Meet David Loh.  The man who developed AAM’s Aggregate Global Equities Fund by tapping into machine learning.

After graduating with a First Class Honours degree in Psychology from the University of Birmingham UK, in 1979, David started out his career in the Singapore public sector, working on analytics and programming.

After over a year there, he went on to complete a Master of Business Administration, and moved on to his next port of call – the world of finance.  

David Loh

Here, he spent over 25 years honing his skills and gaining valuable experience, while holding a series of increasingly senior positions in the Monetary Authority of Singapore, Singapore International Monetary Exchange, and UOB. 

Through these many years, financial analysis and risk management became part and parcel of David’s professional DNA. 

As he approached retirement, David wanted to grow his retirement savings steadily. 

After much research, he decided to invest some funds with Aggregate Asset Management. Little did he expect that this would open a new chapter in his life.

A whole different approach to investing

What drew David to AAM?  Well, their unique approach to investing.

Most fund houses practise qualitative investing, where they invest in 30 to 50 companies. Which means a high degree of volatility and risk – sudden gyrations in one or two stocks (as we have witnessed recently with Facebook) – can greatly impact the value of the fund.

AAM, on the other hand, believes in quantitative investment.

This diversified approach would spread investments across 500 to 600 stocks (currently it stands at 1500 stocks) to ensure good returns, while minimising risk. 

The points accumulation may be small at the start, but over time, the growth and profit potential were tremendous.

Above that, he was also intrigued by whether AAM could deploy machine learning in selecting winning stocks, because machine learning methods were well suited to analysing and choosing many stocks, all at once. 

So taken was David by the potential of machine learning in investment, that upon retirement, he engrossed himself in research (together with AAM) on the use of machine learning for stock selection, and ultimately switched from being a client to being a fund manager with them!  

Now, along with him, he brought vast amounts of knowledge in financial analytics, computer programming and risk management.  All of which were infused into AAM’s machine learning art of stock investing .

This method was not widely used by local fund houses, but AAM – being the fund house that zigs, while all others zag –  was a pioneer in using it.

Traditional Investing vs Machine Learning Investing

In the old days, to do financial analysis for investing, you first had to get a staff member to pour over annual reports (this can take days), before filling up the findings on a spreadsheet – all with the aim of capturing data.

And that’s only for one company?  What if you wanted more?  Well, you simply need to do much more – for every company you want to analyse!

But things can change with the use of machine learning.

Though it may have been around for a dozen of years or so (already in use in a wide variety of industries – medical, agriculture, insurance, economics, online advertising and many more), its use in stock investments can still be considered as quite recent.

What used to take days of research and data collection, can now be done in a matter of hours!

Back then, it was difficult to make accurate stock buying decisions because of insufficient data.

But now, with data being readily available from sources like Bloomberg and Standard & Poor’s, everything is at your fingertips.

And with a great number of software being developed by academia, and powerful tools within reach of AAM, a lot more can be done – affordably, quickly and accurately.

Machine Learning has charged AAM’s ability in choosing winning stocks

There are many variables (eg. PE, PB, Dividend Yield, etc.) that go into determining what are considered winning stocks.

For an individual, even by taking into account all the variables, it’s still difficult to find the right balance to come out with a winning predictive system.

On the other hand, David and team, using the power of machine learning, are able to look across multiple variables, identify certain characteristics, take data, crunch them, and find patterns – to help select stocks that will do well (mix of value stocks or growth stocks).

Machine Learning’s half the story. Humans of AAM were the other.

While machine learning affords AAM with an indispensable advantage, it is still just computer programs.

For it to realise its full potential and promise, you have to team it up with the very best of human expertise.

That’s where David and team come in.

It is this human ability and the brilliance of their judgement that determines success – choosing the right factors and preparing the right data to be fed through the machine learning process.

To be able to do that, you need experience, and AAM has lots of it, starting with Eric Kong, one of AAM’s three co-founders, and now, David.

Human expertise is critical, because if you don’t get it right, the predictions can be misleading, which can result in a disastrous outcome for all.   

Constant adjustments need to be made, before machine learning can predict the optimum results.

But even after predicting the winning stocks, David and team are never satisfied!

“ How do we determine if the results are reliable and good to go?”, asks David. 

Here is where the analyst in David gets obsessive.

What he and the team do is check the results by sending it through a gauntlet of tests.

First up, Back Testing – where they clean up the findings to see if they are reliable.

Next, is the Model Comparison stage – where they compare the results of the various independent machine learning algorithms.

And the final stage is the Meta Model – where they combine the insights of all the independent findings.

After going through all these tests, over and over again, will David and team be truly satisfied with the end result.

And the end result usually means a high probability of predicting winning stocks.

It is this combination of new world technology and old fashion human ingenuity that has made Aggregate Asset Management a formidable champion of winning stocks.

But don’t take my word for it, give them a call and see what you will learn from David and his team of marvellous machine learning minds.


For an in-depth understanding of how David and team go about using machine learning in predicting winning stocks, kindly make an appointment with us.

Featured photo by Kevin Ku on Unsplash

Aggregate Asset Management
1 Kim Seng Promenade
#13-05 Great World City East Tower
Singapore 237994
Tel: +65 6100-2267

If something we share has resonated with you, don't hesitate to contact us or request for an appointment.

Disclaimer: The information on this site is provided for informational purposes only and is not intended to be an offer to sell or the solicitation of an offer or a recommendation to buy in, any fund or security. The information in this website is not intended to nor does it constitute professional advice or services. Accessing this website or using the information in this website does not mean that the financial services that Aggregate Asset Management offers is offered or will be offered in any jurisdiction in which such services would be unlawful under relevant laws of such jurisdiction. It is the responsibility of any persons who access the information contained in this website to observe all applicable laws and regulations of such jurisdiction. Aggregate Asset Management Pte. Ltd. can only serve accredited investors as per MAS guidelines.

© 2022 Aggregate Asset Management. All rights reserved. Privacy Policy.