New perspectives on forecasting inflation in emerging market economies: An empirical assessment

Abstract

We use a broad-range set of inflation models and pseudo out-of-sample forecasts to assess their predictive ability among 14 emerging market economies (EMEs) at different horizons (1 to 12 quarters ahead) with quarterly data over the period 1980Q1-2016Q4. We find, in general, that a simple arithmetic average of the current and three previous observations (the RW-AO model) consistently outperforms its standard competitors-based on the root mean squared prediction error (RMSPE) and on the accuracy in predicting the direction of change. These include conventional models based on domestic factors, existing open-economy Phillips curve-based specifications, factor-augmented models, and time-varying parameter models. Often, the RMSPE and directional accuracy gains of the RW-AO model are shown to be statistically significant. Our results are robust to forecast combinations, intercept corrections, alternative transformations of the target variable, different lag structures, and additional tests of (conditional) predictability. We argue that the RW-AO model is successful among EMEs because it is a straightforward method to downweight later data, which is a useful strategy when there are unknown structural breaks and model misspecification.

Forecasting food price inflation during global crises

“…Most of these benchmark models are suggested by the strands of literature reviewed in Section 1. These include studies focused on inflation forecasting for both advanced and developing economies (see Duncan & Martínez‐García, 2019; Faust & Wright, 2013; Hubrich & Skudelny, 2017; Mandalinci, 2017), as well as those centered on DFPI forecasting (Gómez et al, 2012; Macias et al, 2022; Moser et al, 2007), and the DMA literature (Koop & Korobilis, 2012; Nonejad, 2021).…”

“…We refrain to extend the maximum number of lags beyond six, as this value at a monthly frequency corresponds to the inclusion of two lags at a quarterly frequency. The latter is a standard value established in the literature on inflation forecasting (see, e.g., Duncan & Martínez‐García, 2019; 2023; Faust & Wright, 2013). The rationale behind this decision rests on the potential drawbacks associated with introducing irrelevant lags and large noise due to parameter estimation.…”

“…4 Regarding the forecasting of headline and core inflation during the Global Financial Crisis, a notable exception is Hubrich and Skudelny (2017). 5 Interested readers seeking a more comprehensive understanding are encouraged to explore works such as Faust and Wright (2013), which focuses on advanced economies, and Duncan and Martínez-García (2019), which delves into emerging market and developing economies, along with the references therein. 6 Despite institutions like the National Bureau of Economic Research dating the COVID-19 pandemic recession in the US economy from February to April of 2020, we extend our analysis through June of 2022, which is the last available month in our dataset.…”

Forecasting inflation in open economies: What can a NOEM model do?

“…We report the Faust and Wright (2013) version of the Atkeson and Ohanian (2001) model in because it usually outperforms the typical random‐walk specification without drift in our sample and other studies (see Duncan & Martínez‐García, 2019). That is, our findings on the relative forecasting performance of the NOEM‐BVAR are not sensitive if we use this alternative version.…”

Forecasting Inflation in Open Economies: What Can a NOEM Model DO?

“…We report theFaust and Wright (2013) version of theAtkeson and Ohanian (2001) model in M 3 because it usually outperforms the typical random-walk speci…cation without drift in our sample and other studies (seeDuncan and Martínez-García (2019)). That is, our …ndings on the relative forecasting performance of the NOEM-BVAR are not sensitive if we use this alternative version.…”