To live through a hype cycle in the markets is a privilege. It may not feel like one at the time, but the learning opportunity is immense. Both the stocks that go parabolic and the ones that get left for dead can teach us many things.
Technology companies these days fall into two buckets. One has created generational wealth (for those who have locked in those gains), while the other has plummeted in value.
The first bucket holds companies that sell products based on humans as end users. This includes product companies like Apple and Dell, but also the classic software-as-a-service (SaaS) names that have been at the centre of much debate this year.
The SaaS names were the darlings of the pre-Covid decade, with incredible multiple expansion based on a business model that appeared to be perfect. But as usual, nothing is perfect forever. Names like Adobe have seen their valuation multiples unwind to levels not seen since the global financial crisis — a time well before the SaaS model had flexed its muscles.
The second bucket contains infrastructure-based companies that sell products designed to support the growth in cloud and AI ecosystems. This includes central processing units, memory, data centre components and even graphics processing units.
And a few of the largest technology companies in the world, such as Microsoft, have a leg in each bucket.
Thanks to the US market’s fascination with thematic ETFs, investors can see the chasm in this sector by finding suitable ETFs and examining the top holdings. In doing so, we find fascinating insights into both valuations and the relative lack of diversification in many modern ETFs.
For example, the iShares Expanded Tech-Software Sector ETF ($IGV) promises to “track an index of North American companies in the software industry, along with select media and services firms”.
A few years ago, that’s exactly where you wanted to put your money. From the start of 2020 to the peaks of 2025, this ETF more than doubled in value. But $IGV has fallen 14% over 12 months as the shine has come off many of these stocks.
If investors dig into the top holdings in this ETF, they would find 9.9% in Palo Alto Networks ($CYBR) as the biggest position. This cybersecurity stock has jumped by 73% over 12 months! Most of that return has happened in the past few months as the release of the Claude Mythos AI model gave the market an enormous fright about cyber risks.
Concerningly, this means that the largest holding in this fund is now on an earnings multiple of 268, or a price/sales multiple of 25. It may be less than 10% of the fund, but that’s still a position that is clearly trading at high-risk levels.
It also doesn’t get any better scanning the second-largest holding, with an 8.2% weighting: Palantir ($PLTR).
The “good” news for those looking for a value tilt in this sector is that Palantir is down 11% in the past 12 months. The bad news is that it is still trading on an earnings multiple of 143. If investors are hoping that the price/sales multiple looks more reasonable, not quite — it’s a casual 58!
Among the top five holdings, investors will also find CrowdStrike ($CRWD) and Oracle ($ORCL). CrowdStrike has been a huge beneficiary of the same cybersecurity concerns that boosted Palo Alto, having risen 56% over 12 months and trading on a price/sales multiple of 37. As for Oracle, its best efforts to convince the market of its AI prowess haven’t offset the pain in the share price. Even an AI-focused investor day in Vegas couldn’t do it. Oracle is down 40% over 12 months. That may finally sound like a dip worth buying, but it’s still on a price/sales ratio of six and an earnings multiple of 24.
That accounts for 31% of this ETF with these four names, none of which looks like it offers great value.
The other position in the top five is Microsoft ($MSFT), in third place with a weighting of 8%. Microsoft is now on a price/sales multiple of nine, which still sounds demanding. But because it has a highly profitable business, the earnings multiple is “only” 22.
The rest of the top 10 is a mix of SaaS, gaming, firewalls and other business models. This ETF is essentially the polony of tech. Investors will find a bit of everything if they chew for long enough.
The added challenge is that the market cap-weighted nature of the underlying index forces the ETF to own more of the “expensive” stocks and less of the “cheap” stocks during a period of market evolution.
It’s difficult to argue that $IGV is very appealing right now, so let’s keep looking.
To seek out value stocks in the tech space, it’s tempting to find an equal-weighted index that avoids the issue of the cream rising to the top. This is particularly important when today’s cream is yesterday’s spoilt milk – after all, do investors really want to own the cybersecurity names and Palantir at those multiples?
An eclectic choice would be the State Street SPDR S&P Software & Services ETF, with the ticker $XSW. This promises “unconcentrated exposure across large, mid- and small-cap stocks” with software and SaaS models. With an average earnings multiple of 16 in this fund, investors can immediately see that this is a much cheaper option than $IGV, with its earnings multiple of 35.
Within $XSW, it’s likely that investors have never heard of any of the top five names. Each of them carries a weighting of about 1%, so the concentration risk is nothing compared to that found in a market cap-weighted ETF.
The top holding is Porch Group ($PRCH), which provides B2B and consumer software solutions to businesses that interact with homebuyers during the moving process. This includes insurance and warranty revenue lines. Hardly a household name, ironically. The stock’s compound annual growth rate since the IPO in 2020 is 6.6%, with a return of 16% over the past 12 months. The price/sales multiple of 3.3 is more reasonable than those of many other names, but the earnings multiple is all over the place, as the company hasn’t been reporting profits.
Then there’s Rapid7 ($RPD), which is — wait for it — a cybersecurity company. It’s just a pity that the market doesn’t seem to like it much, as the share price has plummeted 51% in the past year. It’s up 56% in the past 30 days, though, so it hasn’t been completely ignored in the broader cybersecurity rally.
Tenable Holdings ($TENB), Netskope Inc ($NTSK) and Qualys Inc ($QLYS) round out the top five. Guess what? All three are in the cybersecurity bucket, with share price movements over the past 12 months of between -45% and +22%. This is the part where it’s worth recalling that CrowdStrike returned 56% over that period.
CrowdStrike’s outperformance is a reminder of investment writer Keith Fitz-Gerald’s mantra: “Buy the best, ignore the rest”. $XSW is the opposite of this strategy, as it brings exposure to numerous questionable names during a time when technology companies are fighting for survival.
But to be fair, $XSW is down only 4.8% in the past year, while the much flashier $IGV has lost 14%. Buying the best is fine, but only at the right price.
To drive this distinction home, investors need to look at an AI infrastructure ETF for important context. The iShares Semiconductor ETF ($SOXX) will get the job done, boasting names like Micron, AMD, Nvidia, Broadcom and Intel as the top holdings.
This ETF has soared 136% in the past year, driven by the market’s belief that seemingly insatiable demand for AI hardware will continue. The earnings multiple is an eye-watering 71.
The top five holdings represent nearly 37% of the fund, so that’s slightly less concentrated than $IGV. But there’s a critical difference here: the top names in $SOXX are all driven by the same underlying theme, so they are likely to be strongly positively correlated. That’s the opposite of achieving diversification.
The truth of it is that $SOXX is firmly a momentum play. Investors can’t put it in the quality bucket, as describing Intel as a “quality” stock would be a severe application of rose-tinted glasses.
The key feature of momentum is that it shifts.
Investors don’t know exactly when it will happen, but history has taught that a shift is more likely than not. The market is nervous, with the past month seeing a traded range of $552 to $655 based on closing prices. That’s a casual 16% drawdown, which doesn’t sound like a very convincing chart.
There might be an argument to be made for buying the best and ignoring the rest. But the word “best” needs to be carefully defined. Identifying the best names based on share price performance means investors are simply applying a momentum strategy. It works really well, right up until the point where it doesn’t.
With a long-term lens, the “best” could be the names with strong market positions and broken share prices that are trading at depressed multiples. A basket of these stocks would deliver top management teams, household names, wide user bases and a reasonable margin of safety all at once.
This argument would imply bullishness towards yesterday’s SaaS heroes such as Adobe, Salesforce and Intuit. This returns neatly back to the debate raging in global markets at the moment. If investors are going to play in a broken sector, buying the “best” does feel a lot smarter than buying “the rest”. Usually, only the strongest names survive.