When I wrote my update two weeks ago, stock market prices were a sea of red. US markets were red again yesterday but in those two weeks many equity markets around the world have pushed up to new highs and easing bond yields have lifted our fixed income investments.
In the last month our Multi Asset Balanced portfolio has made a little over 2%, 4.6% in the last 3 months, 11.6% in a year and roughly 30% in the last three years. Surely things are getting overheated.
Not surprisingly, and with past bear markets etched in their minds, a few clients are asking: should we be selling down our equities, moving to cash or selling puts on the S&P 500? I will try and answer.
Starting With the AI Bubble
As ever this is the sort of gross simplification and generalisation the press latches on to. Let’s look at what is really happening.
I’m not going to call it AI, again a simplification and generalisation, but clearly IT departments around the world are now able to poke their noses much further into business operations. It is not just spreadsheets and databases any more, the IT geeks have worked out how to write emails, reply to emails, build business plans, write legal contracts, provide research, create more robotics….and so the list goes on. Many argue it’s just IT evolution, that’s correct, but it has just taken IT to a whole new level, it is here, it is a reality, and it’s not going to be stopped.
I think the way to think about the impact of this in terms of equity markets is in three separate strands:
Companies around the world know they will become uncompetitive if they don’t embrace the new IT and to an extent it is being thrust upon us whether we like it or not. There is not suddenly more business for us all to do, but to compete and survive we will need to do it more efficiently and IT can now play a significantly bigger role, our IT spending will inevitably go up.
There will be some developments that this next stage IT will deliver that will provide new sources of revenue. Driverless cabs, robots, medical advances are all being pursued – some will work, some won’t and some have no doubt yet to be thought about. Just like the internet revolution it will take time to see who the winners and losers are.
Delivering the new technology is expensive. AI as we see it and use it today, based largely on Nvidia’s GPUs, is not particularly ‘I’ at all. It is basically just having access to huge volume computations at massive speeds carried out via the cloud in massive energy hungry data centres anywhere on the planet – it is brute force without you seeing either the brute, or the force, only the result of the force. We have been used to IT getting smaller, main frames to laptops to mobile phones to watches, the new IT is currently going the other way.
The first two buckets are still largely unknown as far as markets are concerned.
On Strand 1 – will next-gen IT make, say, American Express more valuable or will it simply be used to keep its market leading position in a finite credit card market? The jury is still out.
On Strand 2 – will Cathie Wood, CEO of Ark Invest, be correct? Cathie predicts there will be 50 million driverless cabs operating on the planet by 2030, half of which will be Tesla’s and she predicts Tesla shares will go up 5-fold by 2029. To put this relatively extreme speculation into context, there are currently 3.5million Uber drivers operating on the planet.
Strand 2 is highly speculative and highly fashionable, particularly on social media and causing bubbles everywhere. On a recent field trip to Altrincham to present on pensions and IHT, our newest adviser, Harry Donoghue and I, found ourselves sharing a sauna with a bunch of lads who had just finished their scaffolding work for the day. Finding out what Harry did, the questions came thick and fast. Should we be buying gold or Nvidia? There is an old adage about the stock market that when cab drivers are giving share tips it is time to exit. I’m sure it holds good for scaffolders!
The biggest issue with Strand 2 speculation is that there are no profits or revenues yet to measure, just hope. This allows irrationality to continue unchecked until the music stops. Harry nailed it the following morning when we were speculating whether Tesla would go up or down on what were obviously going to be painful earnings. Investors don’t care about the current profits from electrical vehicles – they are dreaming about the future and believe Elon will shape it. After an initial drop Tesla shares rose.
On Strand 3 – here we can see the impact today and start to measure its effect. This week we found out that just three US companies: Meta, Microsoft and Google have between them spent around $80bn on AI investment and infrastructure in just 3 months. To put that into context, this is half as much again as the UK has spent on HS2 in the last 5 years. All the amounts spent by the three tech companies were underestimated by stock market analysts. Like HS2 the AI spend looks like being bigger and lasting longer than initial estimates.
The investment team and I looked at this yesterday, and Costa came back with this nice little table highlighting the risks today in the valuations of the Mag 7 companies.
The ratios which are all widely used by investment analysts are:
P/S price divided by sales or revenues
P/E price divided by earnings
P/FCF price divided by free cash flow, this is the cash generated after accounting for capital investment and capital expenditure which don’t immediately come off the profit and loss but are depreciated over time for capital expenditure or revalued in time for investments made.
Source: Tideway, 30/10/2025. Ratios calculated using the latest figures and latest price.
We have compared the seven companies to the wider S&P 500 and to De Lisle America one of our active US funds, which invests more in mid and small caps and has a strong discipline on valuation when investing.
The risk here, highlighted in the table, is that the Mag 7 companies’ current valuations may be more stretched than published earnings ratios suggest and which already look stretched versus historical averages.
Some of the below the line capital expenditure may become more permanent, reducing profit margins and depreciation charges are increasing, reducing cash flow. These companies (ex-Tesla) have enjoyed record profit margins, is that changing?
In the same way not all the investments will reap returns. Deploying capital at this speed carries huge risks.
The good news for investors is that the vast majority of this spend and investment is coming from retained and current profits. This is obviously a lot less risky than if they were having to borrow significantly. The Mag 7 aren’t going to go bust, the risk is in how the market perceives their value.
Those involved in the picks and shovels of AI are seeing their revenues increase. Nvidia is the daddy of all this with its cornering of the GPU market, but Microsoft, Amazon and Google all get revenues back through their cloud computing services.
This week Amazon got the benefit of this with AWS profits surprising and sending the stock up around 10%. Meta got punished as investors worry that all the spending won’t reap returns, it got knocked 10% on its earnings.
Tideway’s Investing Views and Portfolios
Here are the key views shaping our investment decisions:
We generally don’t like the valuations of the Mag 7. Right now, we think they carry considerable risks.
We won’t invest speculatively in anything whether that’s gold, bitcoin or profitless companies hoping to exploit AI years in the future in unproven businesses. This just does not fit Tideway’s conservative approach.
We do like valuations elsewhere in US markets and in the rest of the world and we are happy to invest, with discipline, in the AI supply chain to take advantage of the monster spend.
We are comfortable with fixed income valuations and future returns although we are still wary of long duration bonds given uncertainty around Government finances in the West and future inflation, which we think could surprise again on the upside.
We don’t think selling to cash or buying puts makes sense if clients are properly diversified across equities and fixed income.
All these together make us currently very anti index investing in either fixed income or equities. That may change in the future if we wanted to own the constituents of indices, but for now we would rather pay managers to help us stay away from the worst of the speculation in equities and avoid longer duration bonds.
We have underperformed the passive investments a bit in the last few weeks as longer duration bond yields have dropped a bit and Nivida, the biggest company in the world and biggest constituent of equity indices, has been on a tear, doubling in value from the April lows.
We can live with that. As you all know our portfolios have not been idle, we have been participating in the gains. Our lack of exposure to long duration bonds will help us sleep ahead of the budget. If the Chancellor should fall off her tightrope, we will be relatively cushioned to any downside in bonds. Nor are we struggling with our FOMO to pile in to US indices in equities. The music will stop one of these days and meanwhile we are getting great returns from the rest of the world.
