dashboard Cite

In recent years, machine learning algorithms have been successfully employed to leverage the potential of identifying hidden patterns of financial market behavior and, consequently, have become a land of opportunities for financial applications such as algorithmic trading. In this paper, we propose a statistical arbitrage trading strategy with two key elements: an ensemble of regression algorithms for asset return prediction, followed by a dynamic asset selection. More specifically, we construct an extremely heterogeneous ensemble ensuring model diversity by using state-of-the-art machine learning algorithms, data diversity by using a feature selection process, and method diversity by using individual models for each asset, as well models that learn cross-sectional across multiple assets. Then, their predictive results are fed into a quality assurance mechanism that prunes assets with poor forecasting performance in the previous periods. We evaluate the approach on historical data of component stocks of the S&P500 index. By performing an in-depth risk-return analysis, we show that this setup outperforms highly competitive trading strategies considered as baselines. Experimentally, we show that the dynamic asset selection enhances overall trading performance both in terms of return and risk. Moreover, the proposed approach proved to yield superior results during both financial turmoil and massive market growth periods, and it showed to have general application for any risk-balanced trading strategy aiming to exploit different asset classes.

show abstract

“…8 By using monthly holding period returns, we avoid the issues that the artificial persistence induced from using annual overlapping returns can have on inference procedures (Bauer and Hamilton, 2018). Additionally, the higher return frequency allows us to better capture short-lived dynamics in bond excess returns across economic states (Farmer et al, 2019, Gargano et al, 2019.…”

Section: Predictive Regression For Bond Returnsmentioning

confidence: 99%

“…Moreover, our low (high) activity states are, like the pockets of Farmer et al (2019), more encompassing than the recession (expansion) periods defined by the NBER. We further argue that adopting a conditional view of predictability offers a solution to the puzzling contradiction between statistical and economic evaluations of predictability (Thornton andValente, 2012, Sarno et al, 2016).…”

Section: Introductionmentioning

confidence: 95%

“…Cujean and Hasler (2017) offer a theoretical model for this observation based on time-varying disagreement among investors. Farmer, Schmidt, and Timmermann (2019) show that predictability is mainly local in time and consists of "pockets" with significant predictability. Gargano et al (2019) argue that bond predictability is stronger in recessions, but present in both states.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting Bond Return Predictability

et al. 2020

View full text Add to dashboard Cite

and participants at the Econometrics Workshop at University of Pennsylvania (2017) and research seminars at Aarhus University (2019) for many constructive comments and suggestions. Jonas N. Eriksen and Martin Thyrsgaard acknowledge support from the Danish Council of Independent Research (DFF 7024-00020B and DFF 9033-00003B). Part of this paper was previously circulated under the title "Statistical tests for equal predictive ability across multiple forecasting methods".

show abstract

Pockets of Predictability

Cited by 34 publications

References 93 publications

Time-Varying Parameters: New Test Tailored to Applications in Finance and Macroeconomics

Time-Varying Parameters: New Test Tailored to Applications in Finance and Macroeconomics

Ensembling and Dynamic Asset Selection for Risk-Controlled Statistical Arbitrage

Predicting Bond Return Predictability

Contact Info

Product

Resources

About