Fearmongering – Market Data reveals if COVID-19 Crisis Will Lead to a Market Crash

A lot of financial wisdom gets dispensed in Whatsapp or Telegram groups. Some bits of knowledge may turn out to be true in general, but fail to be established in practice.

In a recent telegram discussion, a group of highly intelligent legal professionals who are friends of mine, said that;

“correlations between asset classes are trending towards one and that may bode ill for financial markets“.

Making this statement is interesting because Evidence Law often bounds lawyers to back their claims with empirical data. Otherwise, there may even be a case to argue such statements as hearsay.

Fortunately for my pals, Whatsapp is not the court and I can take upon myself to judge whether the statement is true.

First, a little background to explain what this statement means.

Why is correlation = 1 “bad”?

Market correlation gives us an idea of how assets are moving in relation to each other. In order to reduce your risk, you’ll want to diversify across assets with low degrees of correlation.

In short; retail investors diversify because correlations are less than one.

This phenomenon allows the bond asset class to mitigate the losses of the stock portfolio when equities start to fall in a recession. But, there’s a catch to this approach of diversification.

In particularly nasty recessions, correlations of all asset classes tend to edge closer towards one, effectively rendering diversification useless in mitigating the losses.

I do not have a known tool to track rolling correlations between different asset classes, so I wrote one on Python and upgraded a multi-stock analytical tool I use to produce slides for my Early Retirement Masterclass.

Correlation between REITs and STI

Let us first look at the rolling correlation between REITs and STI.

I used the Lion-Philip S-REIT ETF (Ticker: CLR) and trace the evolution of correlation figures against the STI ETF (Ticker: ES3.SI) over time:

For REITs, you can observe that between January and April 2020, correlations have indeed spiked closer to one.

However, I would not attribute much of this to COVID-19 because REITs are also getting more highly represented within the STI in recent days. Given that the S-REITs ETF also own the REITs with larger market capitalizations, this may not be a result of fear and gloominess in the markets.

Correlation between Bonds and STI

Next, let us sample the correlation of government bonds against the STI ETF.

I used data from ABF Government Bond Fund (Ticker: A35) and track correlations against the STI ETF over time.

Once you switch to the bond asset class, my friend’s reasoning becomes much weaker.

Bond-Stock correlation in Singapore spiked downwards during the COVID-19 crisis and is currently trending below zero.

Rumors of the death of asset class diversification are thus, greatly exaggerated.

Correlation between Gold and STI

Finally, we turn to gold.

I used data from the SPDR Gold ETF (Ticker: O87.SI) to generate the correlation figures against the STI ETF.

Once again, there was a visible spike during the COVID-19 crisis, but it hardly goes to one. Instead, it spikes to about 0.3 and then turns negative again.

From the evidence, it is clear that the fear about asset classes trending towards one with the equity asset class is limited to the REITs asset class. (But, do note REITs are also a form of equity.)

Based on the current data, Bonds and gold still function effectively as instruments for diversification. We will need the recession to get much worse before diversification fails as an investment strategy.

However, in my opinion, the world is at a much better footing when it comes to virus control compared to March 2020. We even see the launch of a vaccine from Russia.

With equity risk premiums at an all-time high, this may be the time to load up on investments as we edge closer to normality in 2021.