Diversify better than an ETF: optimise your portfolio for violent market crashes

Updated: May 27, 2020

We're several months into the COVID-19 crisis and, whether we anticipated it or not, the stock market remains expensive and volatile. At the time of writing, the stock market experienced one of its most intense drop in history, both in terms of speed and amplitude.

For many, this type of environment is daunting, causing many to quit to reduce their losses as much as possible. Others feed off the adrenaline of such a high-risk, high-reward situation, which can be catastrophic if not executed with a considered, well-thought-out strategy.

Making informed decisions as you invest can help to make sure you're able to make the most of this situation, reducing your losses and gaining control over your portfolio for the long term. By focusing on sustainable companies whose models, practices and figures line up, as well as a bit of machine-driven intuition, you'll be able to make investments again, little-by-little, with some peace of mind.

The VIX, also known as the fear gauge, is a useful tool for assessing market volatility. It measures so according to S&P 500 index options:

When the VIX is above 25, it suggests that the market remains volatile, as indicated by the upward curvature of the line.

Another ratio we like, this time for the pricing, is the Shiller PER: a well-known tool for evaluating stock prices.

The historical average is 17 so, as we're currently around 26, you can see we're a staggering 53.2% higher than average at the time of writing! Simply put, this shows that stocks are still expensive but getting closer to their historical average price.

The two graphs above emphasise the abnormality of the current economic situation, the silver lining is that, yes, the markets are still tumultuous but they're not quite as volatile nor expensive as they were a month ago.

Bearing this in mind, for those of us who use strategic, data-driven research into the stocks we buy, it's likely that you can start investing again little by little, but doing so in a way which means you're able to build a resilient, diverse portfolio that has much more potential in the long run, no matter what the future holds.

How did ETFs on indices perform during the crisis?

Let's have a look at some US, UK and EU indices and calculate their Maximum Drawdown (MDD), defined as follows:

The S&P500 went from $3386 on February 19 to $2237 on March 23:

That's an MDD of 34% for the S&P500 (US).

The same analysis for the FTSE100 (UK) gives an MDD of 32%. The CAC40 (EU) is showing a MDD of 33%. By averaging them, we can see that the main indices had an MDD of 33% between 19th February and 23rd March.

How would have performed a data-driven portfolio ?

Following an index will not give you the best diversification. Indices aren't balanced entities by design.

We decided to use an approach based on machine learning to diversify the portfolio better with what's known as a clustering algorithm. By defining a list of features, such as capitalisation, localisation and sector, the machine is able to automatically find groups of stocks, or clusters, that belong to the same family.

Picking the top companies (basically those with the highest return on equity or lowest debt) by cluster can result in portfolios looking like:

The average performance of such a portfolio is -25% versus -32% for the S&P500.

It would have performed way better during the market drop!


It appears ETFs aren't as interesting as they used to be, especially during difficult times. When the economy becomes highly decentralised and fragile, it's worth considering tailored diversification methods that can lead to higher returns and less risk. AI and, more specifically, clustering is a great tool for such a need. Also, there's another asset that we believe in, coupled to a good diversified portfolio. Hint:

Take care and stay safe at home. Chloe from Stockomatics Disclaimer: This is just information and is NOT given as advice or recommendations.

446 views0 comments