As stock markets crashed recently in response to widescale and unprecedented disruptions brought about as a result of the global spread of Covid-19 and increased uncertainty, the recent, brief and highly unexpected crash in gold prices was mainly driven by four factors, all of which were motivated by a need to liquidate assets to reduce risk, according to financial market data company Refinitiv.

Refinitiv metals and iron-ore global head Bernd Sischka says markets experienced severe volatility levels towards mid-March.

He adds that there was equity market liquidation towards the end of March, but with this, it was counterintuitive to see gold crash, as gold is generally considered a safety asset, with investors typically buying gold stocks in times of crisis.

However, although a range of factors exist, Sischka notes that one of the possible reasons for the falling gold price could be that risk hedge fund and asset managers use something called the “value at risk” (VaR) model, which is heavily connected to volatility and also relies on cross-asset correlation in the portfolio environment.

A possible second factor affecting gold negatively could have been the volatility in all assets – foreign exchange, equities, bonds and commodities – which all valued strongly in March, thereby leading correlation converging to one.

“This forced liquidation across all assets including gold as the VaR models indicated to reduce risk as risk parameters are beginning to be breached,” he says, adding that he thinks this factor played a key role in devaluing gold.

“A lot of risk managers were forced to cut risk across the board, and I think this is what happened. I think the move of value was evidence as to just how quickly people had to reduce risk and liquidate,” notes Sischka.

A possible third factor is risk parity fund liquidations; these were performing well, he says, with a very large allocation to bonds − more bonds than equities.

However, they also have a variety of commodity holdings such as precious metals, gold and silver. “As these funds got destroyed across the asset classes at the same time, you will see some liquidation of some of their gold holdings as people were withdrawing assets.”

The fourth factor is margin requirements – these were substantially going up as the levels of volatility “exploded”, notes Sischka. “This would have gotten people receiving margin calls and liquidating their long-term positions all the way down.”

Basically, all these factors were a result of rapid “liquidity shock”, he explains, adding that there was not enough liquidity across all the markets. Accounting for all these markets collapsing at the same time is an indication as to the systemic danger impacting on the ability of the markets to functioning properly, which unfortunately spilled over into the gold market as well, says Sischka.

He also notes that in looking at why gold crashed recently and then rapidly rebounded, it was important to remember who the players were in the gold trading market.

“[There has been a] vast increase in automation of trading during 2014 and 2018,” points out Sischka. Prior to this, the majority of trading was performed by manual traders, or humans, who interacted and traded with other humans. Trading floors were large, with up to about 4 000 people at certain points in time.

However, with the onset of digitalising trading platform came trading algorithms – automated trading protocols with trigger limits to invest and divest in commodities and other tradables.

In this regard, he highlights that during November 2016 to October 2018, gold experienced a 2.6% increase in terms of trade automation, with 67.6% of trades being generated by algorithms. This is in addition to over 50% of all trades in gold being performed by algorithms trading against algorithms, “only 11% were manual traders interacting with other manual traders”.

This creates a situation where the trading fundamentals are not always easy to tie back to price action, says Sischka.

However, he says this is something humans can do well, taking into account current news about the economy. For example, Sischka says that when something such as quantitative easing is introduced, automatically gold buying should increase.

“That’s the general narrative, but since 70% of all the trade is being undertaken by algorithms, which might not read news or know about economics, maybe all they care about is some sort of propriety indicator on the gold futures price.”

He states that it is important to remember that the majority of players are not in the camp of doing fundamental analysis.