Winding road through a stormy landscape, economic forecasting signs.

Local Projections vs. VARs: Which Economic Forecasting Tool Should You Trust?

Nico Varela in Business & Economy December 2025 • 4 min read.

"Uncover the surprising vulnerabilities of traditional economic models and learn how to make more reliable forecasts in an uncertain world."

In today's economy, everyone from investors to policymakers relies on forecasts to make informed decisions. But what happens when the tools we use to predict the future aren't as reliable as we think? Two popular methods, Local Projections (LPs) and Vector Autoregressions (VARs), have been at the forefront of economic forecasting, each with its perceived strengths. However, new research is uncovering some unsettling truths about their accuracy.

Local Projections (LPs) and Vector Autoregressions (VARs) are time-series analysis techniques used by economists to make forecasts. Vector Autoregressions (VARs) structural vector autoregressions (SVAR) assume that the future value of a variable depends linearly on its own past values and the past values of other variables. Local Projections (LPs) estimate the impact of a predictor on an outcome at various future time points. For instance, economists use these tools to predict everything from inflation rates to the effects of government spending.

A groundbreaking study, "Double Robustness of Local Projections and Some Unpleasant VARithmetic," is challenging long-held beliefs about the robustness of these methods. This article dives into the key findings of this study, revealing the surprising vulnerabilities of VARs and the unexpected reliability of LPs under certain conditions. Whether you're an economist, investor, or simply someone keen to understand the forces shaping our economic future, this is essential reading.

The Double-Edged Sword of VARs: High Risk, High Reward?

Winding road through a stormy landscape, economic forecasting signs.

For years, VAR models have been a go-to choice for economists due to their ability to capture the complex interdependencies within an economy. However, the recent research highlights a concerning flaw: VARs can be severely unreliable, even when the model's assumptions are only slightly off. This is a critical issue, as real-world economic models are rarely, if ever, perfectly specified.

The problem lies in how VARs handle 'misspecification' – small errors in the model's assumptions. The study reveals that even minor misspecifications, undetectable by standard statistical tests, can lead to significant undercoverage in VAR confidence intervals. This means that VARs may present a misleadingly narrow range of possible outcomes, giving a false sense of certainty.

Undercover VARs: VAR confidence intervals can be severely unreliable, even when the model's assumptions are only slightly off.
The double-edged sword of VARs: VARs are high risk, high reward. The worst-case bias is small precisely when the VAR estimator has nearly the same variance as LP.
A Word of Caution: Applied researchers must therefore be careful when selecting one method over the other.

Imagine a doctor prescribing medication based on a diagnosis with a small margin of error. If the potential side effects are severe, the doctor would proceed with caution. Similarly, economists using VARs must be aware of their potential fragility and consider the consequences of relying on potentially flawed forecasts.

Navigating the Forecast Minefield: A Call for Vigilance

The world of economic forecasting is far from perfect. Economic forecasting is a complex and ever-evolving field. By understanding the limitations of VARs and the strengths of LPs, economists and decision-makers can navigate the uncertainties of the future with greater awareness. The road to sound economic planning begins with a healthy dose of skepticism and a commitment to using the most reliable tools available.

About this Article -

This article was crafted using a human-AI hybrid and collaborative approach. AI assisted our team with initial drafting, research insights, identifying key questions, and image generation. Our human editors guided topic selection, defined the angle, structured the content, ensured factual accuracy and relevance, refined the tone, and conducted thorough editing to deliver helpful, high-quality information.See our About page for more information.

This article is based on research published under:

DOI-LINK: https://doi.org/10.48550/arXiv.2405.09509,

Title: Double Robustness Of Local Projections And Some Unpleasant Varithmetic

Subject: econ.em

Authors: José Luis Montiel Olea, Mikkel Plagborg-Møller, Eric Qian, Christian K. Wolf

Published: 15-05-2024

Everything You Need To Know

What are Local Projections (LPs) and Vector Autoregressions (VARs), and how are they used in economic forecasting?

Both Local Projections (LPs) and Vector Autoregressions (VARs) are time-series analysis techniques employed by economists for economic forecasting. Vector Autoregressions (VARs) analyze how variables depend on their own past values and the past values of other variables, creating a system of equations to predict future values. On the other hand, Local Projections (LPs) estimate the impact of a predictor on an outcome at various future time points. Economists utilize these tools to forecast various economic indicators, such as inflation rates or the effects of government spending, to inform decision-making.

Why are Vector Autoregressions (VARs) considered a 'double-edged sword' in economic forecasting?

Vector Autoregressions (VARs) are considered a 'double-edged sword' because, while they can capture complex interdependencies within an economy, they are susceptible to significant unreliability. The recent research reveals that even minor misspecifications in the model's assumptions, undetectable by standard statistical tests, can lead to significant undercoverage in VAR confidence intervals. This means VARs can present a misleadingly narrow range of possible outcomes, potentially giving a false sense of certainty. The worst-case bias is small precisely when the VAR estimator has nearly the same variance as LP, making it a high-risk, high-reward tool.

What is the significance of 'misspecification' in the context of VAR models?

Misspecification refers to errors in the assumptions of a model. In the context of Vector Autoregressions (VARs), even small misspecifications can severely impact the reliability of the model. The study highlights that these minor errors, which might not be detectable through standard statistical tests, can result in significant undercoverage in VAR confidence intervals. This means the VAR model's predictions may be far less reliable than they appear, leading to potentially flawed economic forecasts and decision-making.

How does the study 'Double Robustness of Local Projections and Some Unpleasant VARithmetic' challenge the reliability of Vector Autoregressions (VARs)?

The study 'Double Robustness of Local Projections and Some Unpleasant VARithmetic' challenges the long-held beliefs about the robustness of Vector Autoregressions (VARs) by revealing their surprising vulnerabilities. The research points out that VARs can be unreliable, especially when the model's assumptions are even slightly off. It emphasizes that minor misspecifications, which may not be identified through standard statistical checks, can significantly affect the accuracy of VARs' confidence intervals, leading to potentially misleading forecasts. This study underscores the need for caution when relying on VARs for economic predictions and calls for a greater awareness of their limitations.

What recommendations does the text provide for economists and decision-makers regarding the use of Local Projections (LPs) and Vector Autoregressions (VARs)?

The text advocates for vigilance and a healthy dose of skepticism when using both Local Projections (LPs) and Vector Autoregressions (VARs). Applied researchers must be careful when selecting one method over the other. It suggests that economists and decision-makers should be aware of the limitations of VARs and the strengths of LPs. The path to sound economic planning starts with a commitment to using the most reliable tools available and a critical assessment of the potential for error in economic models. It highlights the importance of understanding the uncertainties inherent in economic forecasting to navigate the future with greater awareness.