Investment Strategy

Can robots invest?

What would you get if you crossed a software geek with a financial nerd? Most likely it would be what’s known as a robo-advisor. These low cost, low hassle ways to invest offer several significant advantages. But they can fall down badly in the investments that they choose for you – at least at the current time.

The financial industry isn’t new to using computers to replace people. In fact it couldn’t exist in its current, massive, global form without vast computing power.

In the old days traders shouted at each other – either in trading pits or over the phone. Buy and sell orders were written on pieces of paper and passed to another department to process, usually involving more shouting as the counterparties tried to decipher what their traders had written. Accounts were kept in huge paper ledgers. Pencils were in high demand.

Those days are long gone in most areas of financial markets. Shouting has been replaced by the gentle tapping of keyboards and the whirring of the fans that keep the computers cool. Most trades are settled without human intervention. Paper ledgers are museum pieces.

The battleground for profitability is largely fought by software developers.

As efficiency has improved, and costs have fallen, commissions charged to customers have collapsed over time. Broking is now a low margin, high volume game. The battleground for profitability is largely fought by software developers.

This still left lots of room for financial advice, investment analysis and decision making. Processing trades may have become something like the low margin textile business. But tailoring the clothes remained highly bespoke.

However, in recent years a new challenger has emerged – the robo-advisors. The idea is to automate asset allocation and portfolio management, cutting out the cost of expensive advisors and fund managers. In turn this allows these services to charge very low fees.

The principle is a good one, at least at first glance. The customer inputs some basic details about themselves and “efficient portfolios” are constructed using the fancy mathematics of portfolio management theory.

The idea is to select asset classes that offer the optimal combination of return for a given level of risk, based on historical price data.

Low cost ETFs are used to fill the asset allocations. A portion goes to stocks of various types, another portion to bonds. The ongoing management of that portfolio is then handled by computers.

Apart from the low fees, there are three main advantages for the investor:

  • Automatic portfolio rebalancing
  • Efficient income reinvestment
  • Tax loss harvesting (for taxed portfolios)

All successful investment strategies have asset allocation at their core. The investor, with or without outside help, chooses what asset classes to invest in, and how much to put in each one.

Then they choose the individual investments to fill each asset allocation. So you may want 20% in US stocks and 20% in Japanese stocks, but you have to work out exactly which individual stocks, mutual funds or ETFs to pack in there.

Once you’ve set it up then you need to make sure the allocations stay relatively constant. They won’t have to be exact all of the time, as prices swing about each day. But the idea is to stay close to the plan.

So let’s say US stocks go up a lot and end up being 25% of the portfolio. And let’s say Japanese stocks go down and end up being 15% of the portfolio. If you haven’t changed your strategy then you’d sell the excess 5% of US stocks and use the money to buy more Japanese stocks.

Portfolio rebalancing is about taking profit from short term winners and investing more in short term losers, in the expectation that their price fortunes will be reversed in future.

This kind of approach works well, provided all the asset classes have good medium to long term prospects. Portfolio rebalancing is about taking profit from short term winners and investing more in short term losers, in the expectation that their price fortunes will be reversed in future.

Numerous studies have shown this approach to be an effective strategy – assuming allocation is made across a selection of good asset classes.

Obviously it would be a disaster if one or more of the asset classes has poor prospects. At each reallocation the investor would be throwing good money after bad, by buying even more of a dud investment. More on that later…

One problem is that keeping track of the portfolio rebalancing is a bit of a hassle. You need to review it at least once a year. But you also need to be aware of big market swings in the interim.

Using a robo advisor allows the investor to pass the monitoring on to an unemotional automaton, the computer, and never worry that the portfolio is getting out of whack with the strategy. There’s little need to follow markets at all – clearly a benefit for most people.

Then there’s the issue of income reinvestment. Stocks pay dividends, bonds pay coupon interest. For best results, and maximum profit compounding, this cash has to be reinvested as it comes in. That way it’s put to work generating more profit.

With a robo-advisor the computers do the heavy lifting. As soon as the cash comes in it’s reinvested, which means it isn’t left idle.

What’s more it’s invested in the asset classes that have become underweight, which helps with the portfolio rebalancing. By using new cash income to buy, instead of selling overweight positions for a profit, there are fewer capital gains, and therefore less tax to pay.

Another potential tax advantage of the robots is “tax loss harvesting”. Here the computers sell loss making investments and the loss is offset against realised capital gains in taxable portfolios.

This reduces the net taxable gains, since realised losses are deducted from realised profits from sales, and only the net amount is taxed. There’s nothing new about this, it’s just automated with a robo-advisor.

The sold positions are then replaced with different investments in the same or very similar assets. This way the tax loss harvesting process – which sells losers – isn’t running counter to the portfolio rebalancing – which seeks to top up losing allocations. So there’s a tax gain, although the extra trading adds transaction costs.

All of this sounds great. The computers do all the repetitive stuff that an investor should do for maximum profit, take away the hassle of having to keep an eye on it, and achieve it all for a very low fee.

But there’s a catch.

That, simply, is the problem of what assets make up the strategy. That part is still defined by the humans that programme the computers, and the choices are poor.

I took a look at two of the market leading robo-advisors in the US, Betterment and Wealthfront. Both offer a range of automated portfolios that are jammed full of US and other developed country stocks and bonds. All of those offer the prospects of poor future returns.

US and developed country stocks are expensive, and trading well above their median valuation levels by any measure. Bonds offer pathetic yields and are mostly in a bubble that’s been expanding since the 1980s.

So it doesn’t matter how slick the ongoing portfolio management algorithm is. If you put garbage investments in, you’re going to get garbage profits out.

So it doesn’t matter how slick the ongoing portfolio management algorithm is. If you put garbage investments in, you’re going to get garbage profits out.

The Betterment website spewed out a recommended portfolio of ETFs for a 44 year old, namely me. It was 42% US stocks, 37% other developed country stocks (mainly Europe and Japan), 7% US corporate and municipal bonds and 4% non-US bonds.

So with 90% allocation to stocks it got one thing right. Any long term investor should have a large allocation to stocks. Just not the ones it chose, at the current time and current prices.

With a bit of further analysis I estimate that this portfolio would return about 6.3% a year, before tax. Not too bad, but certainly not great either.

But that assumes all the developed country stocks stay at their current elevated valuations. If they revert to something consistent with historic average P/Es over the next ten years – which seems likely – my profit estimate falls to 3.7% a year.

The actual result could even be worse. If the developed country stocks perform badly then the automatic portfolio rebalancing would keep throwing more money into them, as it tops up the asset allocation.

Suddenly I’d be looking at 2-3% a year, again before tax. That’s barely enough to cover inflation, and simply not enough. This also assumes that fat corporate profit margins stay at their current elevated levels. If those fall to more average historical levels then this portfolio could offer little more than a breakeven return, which is to say a loss after inflation.

Robo-advisors are certainly appealing for the low cost and lack of hassle. But unless you can find a service that allows you to tailor the portfolio to an attractive asset mix then they aren’t worth it.

It’s absolutely essential to pick the right investments in the first place. Otherwise all the other bells and whistles are pointless.

I’ll keep doing my research. If there was a service that offered the automation benefits and low costs, but with much more flexibility with the asset classes and ETFs, then it could be interesting.

Perhaps you know of such a service, or have direct experience with robo-advisors? If so, please drop me a line at the email address below. As always I’d love to hear your feedback.

For now I’d recommend a large allocation to selected emerging market stock ETFs – where there’s value – another chunk in a gold ETF, and US dollar cash or equivalent as the rest, most likely in the form of a US T-bill ETF.

That cash is ammo for when more value becomes available, such as after a US stock market crash (tick, tock, tick, tock…). It could then be used to scoop up bargains after everyone else has panicked and quit the market.

Stay tuned OfWealthers,

Rob Marstrand

Previous ArticleNext Article
Rob is the founder of OfWealth, a service that aims to explain to private investors, in simple terms, how to maximise their investment success in world markets. Before that he spent 15 years working for investment bank UBS, the world’s largest wealth manager and stock trader with headquarters in Switzerland. During that time he was based in London, Zurich and Hong Kong and worked in many countries, especially throughout Asia. After that he was Chief Investment Strategist for the Bonner & Partners Family Office for four years, a project set up by Agora founder Bill Bonner that focuses on successful inter-generational wealth transfer and long term investment. Rob has lived in Buenos Aires, Argentina for the past eight years, which is the perfect place to learn about financial crises.