Skip to content

Searching for Safe-Haven Assets During the COVID-19 Pandemic: summary and key takeaways

The paper tests whether bitcoin, gold, two currency pairs, crude oil and soybean futures protected equity portfolios in the early COVID-19 crash. It monitors left-tail stability of returns and of pairwise portfolios with equity indices. Gold and soybean futures held up as safe havens; most other candidates lost their protective role.

Ji, Q.; Zhang, D.; Zhao, Y. · 2020 · International Review of Financial AnalysisSummary updated 20 November 2026
Published2020Journal article
Asset focusMulti-assetSafe haven
TypeResearchRegimes & Safe Havens
Reading time4 minthis summary

Searching for Safe-Haven Assets During the COVID-19 Pandemic at a glance

The question
Did traditional safe-haven candidates still protect equity investors during the COVID-19 market turmoil?
Dataset & period
Daily closes for MSCI indices of China, the EU and the US, plus bitcoin, gold, EUR-USD, CNY-USD, WTI crude oil futures and soybean futures. Training sample runs from August 2019; the testing (COVID-19) sample runs from 1 December 2019 to 31 March 2020. The two samples have equal length. Log returns are used.
Method
A sequential surveillance test (Hoga and Wied) learns the left-tail quantile of each return series on the training sample, then checks day by day whether it shifts in the COVID-19 sample. The same test is applied to pairwise mean-variance portfolios of an equity index and a candidate haven, using the Treasury bill rate as the risk-free rate. A cross-quantilogram (Han et al.) then measures whether equity tail losses predict tail losses in the candidate the next day, in both samples.

Key findings of Searching for Safe-Haven Assets During the COVID-19 Pandemic

  • All three equity indices showed a tail change in March 2020. US indices fell more than 7% on one day in early March.
  • Crude oil futures also showed a tail change. The paper links it to a one-day oil price drop of 24%.
  • Bitcoin, gold, CNY-USD and soybean futures showed no tail change in their own return series over the testing period at the 0.05 quantile.
  • Only pairwise portfolios built with gold or soybean futures avoided tail changes at all tested left quantiles and had positive Sharpe ratios.
  • Portfolios with bitcoin, currencies or oil did not avoid tail changes. A few bitcoin and CNY-USD portfolios offset the tail instability but had non-positive Sharpe ratios.
  • Cross-quantilogram results: bitcoin, CNY-USD and oil worked as havens in the normal period but weakened in the COVID-19 period. Gold, EUR-USD and soybeans kept or improved their haven status.

Robustness, as reported in the paper

  • Tail detection was repeated at quantiles 0.10, 0.15 and 0.20, with similar results.
  • The cross-quantilogram approach is used as a robustness check, and the paper calls its results generally consistent with the tail-monitoring results.
  • Rolling cross-quantilograms use a 21-day window; the share of no predictability rose for almost all assets in the COVID-19 period.

Caveats and limits

  • The testing sample is short (December 2019 to March 2020), and the paper notes the worst market days came at its very end, which could change the haven roles.
  • The paper did not back-test the VaR estimates and examined only the 0.05 quantile for VaR, citing limited observations.
  • Portfolios are simple mean-variance mixes with no transaction costs or out-of-sample allocation test (inferred).
  • Only three equity regions and six candidates are covered, and the explanation for soybeans is argued rather than tested (inferred).

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

Why it matters

The study shows that a haven's value depends on what drives the crisis. Here a health shock hit oil and currencies while gold and an agricultural future held up. Practitioners can see how a tail-stability test and a cross-quantilogram can be used to check a hedge in a short crisis window.

Questions about this paper

Which assets stayed safe havens during COVID-19 in this paper?

Gold and soybean futures. Portfolios with them showed no left-tail change, while bitcoin, currencies and oil portfolios generally did not avoid tail changes.

Did bitcoin work as a safe haven?

Its own tail showed no change, but its role as a haven for equity indices weakened in the COVID-19 period in the cross-quantilogram analysis, especially for the US and EU indices.

What quantile did the paper focus on?

The main tail monitoring uses the 0.05 quantile, with 0.10, 0.15 and 0.20 as checks.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.1016/j.irfa.2020.101526. 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.