Are higher rates a red herring for the tech rally?

I always found the expression “moving feast” to be a puzzling analogy – what is the significance of eating a meal in different places? Is this one meal being nibbled at, or the same meal being made each time but served in different places? Why would you even bother following the meal around? 

It turns out, as no doubt readers already know, that it is not an analogy at all, but a reference to events in the Church calendar that fall on different dates each year, such as Easter. I have to admit that it’s slightly more embarrassing to discover this in middle age than it was to discover the lyrics to the hymn are “the Lord of the dance, said he”, not “the Lord of the darned settee”, as I no longer have the excuse of being eight years old. 

Anyway, after a bit of research I am no longer sure it makes sense to describe markets as a “moving feast”, but what I want to find an expression to say is that the drivers of markets are constantly changing, and relying on correlations to continue to work in the same way they have recently is a surefire way to underperform. So it is with the tech industry, the most important in stock markets today.

Technology’s latest surge

Tech has surged once again this year, with Allianz Technology Trust Ord up 29% in 2026, to 19 May 2026. The dangers in such a meteoric rally were illustrated last week with a one-day drop of 4% on Thursday, since largely recovered. These returns are far ahead of those of the MSCI ACWI/Information Technology Index, which is up 8% over the same period.

YEAR-TO-DATE PERFORMANCE

Source: Morningstar. Past performance is not a reliable indicator of future results

ATT’s manager, Mike Seidenberg, has achieved this outperformance while having about half the index weight in NVIDIA Corp, with Nvidia making a sedate 18% so far, after reporting results last week. This is not last year’s tech market, but unfortunately for the bears, it is still where the action is. ATT’s holdings in memory have helped it generate alpha, while the mid-cap focus has also been highly beneficial. Mid-cap tech stocks have outperformed after years of lagging their large and mega-cap peers.

Last week’s sell-off was reported as a response to rallying US yields, and linked to the US 30-year yield hitting highs since 2007. But this doesn’t make much sense given that US bond yields were already significantly up on the year before the sell-off, and indeed tech did best in March when yields rose too. This was also true of the five and 10-year bond yields which are probably more likely to be used in DCF models than a 30-year bond yield anyway. 

I suggest high bond yields reflected a weakening macro environment with a resumption of war on the cards and fears of this sparking inflation and potentially recession. And I think it more likely that is what is responsible for a short, sharp sell-off – a reflexive reaction to a looming negative event of unknown proportion.

Can high rates derail the rally?

High rates do increase the opportunity cost of investing, and they do have an impact on the attractiveness of other return streams. But I think it is missing the point to reduce this to analysts plugging numbers into discount rates (which is just a way to try to mathematically capture the more basic point made above) and I think it is over-simplistic to expect rates to feed through so directly to valuations. In other words, I think investors would be wrong to expect rate hikes, if they come, to stop tech’s rise.

The chart below shows the correlation between US 10-year yields and two US stock indices, the S&P 500 and the Nasdaq, going back over 15 years. If you ever wanted to illustrate the concept of “regime change”, this would do it. 

The broader S&P 500 index was positively correlated to rising bond yields for most of the period. When rates were at rock-bottom levels, it was largely a case of higher yields reflecting higher expected GDP growth, meaning investors were more optimistic on the outlook for equities. After the pandemic, this relationship changed as we entered a high interest rate regime, and the S&P 500 has tended to be negatively correlated to bond yields, albeit weakly and inconsistently. 

However, the most interesting point for our discussion is that the correlation with the tech-focused Nasdaq Composite hasn’t really changed at all: it has been effectively 0 for the whole 15 years. So tech, the supposed “long duration” industry, doesn’t seem to have been driven by bond yields and therefore rate expectations.

CORRELATION

Source: Bloomberg, Kepler calculations. Past performance is not a reliable indicator of future results

Is this surprising? Well, surely not? Over a 10-year period, Microsoft Corp has compounded EPS by 25% a year, Meta Platforms Inc Class A by 33%, Amazon.com Inc by 48%, and Nvidia by 68%. Plug 68% a year growth into your DCF and tell me how much difference a 25bps boost to the terminal discount rate makes to your valuation. 

Sarcasm aside, the maths of compounding at a high rate means that if you could grow EPS by 68% a year for five years, EPS would be 13 times higher than it is today. The valuation of companies that can achieve this sort of growth shouldn’t really be dependent on the rates outlook.

What is going to drive tech in the near future?

Strong earnings from Nvidia last week saw a decent pop in the stock, but a far greater impact on other AI beneficiaries in the hardware space, notably in Asia. Nvidia is now priced for exceptional growth, and the better money has been made in recent months looking for the winners further down the chain. Critically, it is hardware that is seeing the action. 

The below chart is truly extraordinary and shows the relative performance of software versus hardware over 20 years. Software had grown capital 15 times compared to hardware’s 10 times as of July 2025, just 10 months ago, but has now underperformed over the whole period.

RELATIVE PERFORMANCE OF SOFTWARE VERSUS HARDWARE

Source: Morningstar, Kepler calculations. Past performance is not a reliable indicator of future results

The basic reason for this is well-rehearsed: AI threatens the business models of many software companies while also massively increasing their need for capital expenditure. It’s not simply the possibility of new AI-driven software, or companies developing their own in-house alternatives, but a change to the economics of the industry, to shorter contracts and to fewer ‘seats’ per company which reduce the stickiness of revenues and the visibility of earnings.

If it is hardware that is going to make the running in the new regime, this must have some implications for market behaviour. In the pre-ChatGPT environment, software and the mega-caps reigned supreme, offering huge cash generation from capital-light business models, with intellectual property the key to their success. 

Hardware is a much more capital intensive industry, where having control of the means of production is critical to achieving the revolution: do you have the physical capacity to ramp up production and to invest in the next generation of products? As such, the economic cycle could be much more important for tech in the coming years than it has been, as revenues are likely to be linked to capex cycles rather than perpetual subscriptions or licences for services. This might mean that rates end up being more important than they have been. 

If higher rates are implemented to slow an overheating economy, we might see a slowing of investment, and vice versa. But as long as tech companies can continue to deliver absurdly high earnings growth, the impact from rising rates is going to be limited, and as long as the mega-caps gear up their businesses for the new era, and deploy the cash mountains they acquired over the previous decade and a half, there looks likely to be a strong non-cyclical impulse behind AI-related tech.

Conclusion

Those of us who remember the first appearance of Facebook know that its toxicity was obvious from the start, but that didn’t stop Meta from becoming one of the 10 largest companies in the world in about 20 years, with a market cap of over $1.5 trillion. Similarly, the downsides of AI are more than clear already, from seemingly random errors to a destruction of trust in media and images. But that doesn’t mean it isn’t going to revolutionise every single workplace in time. Investors need to vote with their heads, not with their hearts, and the tech story is far from over. Reporters and analysts need to spin short-term narratives, but the rates story is a red herring in my view. The AI rally will only be derailed when companies slow their capex investment, which is more likely to be due to a recession than a few rate hikes. 

As we leave the era of Mag7 and latterly Nvidia supremacy, the mid-cap factor and stock selection should become more important, but the opportunity remains. Whether it be in the US mid-cap space, or in the early stage venture space whereMolten Ventures Ordoperates, AI is going to continue to generate huge growth potential for companies with leverage or a niche. Diversification is good, but let’s not get too cool to invest in the obvious driver of earnings growth and cash flow for at least the next decade.