|
on Forecasting |
| By: | Wagner Piazza Gaglianone; Gustavo Silva Araujo |
| Abstract: | In this paper, we explore machine learning (ML) methods to improve inflation forecasting in Brazil. An extensive out-of-sample forecasting exercise is designed with multiple horizons, a large database of 501 series, and 50 forecasting methods, including new machine learning techniques proposed here, traditional econometric models and forecast combination methods. We also provide tools to identify the key variables to predict inflation, thus helping to open the ML black box. Despite the evidence of no universal best model, the results indicate machine learning methods can, in numerous cases, outperform traditional econometric models in terms of mean-squared error. Moreover, the results indicate the existence of nonlinearities in the inflation dynamics, which are relevant to forecast inflation. The set of top forecasts often includes forecast combinations, tree-based methods (such as random forest and xgboost), breakeven inflation, and survey-based expectations. Altogether, these findings offer a valuable contribution to macroeconomic forecasting, especially, focused on Brazilian inflation. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:bcb:wpaper:561&r= |
| By: | Yilin Qian; Ryan Thompson; Andrey L. Vasnev |
| Abstract: | Expert forecast combination -- the aggregation of individual forecasts from multiple subject-matter experts -- is a proven approach to economic forecasting. To date, research in this area has exclusively concentrated on local combination methods, which handle separate but related forecasting tasks in isolation. Yet, it has been known for over two decades in the machine learning community that global methods, which exploit task-relatedness, can improve on local methods that ignore it. Motivated by the possibility for improvement, this paper introduces a framework for globally combining expert forecasts. Through our framework, we develop global versions of several existing forecast combinations. To evaluate the efficacy of these new global forecast combinations, we conduct extensive comparisons using synthetic and real data. Our real data comparisons, which involve expert forecasts of core economic indicators in the Eurozone, are the first empirical evidence that the accuracy of global combinations of expert forecasts can surpass local combinations. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.07318&r= |
| By: | Jacob Goldin; Julian Nyarko; Justin Young |
| Abstract: | Conducting causal inference with panel data is a core challenge in social science research. Advances in forecasting methods can facilitate this task by more accurately predicting the counterfactual evolution of a treated unit had treatment not occurred. In this paper, we draw on a newly developed deep neural architecture for time series forecasting (the N-BEATS algorithm). We adapt this method from conventional time series applications by incorporating leading values of control units to predict a "synthetic" untreated version of the treated unit in the post-treatment period. We refer to the estimator derived from this method as SyNBEATS, and find that it significantly outperforms traditional two-way fixed effects and synthetic control methods across a range of settings. We also find that SyNBEATS attains comparable or more accurate performance relative to more recent panel estimation methods such as matrix completion and synthetic difference in differences. Our results highlight how advances in the forecasting literature can be harnessed to improve causal inference in panel settings. |
| Date: | 2022–08 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2208.03489&r= |
| By: | Shaswat Mohanty; Anirudh Vijay; Nandagopan Gopakumar |
| Abstract: | The evaluation of the financial markets to predict their behaviour have been attempted using a number of approaches, to make smart and profitable investment decisions. Owing to the highly non-linear trends and inter-dependencies, it is often difficult to develop a statistical approach that elucidates the market behaviour entirely. To this end, we present a long-short term memory (LSTM) based model that leverages the sequential structure of the time-series data to provide an accurate market forecast. We then develop a decision making StockBot that buys/sells stocks at the end of the day with the goal of maximizing profits. We successfully demonstrate an accurate prediction model, as a result of which our StockBot can outpace the market and can strategize for gains that are ~15 times higher than the most aggressive ETFs in the market. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.06605&r= |
| By: | Niccol\`o Ajroldi; Jacopo Diquigiovanni; Matteo Fontana; Simone Vantini |
| Abstract: | Conformal Prediction (CP) is a versatile nonparametric framework used to quantify uncertainty in prediction problems. In this work, we provide an extension of such method to the case of time series of functions defined on a bivariate domain, by proposing for the first time a distribution-free technique which can be applied to time-evolving surfaces. In order to obtain meaningful and efficient prediction regions, CP must be coupled with an accurate forecasting algorithm, for this reason, we extend the theory of autoregressive processes in Hilbert space in order to allow for functions with a bivariate domain. Given the novelty of the subject, we present estimation techniques for the Functional Autoregressive model (FAR). A simulation study is implemented, in order to investigate how different point predictors affect the resulting prediction bands. Finally, we explore benefits and limits of the proposed approach on a real dataset, collecting daily observations of Sea Level Anomalies of the Black Sea in the last twenty years. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.13656&r= |
| By: | Florian Huber; Luca Onorante; Michael Pfarrhofer |
| Abstract: | In this paper, we forecast euro area inflation and its main components using an econometric model which exploits a massive number of time series on survey expectations for the European Commission's Business and Consumer Survey. To make estimation of such a huge model tractable, we use recent advances in computational statistics to carry out posterior simulation and inference. Our findings suggest that the inclusion of a wide range of firms and consumers' opinions about future economic developments offers useful information to forecast prices and assess tail risks to inflation. These predictive improvements do not only arise from surveys related to expected inflation but also from other questions related to the general economic environment. Finally, we find that firms' expectations about the future seem to have more predictive content than consumer expectations. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.12225&r= |
| By: | Mostafa Shabani; Dat Thanh Tran; Juho Kanniainen; Alexandros Iosifidis |
| Abstract: | Deep Learning models have become dominant in tackling financial time-series analysis problems, overturning conventional machine learning and statistical methods. Most often, a model trained for one market or security cannot be directly applied to another market or security due to differences inherent in the market conditions. In addition, as the market evolves through time, it is necessary to update the existing models or train new ones when new data is made available. This scenario, which is inherent in most financial forecasting applications, naturally raises the following research question: How to efficiently adapt a pre-trained model to a new set of data while retaining performance on the old data, especially when the old data is not accessible? In this paper, we propose a method to efficiently retain the knowledge available in a neural network pre-trained on a set of securities and adapt it to achieve high performance in new ones. In our method, the prior knowledge encoded in a pre-trained neural network is maintained by keeping existing connections fixed, and this knowledge is adjusted for the new securities by a set of augmented connections, which are optimized using the new data. The auxiliary connections are constrained to be of low rank. This not only allows us to rapidly optimize for the new task but also reduces the storage and run-time complexity during the deployment phase. The efficiency of our approach is empirically validated in the stock mid-price movement prediction problem using a large-scale limit order book dataset. Experimental results show that our approach enhances prediction performance as well as reduces the overall number of network parameters. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.11577&r= |
| By: | Stefanos Bennett; Jase Clarkson |
| Abstract: | Time series prediction is often complicated by distribution shift which demands adaptive models to accommodate time-varying distributions. We frame time series prediction under distribution shift as a weighted empirical risk minimisation problem. The weighting of previous observations in the empirical risk is determined by a forgetting mechanism which controls the trade-off between the relevancy and effective sample size that is used for the estimation of the predictive model. In contrast to previous work, we propose a gradient-based learning method for the parameters of the forgetting mechanism. This speeds up optimisation and therefore allows more expressive forgetting mechanisms. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.11486&r= |
| By: | Gaetan Bakalli; St\'ephane Guerrier; Olivier Scaillet |
| Abstract: | We develop a penalized two-pass regression with time-varying factor loadings. The penalization in the first pass enforces sparsity for the time-variation drivers while also maintaining compatibility with the no-arbitrage restrictions by regularizing appropriate groups of coefficients. The second pass delivers risk premia estimates to predict equity excess returns. Our Monte Carlo results and our empirical results on a large cross-sectional data set of US individual stocks show that penalization without grouping can yield to nearly all estimated time-varying models violating the no-arbitrage restrictions. Moreover, our results demonstrate that the proposed method reduces the prediction errors compared to a penalized approach without appropriate grouping or a time-invariant factor model. |
| Date: | 2022–08 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2208.00972&r= |