Skip to content

The 2008 GFC and COVID-19 Pandemic: How Safe are the Safe Haven Assets?: summary and key takeaways

The paper tests whether assets that sheltered investors in the 2008 crisis did so again in the COVID-19 crash. The dollar index and the franc index were the more useful havens in the GFC, while gold and silver behaved like risky assets in many COVID settings. Safe-haven character weakened for gold, silver, the dollar and the franc.

Cheema, M.A.; Faff, R.; Szulczyk, K.R. · 2022 · International Review of Financial AnalysisSummary updated 20 November 2026
Published2022Journal article
Asset focusMulti-assetSafe haven
TypeResearchRegimes & Safe Havens
Reading time4 minthis summary

The 2008 GFC and COVID-19 Pandemic: How Safe are the Safe Haven Assets? at a glance

The question
Do assets that acted as safe havens in the 2008 Global Financial Crisis keep that role during the COVID-19 pandemic?
Dataset & period
Daily US-dollar log returns for equity indices of the ten largest economies (Canada, Brazil, Italy, India, France, UK, Germany, Japan, China, US) and for seven candidate havens: gold, silver, the US dollar index, a Swiss franc index, T-bills, T-bonds and AAA corporate bonds. The sample spans 31 December 2003 to 31 July 2020, with the GFC window opening 12 September 2008 and the COVID window 20 February 2020.
Method
Each candidate haven's daily return is regressed on a stock index return, with extra terms that capture the GFC and COVID windows. The error term follows a GJR-GARCH process. The paper adds the baseline and crisis coefficients to get a total crisis beta. It then grades the asset as a weak, intermediate or strong haven using both significance and the economic size of that beta. A second test looks at gold on days of extreme stock losses (the lowest 5% and 1% of returns).

Key findings of The 2008 GFC and COVID-19 Pandemic: How Safe are the Safe Haven Assets?

  • In the GFC, the US dollar showed intermediate safe-haven behaviour and the Swiss franc somewhat so. AAA-bonds, T-bonds, T-bills and gold were weak havens.
  • Gold fell on six of the ten worst S&P 500 days in COVID. On March 12, 2020 the S&P 500 fell 10% and gold lost 4.90%.
  • In the regressions, gold's total COVID betas were positive, significant and not small in every country. Japan was among the largest, at 0.22 in the paper's wording, and China was higher.
  • Silver was the weakest candidate. Its baseline betas ranged from 0.10 for China to 0.69 for Canada, so it did not hedge any of the ten markets.
  • In COVID, silver's total betas exceeded 0.50 for Canada, the UK, France, Germany, Japan and China. The paper says silver then carried equity-like risk.
  • AAA-grade corporate bonds stayed a generally weak haven in both crises, and strengthened a little in COVID. T-bills and T-bonds stayed weak or very weak.

Robustness, as reported in the paper

  • Using the usual 20-trading-day crisis window, or windows of 30 or 40 days, leaves the conclusions intact. Details are not shown in the paper.
  • Results for the S&P 500 worst-loss days are said to be similar for the other nine stock markets. These are not reported.
  • A separate quantile test of gold (lowest 5% and 1% of stock returns) after September 5, 2011 found weaker gold behaviour than before that date. For the US the total 1% beta moved from -0.06 to +0.05.

Caveats and limits

  • The weak, intermediate and strong cutoffs are described by the authors as arbitrarily chosen.
  • The COVID window is short and ends in July 2020, so the paper says the shelter seen for some assets holds only for now.
  • The explanation for gold's weaker role (lost trust after the 2011-2015 price decline, new hedging instruments) is suggested, not tested directly (inferred).
  • No transaction costs or portfolio-level results are reported (inferred). The paper is a preprint that has not been peer reviewed.

Items marked (inferred) are our reading of the design, not statements by the authors.

Why it matters

The results show that a haven label earned in one crisis does not carry over automatically to the next. Hedge sizing based on past gold or silver behaviour may misjudge risk in a different kind of stress. The paper points to currencies and Treasuries as more stable choices, though no asset gave strong protection.

Questions about this paper

Did gold act as a haven in COVID-19?

Not reliably. The paper finds gold's total COVID betas against stock losses were positive and significant across the sample countries, and gold lost value on several of the worst stock days.

Which assets held up best across both crises?

AAA-grade corporate bonds kept a generally weak haven role in both. The dollar and the franc were more useful in the GFC but weakened in COVID.

How does the paper grade a safe haven?

It uses the sign, significance and size of the total crisis beta to label an asset weak, intermediate or strong, instead of relying only on significance.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.1016/j.irfa.2022.102316. We link to the paper and do not host the PDF.

Follow new quant research

Download MacroFinance market data as CSV or Excel, or pull it by API, and test the ideas in these papers yourself. New summaries are added as we review them.

Create a free account
Summaries are written by MacroFinance from the published papers, in our own words, for information only. They are not investment advice and may omit detail. Papers are not hosted here: each page links to the original at its publisher.