Source: Smart Investors
Fundsmith’s semi-annual investor meeting, London time, afternoon of July 15.
In his opening remarks, founder Terry Smith recounted how others had mocked him. One media outlet suggested he might have lost his mind and could even become the next Neil Woodford—the once-celebrated UK active fund manager who later became a cautionary tale in the industry.
“After this presentation concludes, you can judge for yourselves whether they’re right.”
Having attended many of Terry Smith’s client meetings and interviews over the years, I found this one particularly distinctive: an unusually candid conversation from a legendary fund manager navigating a period of severe headwinds, reflecting openly on introspection, adaptation, and perseverance.
Over the past six months, the fund manager renowned for his mantra—“buy good businesses, don’t overpay, and then do nothing”—replaced nearly half of his portfolio holdings. LVMH, Zoetis, Coloplast, Unilever, Novo-Nordisk A/S, Nike, Intuit… the list of sales was extensive. According to the semi-annual letter, the fund’s turnover ratio reached 51.8% in the first half of the year, with trading costs amounting to approximately £11.4 million—more than double the level seen in 2020.
As he put it himself, the fund’s performance “remains painfully divergent” from its benchmark, the MSCI World Index.
(This “pain” is quantifiable. In the first half of 2026, the fund’s net asset value declined by 2.9%, underperforming the GBP-denominated MSCI World Index by 14.1 percentage points—a fifth consecutive year of underperformance. This is the context in which the meeting took place: on one side, a 15-year annualized return of 13.8%; on the other, five straight years of lagging the benchmark.)
The catalyst for these changes was an internal review conducted without any off-limits topics. The team seriously explored directly replicating the top ten holdings of the index and even considered hiring an external manager skilled in momentum investing to manage part of the portfolio.
The review ultimately pointed squarely at himself: “One of the factors behind our actual performance has been me.”
More specifically, his adherence to the principle of “doing nothing” had been taken too far, morphing into portfolio inertia and endowment bias—continuing to hold positions simply because they were already owned, repeatedly rationalizing deteriorating fundamentals.
However, the correction did not breach the bottom line.NVIDIA (NVDA.US)、 $Apple (AAPL.US)$ 、$Amazon (AMZN.US)$、$SK Hynix (SKHY.US)$、$Broadcom (AVGO.US)$, all of which are within Fundsmith’s investable equity universe; purchasing them would eliminate the deviation from the benchmark index and alleviate most of the criticism.
Terry Smith did not buy them, saying, “There is only one scenario in which we would consider purchasing these companies—more precisely, not if, but when a disaster strikes, and amidst the wreckage, we discover a gem-like opportunity in one of them.”
That was his assessment half a month ago. In hindsight, at least those investors who heeded his words that day would now be quietly giving him credit.
Smith repeatedly emphasized that what was added this time was fundamental momentum, not price momentum. The adjusted portfolio has a return on capital of 31%,free cash flowyield of 4.3%, significantly higher than$S&P 500 Index (.SPX.US)$the 2.7%. The three core pillars remain unchanged; what has changed is the rigor of execution.
Fundsmith has approximately 50,000 direct individual investors. Smith stated that it would be irresponsible not to explain to them what is happening in the market and why.
How does an investment veteran with 52 years of experience publicly acknowledge, 'We haven’t executed this principle as well as we should have in the past'? How does one distinguish between temporary setbacks and a deterioration in underlying logic? And where does one draw the line between adaptation and adherence to core principles? For investment managers today—facing similar pressures over performance and scrutiny of their investment style—this exchange offers a rare example of candor.
As for whether this adjustment is correct, he set a verification period: “If I were you, I’d wait until the end of 2027 to make a judgment.”
Approximately 18 months.
Below is a summary of the key points from this discussion. Smart Investors has compiled the highlights of the presentation along with the full Q&A session for your reference. The Q&A segment was particularly insightful.
01. Momentum strategies are currently outperforming the broader market to the greatest extent seen in the past 30 years.
Net inflows into ETFs have been steadily increasing since 2020. After a temporary decline following the pandemic, they have now entered an almost vertical upward trajectory.
Flows into ETFs in 2026 are already about 50% higher than those in 2025—and we are only halfway through the year. The scale of capital shifting from actively managed products into various ETFs can now only be described as a 'tsunami.'
Yet I still see some commentators claiming it is incorrect to suggest that these capital flows affect stock performance and valuations.
To this, I can only reiterate what was said in All the President’s Men: 'If they can’t figure that out, then they’re not very smart.'
Because this capital flow is, by its very nature, a momentum strategy.
You can clearly see what is happening now by comparing the Bloomberg MSCI Momentum Index against the broader MSCI Index.
Currently, momentum strategies are delivering their strongest relative performance against the broader market in the past 30 years.
If you are engaged in active investing but are not employing a momentum strategy, you are currently facing significant difficulties.
And from around mid-2025 until recently, we have been comparing$Russell 2000 Index (.RUT.US)$the performance divergence between companies with negative earnings per share (EPS) and those with positive EPS—by negative EPS, we mean companies that are unprofitable, as is commonly understood. As you can see, the performance gap between these two groups is very large and continues to widen.
What we are witnessing in the market today is essentially a situation where companies with no fundamental support keep rising—unless you believe their fundamentals will undergo a dramatic transformation in the future—while companies that are actually profitable are underperforming.
As I also mentioned in my semi-annual letter, this scenario reminds me of the movie 'Crazy World.'
When all these factors converge, I cannot think of any fundamental-based investment approach that could capture such a market move.
And a fundamental-based investment approach is precisely what we aim to deliver to investors.
02. The current portfolio valuation is clearly below that of the S&P 500.
As you know, our investment philosophy is: invest only in high-quality companies; never pay too high a price; and then do nothing.
This is the strategy we have consistently strived to implement.
As of the end of the first half-year, there have been no material changes—certainly no materially negative changes—in the companies within our portfolio. However, these figures should be interpreted with caution, as some of the metrics exhibit strong seasonality, and a six-month view may not accurately reflect the underlying reality.
The portfolio’s return on capital remains stable at 31%, and gross margin is steady at 62%. Operating margin has actually increased slightly, reaching 29%. Cash conversion of profits remains largely stable at 92%.
Among all the metrics, however, cash conversion of profits is likely the most susceptible to seasonal effects.
The interest coverage ratio has risen to a very high level of 43 times.
As of the end of the first half-year, we hold a highly resilient portfolio. The companies in which we are currently invested continue to meet our criteria for what constitutes a high-quality business.
As of the end of the first half-year, the free cash flow yield of our portfolio stands at 4.3%.
This change is driven primarily by share prices, rather than by free cash flow itself, as free cash flow typically does not fluctuate significantly over such a short period.
Moreover, we do not believe that data covering only six months provides a particularly reliable basis for judgment.
Nonetheless, it is evident that the portfolio’s 4.3% free cash flow yield is beginning to approach the level of the FTSE 100. In my view, however, the FTSE 100 is not a truly comparable benchmark in terms of company quality.
By comparison, the S&P 500’s free cash flow yield stands at just 2.7%.
In other words, the valuation of our current portfolio is already significantly lower than that of the S&P 500 Index.
I reviewed independent analysts’ forecasts for the forward one-year free cash flow yield of the S&P 500 Index excluding financial firms, which came out to 2%.
The reason this figure is significantly lower is very straightforward: several major companies within the index engaged in the artificial intelligence arms race plan to incur extremely large capital expenditures during this period.
Based on the plans these companies have already announced, their total capital expenditures amount to $1 trillion—nearly $100 billion more than the U.S. defense budget.
This clearly illustrates the intensity of corporate enthusiasm for investing in artificial intelligence and the significant impact these expenditures are having on free cash flow.
Regardless of whether other judgments are correct, I am quite confident about one thing: the valuation of our current portfolio is already substantially below that of the corresponding index.
For most of the fund’s existence, our valuation has not been meaningfully below that of the index. Now, for the first time, we have truly entered a phase where our portfolio is demonstrably cheaper than the index.
Of course, it is also possible that we have purchased a set of 'value traps.'
This possibility must be acknowledged, and I must always keep it in mind. However, if that were indeed the case, it would mean we have consistently acquired value traps across a concentrated portfolio of only around 30 companies, relative to those in the index.
If the market’s forecast of a 2% forward free cash flow yield for the index is broadly accurate, and even if our portfolio merely maintains its current price-to-free-cash-flow ratio, then we clearly hold a relatively inexpensive portfolio.
03. We have taken the principle of 'doing nothing' too far.
The list of companies bought and sold in the first half of 2026 is very long. Even compared to the already elevated turnover rate in 2025, trading activity in the first half of 2026 was significantly higher.
We have conducted a very thorough and serious review of our performance over the past period. We considered every possible option we could think of regarding whether action should be taken and, if so, what kind of action.
There were no off-limits topics during the discussion.
Frankly speaking, the only scenario in which a team member would face criticism is if they offered no opinion at all. They are free to propose any solution.
These included directly replicating the top ten stocks by index weight in our portfolio; or replicating those top ten stocks unless we could articulate a clear rationale for excluding any specific holding;
or alternatively, bringing in an external manager who truly understands the prevailing market investment style and excels at capturing momentum in these stocks, to either assist in managing the entire portfolio or oversee a portion of it.
We discussed a wide range of possibilities. Ultimately, however, we decided to continue with our current approach.
We believe that making an aggressive shift at this juncture—transforming ourselves into a closetIndex Fundindex fund, or beginning to chase momentum in large-cap artificial intelligence-related stocks, would be an extremely poor timing decision.
Whether these companies belong to the index we are comparing against or, like$SK Hynix (000660.KR)$to another index, the same logic applies.
We do not do this, first and foremost, because it is simply not our investment approach.
Secondly, whenever people change strategies at times like this, they often find the next day that they have chosen the worst possible timing.
If you have already been wrong about many things up to this point, why would you suddenly believe that your timing judgment this time must be correct?
We ultimately rejected all of these proposals.
However, we did conduct very thorough research into some of these possibilities, including what their implementation would actually look like and how we could execute them.
We considered one key issue as follows: we reviewed the investable equity universe over the past five years and conducted a study to assess whether it would be possible to construct a portfolio capable of outperforming the benchmark index by selecting only from our investable equity universe.
Such a portfolio, of course, has one inherent problem: it is essentially backtested in nature.
Under certain parameter settings, this portfolio might be difficult to manage in practice, particularly given constraints such as concentration risk.
Nevertheless, this study still provides important insights.
Our research revealed that the data indicate it is indeed possible to construct a portfolio from our own investable equity universe that outperforms the benchmark index.
This implies that one of the factors affecting our actual performance was myself—or, more broadly, our own decision-making.
If we intend to continue investing using this approach, we must consider: how can we mitigate the adverse effects caused by our own decisions?
We therefore compared the theoretical portfolio with our actual holdings.
The ultimate conclusion was that we had taken the principle of 'doing nothing' too far, allowing it to evolve into a form of position inertia—or, in behavioral finance terms, an endowment effect.
In other words, we continued holding a company simply because we already owned it.
Among the numerous comments generated by this performance review, some pointed out that the companies we later sold were largely those whose stock prices had underperformed in recent years.
This observation is accurate.
04. Newly acquired companies exhibit higher certainty of growth compared to those divested.
But this is not the issue we consider most important.
The real problem is that the fundamentals of these companies have also fallen short of expectations.
The fundamentals of several companies in which we invested or held for the long term—such as Unilever, Coloplast, Novo-Nordisk A/S, and Atlas Copco—have disappointed us.
However, we spent too much time looking for reasons to continue holding them. In fact, one primary reason we kept holding these companies was precisely because their stock prices had already performed poorly.
From the perspective of$Estee Lauder (EL.US)$and$PayPal (PYPL.US)$Based on our investment experience with companies such as Estee Lauder, when a company’s fundamentals have deviated from our original investment thesis, selling it is usually the right decision.
I see that PayPal appears to have finally received an acquisition offer today.
Even if the timing of the sale comes later than it ideally should have, selling companies whose fundamentals no longer align with our investment logic is typically still better than continuing to hold them.
Therefore, we sold them.
Subsequently, we selected a group of companies from our existing investable universe to replace them.
These are not new companies that we suddenly discovered from somewhere outside; rather, they are companies that have long existed within our investable universe and were eligible for holding all along.
We believe these companies represent better alternatives.
The companies we sold and those we newly purchased can be compared from many perspectives.
However, if we were to summarize the most evident common characteristic of the newly acquired companies, it would be this: compared to the companies we sold, they exhibit greater certainty in growth.
By 'growth,' we mean fundamental, organic revenue growth.
This is the true driver behind this portfolio adjustment and the change we ultimately implemented.
Although companies such as Broadcom, NVIDIA, SK Hynix, Amazon, and Apple are all part of our investable universe, we did not rush to buy them.
Had we done so, we could have largely eliminated the portfolio’s deviation from the benchmark index and continued participating in the momentum of these stocks—assuming such momentum persists.
But that is neither what we do nor what we aim to do.
What we truly intended was to sell a group of companies we had held for a very long time. Precisely because we held them for so long, position inertia and the endowment effect led us to continually convince ourselves to maintain these holdings, even as we observed their fundamentals deteriorating and their share prices declining accordingly.
We then redeployed these funds into other companies. For these newly held companies, we are more confident that they will deliver the fundamental growth we seek in the future.
This is the core of the current realignment.
What you see now are the look-through financial metrics of the adjusted portfolio—recalculated based on the current holding weights.
Return on capital employed is 31%, gross margin is 62%, operating margin is 29%, cash conversion ratio is 92%, interest coverage ratio is 43x, and free cash flow yield is 4.3%.
I do not believe we have deviated—or even significantly departed—from the three pillars of our original investment strategy.
We should have been doing this over the past five years.
Another conclusion we ultimately reached—and a key driver behind this change—is that our inertia from long-term holdings has, to some extent, hindered our ability to properly adhere to the first principle of our investment strategy.
We no longer hold a portfolio composed of over twenty of our best companies—that is, they are not the most outstanding group we could select from the entire investable universe.
You could certainly ask, 'Does this mean you didn’t execute that principle as effectively as you should have in the past?'
My simple answer is: yes.
That is simply the case.
Moreover, as long as the current market environment persists—by which I primarily mean market volatility, rather than other factors—the turnover and trading activity of our portfolio in the future may be slightly higher than they have been over the past fifteen to sixteen years.
However, I believe it will not approach the levels seen over the past six months.
This adjustment should be a one-off event. Nevertheless, we will remain vigilant.
On one hand, we must promptly recognize whether we are again developing new position inertia due to holding a particular stock for the long term.
On the other hand, market volatility may also present opportunities that necessitate action.
As I noted in my semi-annual letter, if the share price of a company with a £200 billion market capitalization can fluctuate by 30% overnight, then a rational response is sometimes required.
Taking all these factors into account, I expect our future trading activity to be somewhat higher than in the past, but still far below the trading levels of most market participants.
The total cost of this adjustment was approximately £11 million. For comparison, transaction costs in 2020 were around £5 million.
Therefore, the actual incremental cost imposed on the fund by this adjustment is not unacceptably large.
Question 1
In the past, Fundsmith was willing to accompany high-quality companies through temporary operational downturns, as long as the long-term investment thesis remained intact.
Now you place greater emphasis on momentum, have increased portfolio turnover, and are less inclined to buy into companies experiencing temporary difficulties. How has your threshold for patience with a company changed? And how do you distinguish between temporary setbacks and a deteriorating investment thesis?
Terry Smith: Regarding the patience threshold, my answer is this: personally, I’ve always had relatively low tolerance for problems at a good company—it’s just that my colleagues have historically been more tolerant.
Going forward, however, we must enforce this standard more rigorously.
In fact, in my view, our tolerance for Intuit, LVMH, Mettler-Toledo, Nike, Novo-Nordisk A/S, Otis, Unilever, and Zoetis has long since reached its limit.
That list is actually quite long.
You must accept one reality: even a good company can be ruined by its management.
Buffett often quotes: 'Invest in a business that even a fool can run, because sooner or later, a fool will run it.'
I rarely disagree with anything Buffett says.
But over time, I’ve come to disagree more and more with that particular statement. Many companies simply cannot be entrusted to fools—the damage they can inflict is severe.
Perhaps this is also related to the types of companies we invest in.
What we are discussing now goes beyond just$Coca-Cola (KO.US)$companies like this. Coca-Cola might still be able to withstand such management issues, but for medical technology firms—for example, Coloplast—it would be far more difficult.
I believe this has already changed.
The world around us has also changed, and it is no longer as certain as it once was that a company can regain its market position after making a mistake.
Take Nike, or even possibly LVMH—they believed they could bypass physical retailers entirely and reach consumers directly on their own, or they misjudged the Chinese market.
Once problems surface, it cannot be taken for granted that, simply by waiting patiently, these companies will inevitably recover.
In some respects, this is already a different world, and we must adjust our approach accordingly.
Determining whether a setback is temporary or indicative of deteriorating investment fundamentals is, after all, inherent to the act of investing itself.
My colleagues would tell you that the most effective indicator they’ve found is to observe when I start getting angry.
However, there are several other factors beyond this.
The first very useful signal is whether management, when confronted with a problem, is willing to analyze it honestly or starts talking nonsense.
Everyone makes mistakes. I’m sitting here today, in fact, describing mistakes we ourselves have made.
But if management starts spouting language that no one can understand—not even the speaker themselves—it clearly won’t help the company solve its problems.
Take Zoetis as an example. If the management there can truly resolve the issues the company currently faces, then I can only say their managerial capabilities must far exceed their communication skills.
Unfortunately, these two abilities are usually closely related.
Therefore, we look for management teams that can communicate clearly and honestly. They’ll tell you in language we can understand: 'Here’s the problem, and here’s how we plan to address it.'
The second factor is the nature of the problem itself.
Have they damaged something critical to the company—and is that damage potentially irreparable?
That is the real judgment you need to make.
In the case of some companies, I believe the answer is yes. Take what has happened between Nike and its physical retail channels, for example. It is far from clear whether it can now regain control of those channels.
Partly because nature abhors a vacuum: after Nike exited, competitors moved aggressively into the physical retail space.
Can Nike win back that market share? I’m not certain.
Finally, there is a way to distinguish between a temporary setback and a structural deterioration.
When the same thing happens repeatedly, you should analyze it like Auric Goldfinger in 'Goldfinger.'
He said, 'Once is happenstance. Twice is coincidence. The third time is enemy action—or a trend.'
This has already occurred with some companies we have held.
In this regard, I would particularly single out Unilever. Sorry, but I’ve heard the same story before, and I no longer believe it.
If you continually hear management repeat practices that fall short of our standards—particularly in capital allocation and corporate governance—it’s time to walk away.
Question 2
Many clients invest in Fundsmith precisely because you consistently adhere to a disciplined, low-turnover strategy: buy good businesses, don’t overpay, and then do nothing.
Now, the investment process has changed, and trading activity in the portfolio has increased. Why should long-term investors still believe that the original investment thesis remains unchanged?
Terry Smith: That’s a good question.
First, if you look at how we’ve constructed the portfolio today, you’ll find that, by all financial metrics, it is very similar to our previous portfolios.
What we hold now is not an entirely different portfolio.
It hasn’t suddenly become a collection of unprofitable companies with extremely low returns on capital, after which we tell you, 'Don’t worry—once network effects kick in or data centers start generating profits, everything will be fine.'
That’s not the case.
From this perspective, the new portfolio is very similar to those of the past.
Its growth profile may be slightly higher than before. Interestingly, its valuation may also be slightly lower.
What haven’t we done? Or, put another way, why should investors still trust us?
We could have simply gone out and bought NVIDIA, Apple, Amazon, Arista Networks, SK Hynix, and Broadcom outright.
After doing so, we might perform exceptionally well—or we might not. Who knows?
Personally, I don’t believe it’s guaranteed.
But I am certain that if we bought these stocks, far fewer people would complain about our deviation from the benchmark index.
But that is not our investment approach.
All of these companies are already in our investable universe, and we could easily buy them.
Yet we have chosen not to do so.
I believe there is only one scenario under which we would consider purchasing them—more precisely, not ‘if,’ but ‘when’ a disaster strikes, and amid the wreckage we uncover a gem-like opportunity in one of these companies.
Only then would we buy into these companies.
Finally, I would like to note that this turnover rate is indeed quite high. However, I hope I have been sufficiently candid in explaining that these adjustments should have occurred gradually over the past five years.
Clearly, the market has undergone some changes over the past five years.
Perhaps we should have raised the annual portfolio turnover to around 10% from the outset, rather than maintaining the previously extremely low level. Had we done so at the time, we likely wouldn’t be discussing this turnover rate today.
Moreover, I don’t believe an adjustment of the same magnitude will happen again in the future. I don’t think that six months, a year, or even five years from now, we’ll be sitting here saying, ‘The portfolio has turned over another 50% in the past six months for the following reasons.’
No. That kind of event won’t happen again.
Question 3
What gives you confidence in the future growth of the portfolio? Are you primarily focused on revenue growth at the moment?
Terry Smith: Regarding revenue growth, apart from analysis, there’s really not much else that can give you confidence.
However, within our analysis, we can focus on areas where we have relatively greater conviction.
For example, some companies have historically exhibited relatively low growth rates and may also face potential disruption from artificial intelligence.
Wolters Kluwer is a good example. Its growth rate is roughly in the low to mid-single digits, and there are concerns that AI could intervene and displace parts of its business.
By contrast, Sage grows at approximately twice that rate, and I have far less concern about AI disrupting Sage’s business.
Even if we assume both companies are affected by AI to exactly the same degree, one company’s growth rate is still double that of the other.
Therefore, we conduct this type of analysis company by company to determine which ones give us greater confidence that their future growth prospects are superior to those of our original holdings.
This is not to say that we have absolute confidence in their growth per se, although that is certainly important as well.
More precisely, our confidence in these companies is higher than in the companies they replaced.
Another good example is our sale of Atlas Copco and purchase of Legrand.
Atlas Copco primarily manufactures compressors, vacuum equipment, and similar products. Part of its business is linked to the semiconductor cycle, but a significant portion is unrelated to it.
We replaced it with Legrand. Legrand does have some exposure to the data center business—for example, through products like cables and cable trays—but it also has a substantial low-voltage electrical equipment business.
These products exhibit an excellent business characteristic: they represent a very small share of total project costs, yet they are critical to the successful completion of the project.
Moreover, customers prefer to buy branded products because they are unwilling to risk their own reputations.
Whether electricians or quantity surveyors responsible for managing electrical engineering costs, professionals feel more confident specifying Legrand’s products.
It is in this manner that we compare companies against one another.
Revenue growth is indeed the first pillar of this analysis. Improving profitability without revenue growth is certainly commendable, but ultimately, this approach has its limits.
That is precisely the issue. Relying on margin expansion to drive growth is a nice-to-have, but frankly, it can also be quite risky. Moreover, it is not a gift that can last indefinitely—profit margins inevitably face an upper bound.
Therefore, revenue growth is the first pillar—particularly organic revenue growth, which we can clearly identify.
Of course, we do not favor revenue growth unsupported by profitability, nor do we like companies whose losses accelerate as their growth speeds up.
Nevertheless, our analytical starting point is indeed revenue growth—especially organic revenue growth.
Question 4
Over the years, you have criticized the management teams of certain holdings. Do these issues primarily stem from individual managerial shortcomings or external events? What steps will you take going forward to minimize similar situations?
Terry Smith: In a sense, it’s both.
I don’t believe we must attribute problems entirely to external events or solely to individuals. However, ultimately, responsibility typically comes down to specific people.
Consider the companies we sold this time, including Coloplast, Intuit, Nike, Novo-Nordisk A/S, and Unilever.
Undoubtedly, many of these companies encountered some highly unusual and particularly challenging external circumstances.
Take Nike as an example.
It had to navigate the entire pandemic environment. During that period, people were clearly restricted from shopping at physical athletic footwear stores as they normally would. So, external events certainly had an impact.
But in all these cases, I believe the real issue lies with individual managers and how they respond to such events.
Coloplast’s problem was that after making an acquisition, it failed to properly manage the operational integration of the acquired business.
Intuit made one or two very large acquisitions that turned out poorly, yet management buried its head in the sand and refused to confront the reality.
Nike, for its part, essentially ignored one of its two major sales channels.
There are many similar examples.
So, if I had to choose between individuals and external events, I would say that individual factors matter more than any other consideration.
When evaluating management, I repeatedly arrive at the same conclusion.
It’s hard to believe that this year marks my 52nd year of involvement in business activities in various capacities. Some of my colleagues also have exceptionally extensive experience.
For example, this year marks the 40th anniversary of my working with Julian. Over such a long period, we’ve certainly seen many things and gradually developed an intuition about people.
Sometimes, we simply don’t place enough trust in our own instincts. When we sit there thinking someone is exceptionally capable, more often than not, our judgment is correct.
And when we believe someone is truly terrible, our judgment is even more likely to be accurate. No matter how much management tries to dress up the situation with commentary or sweeten it with flattery, the fact remains unchanged.
I’m answering this question while thinking through it.
But ultimately, the issue stems less from the environment and more from specific individuals.
Question 5
As you place increasing emphasis on momentum, should investors expect that you will hold winning positions in your portfolio for longer periods going forward?
Terry Smith: Yes, I believe investors should expect that we’ll be more inclined to let winners run.
Historically, this approach has worked very well for us, and I believe it will continue to do so in the future.
However, we also need to remain mindful of position concentration limits.
I can assure everyone that we have never breached these limits in the past, but there have been several instances where allowing winners to appreciate further caused our holdings to reach the concentration cap.
In the future, we will seek more creative solutions to address this issue.
For instance, at present—and particularly given current valuation levels—we believe there are attractive opportunities in the payment processing sector. Therefore, for the first time, we have simultaneously purchased$MasterCard (MA.US)$cards andVisa (V.US),。
This allows us to continue holding companies that are performing well—letting our winners run—without quickly approaching concentration limits on individual positions.
How to better manage winning positions so that we can maintain them over the long term is indeed a key consideration for us.
Question 6
Won’t the issues faced by some of the companies you sold also affect the newly acquired ones? For example, switching from Intuit to Sage, or from Wolters Kluwer to Veeva?
This question primarily concerns companies impacted by the recent selloff in software stocks. In a momentum-driven market, aren’t you at risk of being burned again?
Terry Smith: That is certainly possible.
I cannot guarantee that we will be immune to stock price momentum.
I do not know whether switching from Wolters Kluwer and Intuit to Veeva and Sage will necessarily lead to better share price performance in the foreseeable future.
However, I am relatively confident that, from a fundamental perspective, we are in a better position now.
The core concern behind the current so-called 'SaaS doomsday narrative' is that artificial intelligence could replace certain software products. However, compared with Intuit and Wolters Kluwer, the products offered by Sage and Veeva are less susceptible to substitution by AI.
Take Intuit as an example: approximately 37% of its revenue comes from its TurboTax tax-filing business.
I believe the tax-filing business is more vulnerable to AI substitution than accounting software. Tax law is essentially akin to a language, and the term 'large language model' itself offers some indication of this.
Tax law can be encoded, understood, and applied. Therefore, I consider this segment of the business to be more vulnerable.
Beyond that, Intuit also faces challenges stemming from Credit Karma and Mailchimp—particularly Mailchimp.
None of these issues exist for Sage. I would argue that Intuit faces higher risks from AI disruption than Sage does.
Even now, despite our discussions with Intuit’s management, they still do not appear to fully recognize—or candidly acknowledge—that the Mailchimp acquisition has encountered serious problems, nor do they seem committed to addressing it seriously.
When reviewing our own investment mistakes, we strive to confront the issues head-on.
But I have not seen Intuit approach the Mailchimp situation in the same manner.
The same reasoning applies to the comparison between Wolters Kluwer and Veeva.
Wolters Kluwer is a professional information publisher whose content is relied upon by various professionals. But ultimately, we are still talking about documents and language.
I have been emphasizing this point repeatedly: large language models are exceptionally good at handling documents and language.
At least at this stage, the physical world represents a more difficult boundary for them to cross, but whenever language and documents are involved, that is precisely where they excel.
By contrast, Veeva’s software manages the entire workflow—from clinical trial data to manufacturing data—and has already become deeply embedded in pharmaceutical companies’ day-to-day operations.
This goes far beyond merely processing language; it extends well beyond the realm of language.
From a fundamental perspective, I have greater confidence in Veeva.
However, that does not necessarily mean I will be correct about its stock price trajectory from now until the foreseeable future. In this regard, I suspect others’ guesses are at least no worse than mine.
Question 7
Regarding the principle of ‘doing nothing,’ what level of portfolio turnover do you expect to eventually stabilize at? Following this adjustment, how do you anticipate the portfolio’s volatility will change?
Terry Smith: I estimate it will be around 10%.
I’m not sure exactly how the portfolio’s volatility will change. But if I were to measure it, I would focus primarily on the Sortino ratio rather than volatility alone.
I am relatively confident that the new portfolio’s Sortino ratio will be higher than that of the old portfolio. In other words, the new portfolio should generate higher returns relative to the downside volatility it assumes.
The reason is simple: some of the companies we sold have already shown unfavorable fundamental characteristics.
After all, volatility can be triggered by many factors.
However, IBM’s situation this week illustrates that even a modest negative shift in fundamentals can trigger significant market volatility. IBM’s revenue was only about 2% to 3% below consensus expectations, yet its share price dropped by 25%.
Therefore, after switching from the sold companies to the newly acquired ones, I am relatively confident—when measured by the Sortino ratio—that the new portfolio of holdings will outperform the companies we sold.
Question 8
The current portfolio holds 28 stocks. Do you think this number is appropriate, or might it increase or decrease in the future?
Terry Smith: I believe this number is appropriate.
Reducing the number of holdings to just over twenty would make management extremely difficult, due to the concentration risk I mentioned earlier.
Therefore, we would likely aim to keep the number of holdings in the latter part of the twenties, rather than the early twenties.
Moreover, among the companies currently available for purchase, only one or two in our investable universe still capture my interest relative to our existing holdings.
Similarly, among the companies we have recently added to the portfolio, there are also one or two that fall toward the more speculative end of the spectrum.
After all, this is a portfolio, ladies and gentlemen.
For example, the way I view$Applovin (APP.US)$is entirely different from how I view TJX, the off-price retailer of branded merchandise.
These two companies are very different. This is also evident from their portfolio weights—one company’s weighting is clearly lower than the other’s. If these relatively more speculative companies fail to deliver the growth we anticipate, we would be willing to sell them.
Question 9
Last quarter, you purchased your first Asian equity. Are there currently companies in other regions under consideration?
Terry Smith: We continue to monitor opportunities globally.
There are several Japanese companies we like. These include a Japanese pump manufacturer, a few Japanese firms involved in the semiconductor industry, and a Japanese factory automation company.
In Korea, there are several memory chip companies. However, not all of these companies are currently suitable for investment.
The issue with most of these companies is that they are too closely tied to the current state of the semiconductor industry.
I believe it is quite risky—and inconsistent with our investment approach—to be directly exposed to this sector’s cyclicality.
Question 10
How long, in your view, should investors wait before fairly assessing whether this adjustment has yielded positive results?
Terry Smith said seriously, “If I were you, having sat here and listened to everything discussed today, I would wait until the end of 2027—that’s roughly 18 months—to make a judgment.”
Question 11
You previously invested in Meta and experienced its massive capital expenditures on the metaverse around 2021, followed by a swift reversal of course. Do you think similar strategic pivots could also occur among these tech giants?
Terry Smith: Yes, I believe such rapid strategic shifts could indeed happen among these tech giants.
Because my assessment has remained unchanged from the beginning: achieving sufficiently high returns from investments made in this manner is, if not impossible, at least extremely difficult.
Spending $1 trillion annually means these companies would need to generate an additional $200 billion in profits and free cash flow each year just to achieve a reasonable return.
I simply cannot see where these returns would come from.
For example, as evidenced by IBM’s results yesterday, we are already starting to see how these expenditures are acting like a massive pump, diverting funds that might otherwise have been allocated to other areas.
But this figure is simply too large, and the pool of capital available for it to absorb will eventually be exhausted.
So yes, I believe such a reversal could happen. If it does, I think it would create an apocalyptic scenario on two levels.
You could certainly say, 'I’m no longer building these data centers, purchasing GPUs, buying racks, cables, cooling systems, or power infrastructure, and I’m halting model training altogether. That’s it—I’m shifting into reverse and pulling out now.'
As you mentioned, others have done exactly this in the past—for example, Zuckerberg’s strategic pivot away from the metaverse.
I believe it could happen again this time.
But if it does happen, it will be a catastrophic disaster. These companies have clearly signed numerous binding contracts, and a reversal on such a massive scale would have extremely severe consequences for certain firms.
This is especially true for companies entangled in the current circular shareholding and circular financing structures, such as$CoreWeave(CRWV.US)$.
This won’t be as simple as flipping a switch and saying, 'Alright, we’re reversing course now,' without any repercussions.
I believe there will be significant consequences. Some companies may lose access to financing or even go bankrupt, and related entities could end up suing one another in various ways.
I’ve also seen some commentators argue, 'If these companies sharply cut capital expenditures, that would actually be a good thing, because we wouldn’t have to worry about them becoming capital-intensive, low-return businesses. Instead, they could revert to being light-capital, high-return software, social media, and advertising companies.'
I agree. Fundamentally, this is indeed very attractive.
The only issue is that this runs completely counter to the narrative the market currently believes, so I think it would trigger a severe market sell-off.
The market has been rising continuously because this capital expenditure is taking place, and the market views it as a positive development.
If one day this capital expenditure is not only no longer seen as beneficial but is entirely abandoned, the market won’t instantly and effortlessly switch to a different mindset:
“Isn’t this great? We now have a new batch of software companies again—with gross margins of 80% to 90%, minimal capital investment, and 100% cash conversion on profits. Fantastic! I can go have dinner and relax completely.”
It won’t happen that way.
Such a massive shift in narrative would inflict serious damage on the market. Moreover, we are already seeing some early warning signs in the market.
Everyone has surely noticed Meta’s recent announcement.
Initially, it stated it would develop its own large language model—and specifically an open-source one. Subsequently, it committed substantial funds to building data centers to run these models.
Now it has announced plans to sell excess data center capacity beyond its own needs to other clients, effectively launching a cloud computing business.
That is quite interesting.
First, there are already about three cloud computing companies dominating this market. Entering the market and competing with them would not be an easy task.
Consider how long it took Google before its cloud computing business finally reached a reasonably profitable level.
What reason do we have to believe that a fourth or fifth company entering the market could successfully achieve this goal?
Of course, you could always say, 'I’ll sell this computing capacity to external customers.' But will that generate sufficiently high profits?
I very much doubt it.
Especially since Meta is likely not the only company entertaining this idea. This could well become a common industry-wide practice.
I recall Jim Cramer once said on his show 'Mad Money' that he believed Meta’s announcement could drive the company’s stock price up by $100.
I don’t share that view myself. To me, it sounds more like whistling in the dark to summon courage.
So yes, I believe such a reversal could happen. In fact, I think it inevitably will.
And when it happens, it will cause severe disruption across multiple dimensions, including financing issues, the viability of certain businesses, related legal disputes, and the market performance of these companies’ shares.
Question 12
You previously seemed quite skeptical about the semiconductor industry’s outlook. What has changed now?
Terry Smith: I remain quite skeptical about the semiconductor industry’s prospects, as I believe the current nature of this frenzy of investment renders it unsustainable.
Will there be competition in the GPU space? Certainly—indeed, competition is already emerging. We already know that companies such as Broadcom will enter the GPU market to compete.
Will the current capital expenditure boom in this industry prove unsustainable? Yes, I believe it will.
However, we have already invested in some companies that, while affected, are not exposed to these dynamics in such a direct manner.
$Texas Instruments (TXN.US)$is a good example. It primarily engages in analog chip business.
Other examples, though not entirely identical, include companies like TSMC. For a long time, I have internally advocated within our team that we should hold TSMC.
Can we predict the semiconductor cycle? I am doubtful.
Can we predict who will ultimately emerge as winners in the semiconductor cycle? I am even more certain that I cannot.
But I know that TSMC is a primary beneficiary of the current trend, and it is likely to continue benefiting for the foreseeable future.
The same applies to power equipment suppliers. The current data center construction boom has clearly provided significant momentum to$GE Vernova (GEV.US)$and other such companies. However, I believe their businesses would continue to grow even without this AI-driven data center boom.
If all of this were to end abruptly, it would certainly hurt their businesses. But I don’t think it would fundamentally derail these companies’ core fundamentals.
They are continuously expanding their installed equipment base and are still at a very early stage with modular nuclear reactors. Regardless of what happens, I believe these businesses will continue moving forward.
Question 13
How closely do you monitor correlations among different stocks in your portfolio?
Terry Smith: We do monitor them.
Of course, we could sit here and say, “We hold a diversified portfolio. We consider Amadeus distinct from Sage, from Veeva, and from other companies as well.”
But the reality is that, at the end of the day, they are all software companies.
If other market participants perceive them as correlated, then—at least in terms of short-term stock price performance—our belief that their fundamentals differ won’t be of much help.
We do indeed observe how stocks in the portfolio have historically moved together when major factors impact the market.
We use this to assess whether we have acquired companies that may appear fundamentally unrelated but are, in fact, treated by the market as belonging to the same category.
Therefore, we continuously monitor correlations.
Question 14
Terry, do you still enjoy managing money?
Terry Smith: Yes, I still do. Although sometimes doing this job feels like having a headache during a thunderstorm—it’s really not very pleasant. But I suppose that’s true of any job; I don’t think there’s any job entirely without its problems.
I still enjoy learning new things and trying to apply those insights to investing. Were it not for all the critics, I’d probably enjoy the job even more.
We have a client who also enjoys rugby league matches, and I’m a fan myself.
He once asked me, “How do you deal with it when people write all sorts of nasty things about you?”
I said, “It’s simple—I don’t read them.”
I usually only read the headlines. That might be a bit lazy. But if, after reading the headline, I can already see it’s clearly drawing a conclusion in a particular direction, I sometimes think, ‘I probably don’t need to read the rest, because I already know what it’s going to say.’
It’s just like reading a lot of newspapers or watching certain news channels these days.
After seeing the headline, you think to yourself, 'Alright, the headline has already told me the stance, so I know exactly what the rest of the article will say—there’s no need to keep reading.'
This is probably my only real complaint about this job.
Question 15
Based on today’s discussion, my conclusion is that your investment principles, philosophy, and process haven’t undergone any substantive changes—you’ve simply incorporated a slight element of momentum. Is that a fair assessment?
Terry Smith: That is a fair assessment.
However, I’d like to add one point.
When we say 'incorporated a bit of momentum,' the first question we should ask is: what kind of momentum are we actually talking about?
My answer is: fundamental momentum, not price momentum.
I acknowledge that some of the significant companies we’ve bought, just like those we’ve sold, still exhibit negative price momentum.
The comparison between Sage and Intuit is a case in point.
However, I believe the fundamental momentum of these two companies is entirely different, and the fundamental risks they face from AI disruption could also be vastly distinct.
Moreover, I have also candidly stated that this recent shift means we must acknowledge that, over the past five years, we did not apply the rigorous standards we should have in evaluating our holdings.
This refers to the first principle of our investment strategy: invest in high-quality companies, with fundamental growth momentum forming an integral part of our definition of a high-quality company.
Historically, we have always been very clear that for a company to qualify as high-quality, it needs two things: first, strong returns on capital; and second, a source of sustainable growth.
We have consistently articulated this point with great clarity.
However, during our long-term holding of certain companies, we somewhat lost focus on the second criterion.
This is my most candid perspective on our current portfolio adjustment process.
Editor/rice