NMRQL Research – a fintech company co-founded by Michael Jordaan – launched South Africa’s first machine-learning powered unit trust fund in October 2017.
What made the fund unique is that it uses machine learning to drive research, analysis, and stock selection.
At the time, NMRQL said the machine-learning powered, computational investment process allows it to discover hidden, patterns in underlying big data.
Once discovered, these patterns can be exploited to forecast returns across all asset classes and markets, resulting in the steady long-term growth of capital and income.
The fund invests in a combination of assets, including money market and interest-bearing instruments, bonds, corporate debt, equity securities, property securities, preference shares, and convertible equities.
NMRQL co-founder and CEO, Tom Schlebusch said using their unbiased algorithmic-driven process takes the emotion out of investment decisions.
“We changed the investment management process from a biased, human-centric investment process to a non-emotive, unbiased algorithmic-driven process,” he said.
Schlebusch added that the system is continuously learning and adapting to changing environments.
NMRQL processes around 2 million data points each time they rebalance their portfolio.
“This could include quantitative, fundamental, economic, or technical variables to discover and exploit repeatable patterns, helping us to deliver superior returns,” he said.
Buzzwords like machine learning and AI get many people excited, but it does not automatically result in better performance than humans.
To assess the performance of NMRQL’s funds, MyBroadband asked Schlebusch for the numbers.
He said NMRQL Research now manages three unit trusts – one under its own brand and two on an advisory agreement with ABSA Alternative Asset Management.
Here is a description of each fund and their performance.
Absa Smart Alpha Defensive Fund
The Absa Smart Alpha Defensive Fund is a “multi-asset low equity” fund which can invest across multiple asset classes and can hold offshore as well as local assets.
The percentage allocation to equity as a whole is, however, limited to 40% while the maximum allocation the fund can have offshore is 30%.
This fund is therefore categorized as having a low to medium risk appetite.
NMRQL started managing this fund around 16 months ago. The fund is up around 10.5% year to date (end August 2020) and around 12.3% on a one-year basis to the end of August 2020.
This is significantly higher than the JSE Shareholder weighted index (Swix), which is down around 5% year to date and down around -0.2% on a one-year basis to the end of August 2020.
The fund carries only around 40% in equity, of which roughly half is invested in offshore equity ETFs.
“We like the Sygnia 4IR ETF and other similar tech exposures, as it gives us exposure to exponentially growing offshore technology companies,” said Schlebusch.
The fund also carries a strategic holding in gold of around 10% while the rest of the assets are invested in bonds and cash instruments.
This fund has performed well and is currently ranked around fourth in its category on a one-year basis.
The NMRQL SCI Balanced Fund
The NMRQL SCI Balanced Fund is a “multi-asset high equity” fund which invests across multiple asset classes and can hold offshore as well as local assets.
The percentage allocation to equity as a whole is, however, limited to 75%, while the maximum allocation the fund can have offshore is 30%.
This fund is therefore categorized as having a medium-to-high risk appetite.
Schlebusch said the fund protected capital extremely well during the market downturn from 19 February to 23 March 2020.
While the Swix was down around 35% over the period, this fund only lost around 18% over the same period.
The fund returned around 3.5% year to date ending August 2020, compared with the Swix which returned was down around 5% over the same period.
On a one-year basis, the fund is up around 2.95% compared with the Swix at -0.2%.
Since it started trading on 9 October 2017, the fund is down around 2.8%, compared with the Swix index, which is down around 3.4% over the same period.
ABSA Smart Alpha Equity Fund
The ABSA Smart Alpha Equity Fund is a “general equity” fund which maintains a tracking error to the JSE Shareholder Weighted Index (Swix) of between 3% to 4%.
The idea is to take small bets around the index holding, and as such the fund does not at present take exposure to offshore equity markets.
On a year-to-date basis, the fund is down around 3.1% compared with the Swix, which is down around 5% over the same period to the end of August 2020.
On a one-year basis, the fund is up around 0.54% compared with the Swix Index which is down around -0.21%.
“The hope is that we can start taking positions in offshore instruments going forward as well as have the ability to take slightly more risk on the fund,” said Schlebusch.
“Nevertheless, the fund is doing exactly what its mandate requires and we hope that the changes mentioned will help us to put more outperformance on the table.”