Should you trust a computer algorithm to run your portfolio?


Computer-generated results have not received a good press in recent weeks, particularly with the A-level results debacle. So it is no surprise that some investors are questioning whether they should be willing to trust a computer with managing their hard-earned cash? 

Applying algorithms – or predefined sets of rules – to stock market data in order to make investment decisions is nothing new. The Medallion Fund, managed out of New York by Renaissance Technologies since the 1980s, has a stellar performance record. 

Although now not open to new investors, the fund recorded gains in March this year when global stock markets all fell sharply. 

Automation: Most algorithm-based funds in the UK are easier to fathom and have the advantage that they can be cheaper than traditional funds run by investment managers

Automation: Most algorithm-based funds in the UK are easier to fathom and have the advantage that they can be cheaper than traditional funds run by investment managers

Renaissance founder James Simons, a Cold War codebreaker and maths genius, used a secret strategy to decide what to trade when setting up the fund. 

Most algorithm-based funds in the UK are easier to fathom and have the advantage that they can be cheaper – in terms of annual charges – than traditional funds run full time by investment managers. 

Martin Bamford, a financial adviser at Informed Choice, is a fan of such funds. 

He says: ‘Where a computer can cheaply and consistently apply a pre-determined set of rules to investing, that’s preferable to investing blindly in the most expensive companies in a given index, or trusting the judgment of a fund manager.’ Bamford adds: ‘Algorithm funds have a bright future.’ 

A MIX OF MAN… AND MACHINE 

Just like the creators of the A-level algorithm, those who champion algorithm funds like to position them as the ultimate mix of man and machine. 

Man sets the parameters – deciding what characteristics will make a fund fly – and then sets the machine to do the work of deciding which shares fit a set of strictly defined criteria. 

READ  Saudi oil attack/Aramco: drone prone

The result is a portfolio tilted towards certain financial characteristics, managed in a way that is a hybrid of traditional active and passive funds. Funds that use this approach might be described as ‘factor investing’, ‘smart beta’ or ‘strategic beta’. 

Some funds are ‘multi-factor’, which means they construct the portfolio using several characteristics. 

So, for example, they might set the computer parameters to look at the price volatility of individual stocks as well as the dividends they pay and their cash flow. Others are ‘single factor’, so may simply look at a stock’s dividend record when picking which shares to include. 

Many algorithm funds are set up as Exchange Traded Funds. This means they can be traded like shares. Jason Hollands, a director of wealth manager Tilney, says the funds should be viewed as ‘complex trackers’. 

He says: ‘They hold a basket of shares or bonds selected on objective criteria or factors – for example, profitability, dividends, balance sheet strength, valuations or a combination of these. 

‘Such funds therefore provide investors with a lot more control over the characteristics of companies they are exposed to, but at lower costs than paying for active management.’ 

Hollands gives the example of Invesco FTSE RAFI US 1000. This is an exchange traded fund that provides exposure to the 1,000 largest stock-market-listed US companies. 

Individual holdings are not determined by market capitalisation but a basket of other factors – including revenues, cash flow and dividends. 

Such an approach, says Hollands, can help investors avoid the dangers of market bubbles. 

This is because the fund’s algorithm skews the portfolio in favour of more conservative companies and away from stocks with high market valuations and limited revenues or cash flow. 

A similar exchange traded fund, recommended by Teodor Dilov of wealth manager Interactive Investor, is SPDR S&P Global Dividend Aristocrats. This tracks high dividend-yielding companies within the S&P index that have increased or held their dividends for at least 25 years. 

READ  After Market: TaMo falls for 7th day; YES Bank looks up; 97 stocks look all set to fall

Similar products are offered by Vanguard. It provides funds that are weighted towards companies that pay dividends, or shares that have ‘momentum’ – their prices have been rising for a while.

ARE ALGORTHM FUNDS VALUE FOR MONEY? 

Ongoing charges for these funds sit between the low fees levied on index tracking funds and the higher fees imposed on actively managed funds. 

For example, S&P Global Dividend Aristocrats has an annual charge of 0.35 per cent, compared with 0.07 per cent for the iShares tracker fund that replicates the performance of the S&P 500 Index.

Invesco FTSE RAFI 1000 charges 0.39 per cent – lower than active funds such as Terry Smith’s Fundsmith Equity which charges 1 per cent. 

Opinion is divided on the value for money offered by algorithm funds. 

Ben Willis, head of portfolio management at financial planner Chase de Vere, says they are ‘usually too expensive for what is basically a plugged-in computer programme, especially when you compare them to standard passive funds which are keenly priced’. But Informed Choice’s Bamford describes them as ‘a sensible middle ground between expensive actively managed funds and dirt cheap index trackers’. 

Hollands says the funds are not perfect. He adds: ‘They are not a panacea and they do have limitations. 

‘The algorithms are set up on historic data and – as all investors know – the past does not guarantee the future.’

HOW HAVE THEY DONE IN THE PANDEMIC? 

Like Boris Johnson’s ‘mutant’ exam algorithm, most algorithm funds have been tested by the pandemic. 

Darius McDermott, managing director at Chelsea Financial Services, says they have ‘little flexibility to adapt to changes in markets’. 

Some did well. For example, Vanguard Global Momentum Factor has registered gains of 16.2 per cent over the past six months – 26.7 per cent over the past three years. 

READ  BRP reports Q4 results

This is because momentum investing – buying shares that have already performed strongly – has very much been in vogue. 

By way of contrast, Vanguard Global Value Factor – which favours companies with strong cashflow in out-of-favour sectors – has recorded three-year losses of 8.7 per cent. 

IS IT TIME TO LET THE ALGORITHM RUN? 

Algorithm funds are relatively low cost, but if you are going to use them it is important to understand what is going on in the ‘black box’. 

Hollands says the funds work best in more established stock markets due to the widespread availability of company information on which to base the algorithm. A good starting point is the Lifestrategy range of funds from Vanguard. Here, the funds simply invest in a set mix of shares and bonds. So, investors can choose a mix of 20 per cent bonds and 80 per cent equities – or 40 per cent bonds and 60 per cent equities, depending upon appetite for risk. 

The funds have a low 0.22 per cent charge. They are run by computers that re-balance the portfolios daily. 

Vanguard’s 80 per cent equity fund, 20 per cent bond fund has made profits of 59 per cent over the past five years – 17 per cent over three year. Even in the last six months it is up 12 per cent. 

Some links in this article may be affiliate links. If you click on them we may earn a small commission. That helps us fund This Is Money, and keep it free to use. We do not write articles to promote products. We do not allow any commercial relationship to affect our editorial independence.



READ SOURCE

LEAVE A REPLY

Please enter your comment!
Please enter your name here