Are roboadvisors a perfect solution?
The concept of roboadvisors has become very fashionable in financial circles: they are seen as a cost-efficient way to serve smaller clients by replacing human advisors.
However, like many new things in the financial sector, they are misleading in three ways: etymologically, technically and financially, as we outline below.
The word “advisor” implies giving advice — normally, helping someone to make better decisions. Roboadvisors do not, however, provide advice. They take decisions on your behalf. They are discretionary fund managers, just like a ‘real’ fund manager, but with a glorified digital front end. Their very name is misleading.
Does this mean they are ‘robomanagers’? Many, like Nutmeg or Wealthfront, are not just robomanagers: they run investment committees that try to outwit markets via the usual processes of conjecturing the markets’ next moves, based on a set of historic, databased lead indicators. That generates transactions, and therefore additional costs, for a very hypothetical benefit. Additionally, each manager tends to favour one indicator over another when not building their own, creating strong biases.
Put simply, they are just a form of traditional, discretionary fund managers with a different client interaction (what economists call a process innovation, not a disruptive one). They add extra choice to an already fragmented market, creating more confusion. Is this what is most needed to make the sector work better in the client’s interests? How effective are they in answering the key question that keeps investors awake at night — is this the right provider offering the right risk allocation?
Roboadvisors — if we can call them that — promise customised solutions at a fraction of the cost. They all use the same technology which applies the 65-year-old Modern Portfolio Theory (MPT) developed by Markowitz, who himself said that his theory was not a good predictor of portfolio returns. Indeed, this method works on the basis that both returns and risk figures are constant, proportional and predictable. Research has shown this not to be true and leads to bad risk management.
In terms of customisation, roboadvisors simply seek to allocate everyone into any of the three or five model portfolios they have manually created (or outsourced to external consultants) according to so-called ‘level of risk appetite’. There are in fact no homogenous level of risk appetites as everyone is unique, with risk appetites varying according to their current circumstances.
Moreover, due to the methodology they follow, roboadvisers are not able to maintain a constant risk profile in actual monetary terms. Mathematics shows that anyone contributing regularly to their investment pot should take different levels of risk to keep their risk constant in monetary terms. By keeping anyone within the same so-called ‘constant percentage risk portfolio’, investors may end up exposed to higher risks in actual money terms while earning less…
Roboadvisors are therefore not technically able to offer the level of customisation that ensures risk allocation is truly personalised to each individual investor.
A favourite roboadvisor rule is to use ETFs on the basis that they are cheap and liquid, unlike active fund managers who add no value. There is no point arguing here about the statistically proven fact that many active managers are just an expensive form of passive investing. Still, managers (most roboadvisors source their products from a very narrow list of providers) and style diversifications should not be ignored. Beware statistics about comparing ETFs with active managers: their individual behaviours are prone to substantial biases induced by external policies such as quantitative easing, now renowned for distorting markets and thus leading to adverse selection — a costly risk.
More importantly however, ETFs are far more expensive than what is shown to investors. While ETFs may charge lower management fees, they indirectly increase costs in other ways, such as a liquidity premium called the ‘bid-ask spread’, which is expensive and goes largely unnoticed. This means that someone buys an ETF for a price higher than the ETF’s intrinsic value and, inversely, sells it at a price lower than what it is truly worth. It’s true that the more popular the ETF, the lower the spread, but there’s more. Governance costs are another issue — ETFs have no view on the quality of a company’s management, therefore potentially encouraging bad governance at the expense of staff, customers and, more widely, the human community. Indeed, some ETFs track ESG-compliant indices. Nonetheless, without a thorough analysis of a company’s governance practices, the chance of having a positive impact becomes increasingly limited.
Furthermore, ETFs do not remunerate investors for the risks they are exposed to, most notably:
- Liquidity risk: the more popular an ETF, the more market share it owns, meaning that it becomes a major source of volatility and asset mispricing. A major investor pulling out of an ETF may create technical turbulence at the expense of other investors.
- Counterparty risk: many European ETFs do not actually buy the underlying stocks but enter a contractual agreement, called a total return swap. The ETF does not buy the asset but receive from a bank the return and coupon of the so called index, for a fee, usually the money market rate plus a spread reflecting the quality of the asset. This means that if this bank fails, the ETF will no longer receive the synthetic replication of the underlying index performance. . Remember Lehman? Meanwhile, where does sit the cash invested in the ETF? With another entity. The bank paying the index return has hedged itself by buying the underlying assets. Or not via another swap. It becomes a complex circuit whose only rationale is for banks to simply keep on charging more and more fees, which all go undisclosed.
In case of ETFs owning the underlying stocks, this counterparty risk plays differently: the ETFs themselves lend those assets to other counterparties, usually speculative hedge funds betting against a company, an economy or the markets… If they too fail, what happens?
- Index risk: ETFs do not track markets per se, but a representation of such markets called an index. Indices are prone to many biases such as favouring the most expensive stocks (the so-called price-weighted indices that give more expensive stocks a higher weight). So by not knowing the details of the index that the ETF tracks, investors may not get the optimal exposure to a category of stocks.
So… are roboadvisers a perfect solution?
Roboadvisers in their current configuration are not a one-stop shop for advice on your overall investments and wealth. They instead appear to be a repackaged version of an old model that led to much discontent amongst investors. It is possible, even likely, that many roboadvisors fail to understand what they are exposing their clients to. They all overlook the essence of financial advice by limiting user choice to a limited number of poorly described, subjective risk profiles.
Why should investors be deprived once more of the ability to make their own decisions? This leads to another, underlying question: why pay a fee to an entity that relies on outdated technologies, of which it has questionable understanding, to make decisions on one’s behalf in a non-transparent manner?
This is not the efficiency one should expect from a technological advancement that exists to enable individuals to make better decisions on their own. The only way to avoid this inefficiency is for you to become your own investment manager — the only way to regain control over your money and investments. However, how do you overcome the challenge of not spending too much time and money researching the right risk selection? That is the million dollar question… that we, at S:YB, have been working hard to solve!