Artificial Intelligence Vs. Genuine Stupidity: how advisers can beat AI

There is a dual threat to financial advisers when it comes to AI; the risk of being replaced, and the risk of poor investment decisions in an AI bubble.

‍By now, most of you will have played about with Artificial Intelligence (“AI”) software, so you know how gobsmackingly revolutionary it is. But if you haven’t, the picture above is a good example: It took me 30 seconds to create using AI, when previously it would have taken weeks using an expensive squad of creatives.

Therein lies the fear: It’s going to replace a lot of jobs. If you haven’t already, then, when you do get to play with AI, I bet you’ll be gripped by the same fear as the rest of us: “Holy smoke – will I even have a job in five years’ time?”

The honest answer to that question is “I don’t know”. But we can at least look back at what, and who, survived previous waves of disruption. In doing so we can guess who might survive the onslaught, and how we can be one of them.

I’ll have a stab at that through a financial advice lens. Clearly not concerning the deeper technicalities of your profession, because what the hell do I know? But the high-level stuff that will probably apply across most white-collar services.

Firstly, what’s instantly clear is that AI is already capable of giving something that at least sounds like good financial advice. Whether it’s actually good financial advice is another matter.

Here I have to admit that, faced with a choice between an artificially intelligent robot and a misguided human, I’m on Team Cyborg. I suspect AI could already give less-bad financial advice than the worst decile of financial advisers, and that alone – the avoidance of the worst potential outcome – will be a powerful driver in favour of the adviserbot.

In some ways this is a good thing, as the new AI-powered advisers might finally fill the advice gap. But, on account of being cheap, convenient, and better than the least skilful human advisers, they will likely take a chunk out of the current advice industry too. So how might you stay the right side of that?

Building human relationships

The most obvious advantage is personal relationships. AI can give impressive-sounding answers, but for stuff that really matters, like health or money, it’ll be a while yet before we’re comfortable not being advised, or at least sense-checked, by a trusted human. Any acts to strengthen your client relationships, therefore, are bricks in your defensive wall.

Clients value in-person interactions

Naturally then, this throws the emphasis back to face-to-face, non-virtual meetings. Inefficient? Time consuming? Maybe, but it’s the big edge you have over a digital app, so I’d be majoring on them. Get some nice biscuits in too.

Likewise, you can take a leaf out of the music industry’s book, which faced its disruptive nemesis -streaming - more than a decade back. What had previously been the product – their songs (in CD/tape/record form) - was commoditised and sidelined, with the real money now earned through live events.

Obviously I’m not suggesting you belt out Livin’ on a Prayer in a client review. Not in 99% of cases anyway. But I’ve worked with a few advisers who run regular seminars, and their clients see this as a valuable thing. Not just from the insights and reassurance they gain, but from the chance to mingle with peers over wine and cheese. So much so that I suspect some stick with their advice firm purely because they don’t want to miss the annual shindig.

What could future AI investment advice look like?

On this front I don’t think investment product choice will veer much from its current course. This is because – I speculate – AI advice will be driven by a couple of factors: Firstly, it will scrape the web for what the advisosphere, on average, thinks clients should be investing in. I’d suggest, uncontroversially, that this is the vanillamost¹ of passive market trackers, probably in some variation of a 60/40 equity/bond split (with that split depending on age factors and risk tolerance).

Secondly, it will most likely be hosted by private companies, so it will want to cover itself legally. This too will push it towards the largest market-tracking products, based on the time-tested principle of ‘nobody ever got sacked (or sued) for buying IBM (or Vanguard)’.

This is worth considering, because if your offering already looks like the kind of thing a half-decent AI would recommend - but at a fraction of your charge - clients will begin to question what they’re paying for. None more so than higher net-worth clients. If they twig that they’re getting the same advice and products as the entry-level clients (who were previously lost in the advice gap), I doubt they’ll be pleased (many things will perish in the forthcoming AI world, but snobbery won’t be one of them).²

So, now is the time to think about all the aspects of your offering. Can you find something a cut above the basic tracker, but that retains the reliability and reasonable charges that make them such a popular choice?

You can smell a sales pitch a mile off, so I’ll save you the detail here. But naturally I think Downing Fox slots neatly into that role. You’ll be the judge of that though, so let’s move on.

What are the investment implications of AI?

I say all this to provoke a little thought, but also to show I’m no AI denier: I believe artificial intelligence will be huge. It will change our world every bit as much as other recent tech developments, most obviously the internet.

But I also believe this means that, like the arrival of the internet (and railways, cryptocurrencies, canals, tenpin bowling etc.), most people who try to make money from this theme will, instead, end up losing from it. Especially retail investors.

Why?

Because that’s what always happens. It’s the nature of a bubble. There is money to be made, but it gets made quickly by the insiders and early-bird professional investors. By the time retail investors catch on, waves of other investors have already got excited about the obvious ‘winners’, bidding the prices up, thereby eating all of their future gains through enormous price rises.

Price rises which, crucially, have already happened, as the chart below illustrates. The late 90s spike in Intel’s share price was when retail investors really started to get involved, and you can see that, from the point of peak excitement in 2000, they are still – 24 years later – still sitting on a loss³ despite being right on their reason for investing: That the internet would change the world.

This chart also shows another problem with AI speculation: We don’t know who the big winners will be yet. Back in 1999 Google was still a fledgling start-up that wasn’t listed on the stock market. You would have been exceptionally unlikely to pick it out of the many other start ups at this time, and even if you did you wouldn’t have been able to invest as it was a private company. Will Nvidia be the big beneficiary of AI? Or will that be a company that doesn’t yet exist? We won’t know until we know.

If it is a bubble, when will it burst?

When people ask me if the recent AI bubble has started to burst⁴, what stops me saying anything more certain than ‘I hope so’ is that too many industry outsiders are as yet insufficiently amazed/terrified about what AI can do. It’s the absence of this cabbieflag (the warning provided by a taxi driver when he says you’d be mad not to invest in the latest big thing) that makes me think there’s a puff or two yet to be blown into this particular balloon.

But this level of insight is to investing what a piece of seaweed is to weather forecasting. The truth is this maybe a bubble that has popped, or a bubble that’s still expanding. Or maybe it’s not a bubble, but will become one soon. The only thing you can rely on with any certainty here is Sod’s Law: The moment you decide to pile in will be the very worst time to do so. This is why I’m not piling in, professionally or otherwise.

Right, enough with the is-it-isn’t-it, let’s assume you’re onside, and you’re not going to punt your clients’ entire life savings headlong into the AI frenzy. Just keep calm and carry on, and all will be well, right?

Well, not necessarily. The trouble is that, just as in 1999 with internet stocks, the AI theme has become so huge that it’s started to influence the market itself. Nvidia briefly became the largest stock in the world in June, and other index giants like Meta, Microsoft and Alphabet also seem correlated with the waves of speculation that drove Nvidia higher. These mega caps, and others like them, now make up a significant part of a US or global market tracker.

So even if you don’t actively take part in the AI frenzy, you may end up doing so passively. Not to the same extent of course, but by enough – if it does turn out to be a 90s-level bubble – to put a nasty dent in your clients’ wealth. In the noughties, the tech bubble bursting left pure passive equity investors nursing annual portfolio valuations that, for the next six years, were lower than they’d been before (it took about seven years for the global index to reclaim its early 2000 peak, as so many of its largest constituents, like Cisco, Intel or Microsoft, had been overextended by the internet frenzy, and then walloped as it reversed).

So what can you do?: Navigating the AI revolution

To witness the hand-wringing from some asset allocators, you’d think this is a hard call. They complain that equities are the best asset for beating inflation, but that stock markets are expensive, so they’re stuck between a rock and a hard place.

Well, equity markets may be expensive, but most individual equities are not. The markets only look expensive because of the valuations of their largest components. So the answer seems simple to me – hold equities, but avoid the dangerously expensive ones. Even if they’re large, brilliant companies that look like they’ll win from AI.

To do that you’ll need to get a little active (as in ‘selective’) though, which is an increasingly unusual thing to do. Maybe that’s swapping your market-weighted tracker for an equally-weighted or small-cap version. Or, and this is obviously my preferred method, by paying a little more for sensible active managers to move you into attractively valued stocks.

On this front, I have rarely heard the active managers we hold more excited about the prospects for their portfolios: They are able to pick up fantastic companies, operating in all areas of the economy, at attractive valuations. Their prospects for the next seven years of returns are, in my opinion, bright.

The tricky bit is that, as long as the AI bubble grows, they will continue to look like tortoises, particularly compared to tech-heavy funds (which increasingly includes market trackers, in content if not in name). You could, of course, try to ‘time’ your jump from passive to active, or tech to non-tech. But you should know that investors’ record here tends towards the disastrous. Glancing at the Intel chart above, it might look easy, but then - after the event - so does picking last Saturday’s Lotto numbers.

I am resigned to hold this position. This, if nothing else, brings peace of mind from the agony of deciding. So, by design, are the Downing Fox funds (in which my own capital is invested). So I will wait, picking up what have so far been perfectly respectable returns as I do so.

And in the meantime, I’ll be listening closely on cab rides.

Good luck out there,

Simon Evan-Cook

Fund Manager, Downing Fox‍

Keep up to date with the latest Downing Fox news by registering for our newsletter.

References

1. Using ridiculous made-up words like ‘advisosphere’ and ‘vanillamost’ are just one of the ways I seek to reassure you that I, Simon the human, wrote this, not ChatGPT. (Irrelevant, tangential footnotes are another.)
2. Would you stick with Sainsbury’s if you’d been buying their ‘Taste the Difference’ range for years only to discover it was their ‘Basics’ range with a different label slapped on it? (I’m not suggesting it is by the way – it’s just a thought experiment).
3. Although they’re probably not sitting on it anymore, chances are they sold near its near 90% trough in 2003.
4. It’s debateable whether AI stocks like NVIDIA are in bubble territory. Some experts think they are, some think they aren’t. In truth, you only know for certain well after the event. From my own perspective, I have certainly seen some bubblicious behaviour, but I’m not a stock analyst - and never have been - so my opinion should carry little weight. NB – I’m writing this in early August, at a time when NVIDIA and other large tech stocks have fallen fairly sharply, but not especially deeply.

‍Opinions expressed represent the views of the fund manager at the time of publication, are subject to change, and should not be interpreted as investment advice.

Important notice: this document has been prepared for professional investors and has been approved as a financial promotion in line with Section 21 of the FSMA by Downing LLP (“Downing”). Capital is at risk and investors should note that their investments and the income derived from them can fall as well as rise and investors may not get back the full amount invested. Past performance is not a reliable indicator of future performance. Any subscription to the fund should be made on the basis of the relevant product literature available from Downing or from the ACD, Valu-Trac; and your attention is drawn to the charges and risk factors contained therein. Downing does not offer investment or tax advice or make recommendations regarding investments. Downing is a trading name of Downing LLP. Downing LLP is authorised and regulated by the Financial Conduct Authority (Firm Reference No. 545025). Registered in England and Wales (No. OC341575). Registered Office: 3rd Floor 10 Lower Thames Street London EC3R 6EN