40 years of bubbles, booms and busts: tips from 6 veteran investors

Talk of bubbles and busts is in the air. Financial markets look frothy and investors everywhere have started to wonder aloud whether the current enthusiasm for all things AI is a rerun of the dotcom boom at the turn of the millennium – and whether the denouement will be the same.

If you want to know what we can learn from previous bubbles and busts, who better to ask than professional investors who lived and worked through them? Here six veteran London-based fund managers recall those crises and tell us what lessons they hold for savers today. Read on or jump straight to any of the managers below.

• Job Curtis - City of London Investment Trust
• Katie Potts - Herald Investment Trust
• Ben Rogoff - Polar Capital Technology Trust
• Peter Walls - Unicorn Mastertrust Fund
• Peter Spiller - Capital Gearing Trust
• James Harries - STS Global Income & Growth Trust 

‘Bubbles can be more powerful and last longer than you would imagine’

Job Curtis – Manager of the City of London Investment Trust since 1991

I started work with a firm of stockbrokers in the early 1980s before moving into fund management in 1985. The stock market star at that time was Japan. Its export success led to appreciation of the yen, which was countered with an expansionary monetary policy. This fuelled a speculative asset bubble, with soaring property and share prices.

I was very aware of the strength of the Japanese stock market, and its premium valuation relative to other markets, and when the ‘Black Monday’ crash came in 1987 I was expecting Japan to fall furthest. I couldn’t have been more wrong: the Japanese stock market was barely affected. This shows that bubbles can be more powerful and last longer than you would imagine.

Japanese stocks finally peaked in 1989, at which point they were worth in aggregate slightly more than the US stock market. Today the US market is worth over 70% of the global index and Japan less than 10%. At the peak in 1989 the largest Japanese company was NTT, the telecoms operator, followed by six banks. The average price-to-earnings ratio (p/e) of those seven companies was 67.

The dotcom bubble of the late 1990s had a major effect on shares listed in London as well as the US and elsewhere. At one point Vodafone was worth some 12% of the UK index. Its share price peaked at 346p in March 2000, compared with 94p today. I can remember presenting to clients during this period and talking about the virtues of dividend-paying stocks. I would be politely received, before the clients asked my accompanying salesperson for some more technology fund application forms.

The dotcom bubble peaked in 2000 and the largest seven companies in the US, which had an average p/e of 52, were Microsoft, Cisco, Intel, Oracle, IBM, Lucent and Nortel. Few were among the biggest beneficiaries of the growth of the internet and mobile communications in the years that followed, as the list excludes Apple, Amazon, Alphabet (Google) and Meta. This could be a lesson when considering which companies might be the AI winners.

• More on City of London Investment Trust

‘In the dotcom bubble everything went up but today there are buying opportunities’

Katie Potts – Manager of the Herald Investment Trust since 1994

Our focus at Herald is smaller companies in the technology and communications sectors. It is widely appreciated how a small number of AI-related companies – Nvidia, Microsoft, Alphabet and Amazon in particular – have led the performance of the S&P 500 index but the impact of this AI supply chain on the smaller companies’ sector is less appreciated.

Almost half of our returns since 2000 have been delivered by four stocks: Super Micro Computer (data centre servers), Celestica (high-performance switches), Fabrinet (optical networking) and BE Semiconductor Industries (chip packaging equipment). All have seen strong growth driven by demand for the building of AI infrastructure. Unquestionably they have been supplying the equivalent of shovels in a gold rush. They were all purchased some years ago on p/e ratios of 7 or 8 and now trade at an average p/e of 42.

It is hard to predict the top, and capital expenditure growth will inevitably slow. I am more positive about the larger companies that benefit from recurring revenues from renting their AI infrastructure and will continue to grow even if capital expenditure slows.

In the early days of the internet Cisco (networking) and EMC (storage) were the winners that supplied the hardware, while the applications came later (search, social media, retail etc). This will be the case for AI. In a decade or two we shall marvel at the applications. For now it is the picks and shovels.

Those big returns equally demonstrate the headwind to the rest of the portfolio, whose p/e ratios have generally fallen. This includes technology stocks perceived as ‘losers’ from AI, such as many software names; falling valuations have provided useful buying opportunities. In the internet bubble of 2000, by contrast, every holding gained and many loss-making public companies had lofty valuations. That was a valuation bubble. This AI boom is a capital expenditure bubble generating immediate earnings for the hardware suppliers.

The overvalued stocks in the technology sector tend to be those where material stock-based remuneration for employees is overlooked, those of a size to benefit from the increasing flows into index trackers and those companies not adapting to the AI world.

• More on Herald Investment Trust 

‘We see AI as the next general purpose technology, a transformative platform that enables waves of innovation’

Ben Rogoff – Lead manager of the Polar Capital Technology Trust since 2006

In early 2000 I was a fund manager in my 20s, helping to run one of Britain’s most popular technology funds at Aberdeen. When I joined in 1998, our funds had strong numbers but were still relatively small; a year or two later investors were clamouring for technology funds, swept up in the extraordinary promise of the internet. During the 1999–2000 ISA season we received inflows in just three days equivalent to half of the total amount we were managing when I arrived. Unless you were there, it’s difficult to convey the intensity and uniqueness of that moment.

Of course, we know how that story ended. The dotcom bubble burst, paper fortunes evaporated and the vast majority of companies launched during that period simply disappeared.

It is said that ‘bull markets reward you, but bear markets teach you’ and the years following the bust were among the most formative for me as an investor. I joined Polar Capital in 2003 and became lead manager of its technology trust in 2006. During those early years we expected the sector to remain in the doldrums as it worked off the excesses of the dotcom period. So, we focused on performance and, as believers in creative destruction, tried to identify where the next technology cycle was likely to emerge.

In time it became apparent that the excess capacity from the dotcom period, particularly the bandwidth built to carry burgeoning internet traffic, laid the foundation for both the cloud computing and smartphone cycles that followed. We were day-one investors in Google and Facebook (which we’ve held ever since), companies that were made possible by the earlier excess and that helped create vast new markets that changed the world far more than even those frenzied earlier years might have suggested.

Today, as a self-described ‘AI maximalist’, I see striking parallels with that earlier period. We consider AI to be the next ‘general purpose technology’, a transformative platform that enables waves of complementary innovation, is widely applicable across sectors and is capable of driving economy-wide productivity gains. However, just as early communication networks were not designed to carry vast quantities of internet traffic, existing CPU-based computing is incapable of training AI models or running AI workloads at scale. The need to build a new, high-performance computing infrastructure explains why we have so much exposure to AI enablers today.

But there are also crucial differences. The AI revolution is unfolding far faster than the dotcom era, building on decades of advances in smartphones, cloud computing and digitised knowledge. We also see a profoundly different ‘buyer’ this time, with the vast majority of today’s AI investment accounted for by the giant technology companies. Not only are they among the best capitalised companies on earth, in stark contrast to the debt-funded telecom operators of the 1990s, but they are also among the most technologically proficient, with a privileged vantage point on AI progress and on the potential markets being unlocked.

The magnitude of their spending, hundreds of billions of dollars this year, supports our view that AI is likely to prove the next general purpose technology. While we understand why the scale and pace of this investment have left some investors concerned about ‘another bubble’, we continue to see more similarities with the mid, rather than the late, 1990s.

Just as we consider risk as well as return when constructing our portfolios, investors should size their positions and adopt investment horizons appropriately. Doing so reduces the likelihood of being ‘shaken out’ of an AI cycle that we believe will transform the world by redefining the relationship between man and machine, just as the agricultural, industrial and information revolutions did before it.

• More on Polar Capital Technology Trust 

‘I identified the most overvalued stocks and the orders flowed – but they were buy orders’

Peter Walls  – Manager of the Unicorn Mastertrust Fund since 2008

Today’s valuations of the Magnificent Seven, precious metals, commodities and crypto can only point to elevated levels of speculation. When I watched a Financial Times video on Michael Saylor’s Strategy (formerly MicroStrategy) I reached for my copy of Extraordinary Popular Delusions and the Madness of Crowds. I concluded that the best description of Strategy was, with Holland’s 17th century tulip mania in mind, Tulips Squared.

Investment capital gets attracted to wherever it thinks it will achieve the greatest return. Throw animal spirits and FOMO into the mix, add a great story (true or false, it doesn’t seem to matter) with plenty of momentum and before long a bubble will appear. Figuring out if a bubble is close to bursting is not an exact science as countless episodes over the centuries remind us. However, when fundamental valuation metrics get seriously out of kilter it usually makes sense to seek greater diversification.

I often refer to the Tokyo stock market of the 1980s. In today’s parlance we would probably call it the period of Japanese Exceptionalism, when entrepreneurs were lauded for their acumen and the market traded at a p/e ratio of 60. Japan represented 45% of the global index. The market peaked in December 1989. Thankfully I do not recall British investors having much money in Japan back then. Just as well, as the Nikkei 225 delivered a negative return in real yen terms in the ensuing 35 years!

So, with the US market now representing 68% of the global index and many British investors having that level of exposure, there must be a case for increased caution. For me, there are some similarities today with the dotcom bubble which popped in March 2000.

Back then I was writing investment trust research at a stockbroker. Listening to a discussion about the prospects for a potential technology company flotation, which was deemed attractively valued on only 15 times sales (not earnings), I turned my attention to the trusts most heavily weighted to the Telcoms, Media and Technology (TMT) sectors.

For each trust’s largest holdings, I pulled together the standard valuation metrics. I then listed the trusts with the most demanding valuations. The note, entitled ‘Who’s got the biggest bubbles’, hit clients’ desks in early February. And the orders flowed. The only problem was that most were buy orders!  This taught me that you need to be a contrarian investor…

From its 2000 peak the Nasdaq bottomed out 77% lower in October 2002 and it was 15 years before a new high was established.

• More on Unicorn Mastertrust Fund 

‘Bitcoin, a pure expression of risk sentiment, has fallen by 25%, which might be a hint’

Peter Spiller – Manager of the Capital Gearing Trust since 1982

At present we are facing twin crises: stretched valuations in stock markets and oversupply in bond markets.

US stock market valuations are at their highest since the dotcom crisis, particularly for AI-focused companies. Meanwhile, more debt is being used to finance the AI sector. AI companies are also becoming increasingly interconnected, for example through ‘vendor financing’ – in effect, borrowing from your suppliers. The longer current heightened valuations persist, the greater the risk of a sharp correction. High levels of leverage in the financial system have the potential to amplify its impact.

Similarly, global government debt is elevated and budget deficits are at levels unknown in peacetime. This is reflected in rising government borrowing costs. Furthermore, the ownership of government bonds has moved decisively towards non-bank institutions such as hedge funds, whose leveraged positions appear to be rented rather than owned.

Despite these risks, financial conditions remain easy. Short-term interest rate expectations continue to fall, the extra interest demanded of riskier borrowers is small and stock market conditions remain supportive. But bitcoin, a pure expression of risk sentiment, has fallen by 25%, which might be a hint.

Faced with these risks, how should an investor respond? We have always been more concerned about avoiding stock market downturns than participating fully in bull markets. In the late 1990s we were early to reduce our exposure to shares, but our net asset value (NAV) increased in both 2000 and 2001 against the broader market trend. We repeated this pattern in 2007, when we reduced our allocation to stocks before the final leg of the bubble.

Fund managers who construct a portfolio meaningfully different from the market always risk losing their job, particularly when the market is a bubble that continues to run. But for an individual investor the calculation is often very different and the tolerance for large falls is much less.

US stock market valuations have looked stretched for several years and again we have been too early to reduce risk in our portfolio. However, the dotcom crisis and the global financial crisis saw falls in the S&P 500 of 49% and 57% respectively – far greater than the 10%–15% corrections that investors often expect. So, given the threats to both bonds and stocks, we think the best place to hide is in short-duration US government inflation-linked bonds (known as TIPS). Indeed, we are fortunate to have relatively low-risk assets available with a decent yield (about 2% in real terms) at a time when long-term prospects for most financial assets look negligible.

• More on Capital Gearing Trust 

‘Having witnessed three very damaging bear markets in the first decade or so of my career, I am perhaps conditioned to expect another’

James Harries – Co-manager of the STS Global Income & Growth Trust since 2020

My career began managing money for charities in the shadow of the Asian financial crisis. The so-called Asian tiger economies had achieved phenomenal performance but by 1996 they were overstretched.

Once the dust had settled, attention shifted to the US. In the mid-1990s, the ChatGPT of the day was the Netscape browser. After the browser’s launch in 1994, the company listed in 1995, firing up investors’ imagination and investment. The technology boom was born. Ignoring the Federal Reserve chairman’s warning of ‘irrational exuberance’, the markets put on a tear, peaking in 2000. Then, as now, capital markets had a distinctly fin-de-siècle feel. A long-running theme had led to extreme sentiment and valuations, which, as with all booms, resulted in a bust.

The two-year bear market that followed felt long and grinding. It was a reminder that when a market theme becomes overextended, it can take years for the effects to wash through. There was money to be made in high-quality, left-behind companies such as consumer staples and in government bonds. But technology was crushed and economies suffered a recession. Once again, investors who overstayed their welcome in the most popular but overvalued companies suffered brutal losses.

To offset the effects of the bust, the authorities put in place a very accommodative policy and encouraged a liberal regulatory framework for banks. This bolstered US house prices via exotic forms of financing, which triggered the global financial crisis of 2008. Following the failure of Lehman Brothers, global capital markets were thrown into turmoil. We saw massive government bailouts, quantitative easing and zero interest rates around the world. By the end, shares were cheap and set up a wonderful opportunity for investors, which we have been enjoying – interrupted briefly by the pandemic – to this day.

Having witnessed three very damaging bear markets in the first decade or so of my career, I am perhaps conditioned to expect another. It is sobering to think we have not witnessed an orthodox recession, one induced by rising interest rates to offset inflation, for 25 years. Stock markets are highly concentrated in companies that are highly correlated and spend investors’ capital on projects with uncertain payoffs. Valuations are at generational highs in the US while the world is in flux as the post-war order breaks down.

It is ironic to us that steady, dependable, high-quality companies that make up our portfolio offer decent value, just as they did in 2000, and are well placed to weather whatever may be coming.

Maybe history really does rhyme and we are, once again, about to be reminded of the lessons of the past.

• More on STS Global Income & Growth Trust

Got a burning question you want to ask? Why not drop us a line. Click here to ask your question.

• Read: The winners and losers on 2025 - in charts
• Read: Top 10 best-selling investment trusts of 2025
• Read: Lifting the lid on our UK investments

Important information - investors should note that the views expressed may no longer be current and may have already been acted upon. The shares in these investment trusts are listed on the London Stock Exchange and their price is affected by supply and demand. Investment trusts can gain additional exposure to the market, known as gearing, potentially increasing volatility. Overseas investments will be affected by movements in currency exchange rates. Investments in emerging markets can be more volatile than other more developed markets. Reference to specific securities should not be construed as a recommendation to buy or sell these securities and is included for the purposes of illustration only. Eligibility to invest in an ISA and tax treatment depends on personal circumstances and all tax rules may change in the future. This information is not a personal recommendation for any particular investment. If you are unsure about the suitability of an investment you should speak to one of Fidelity’s advisers or an authorised financial adviser of your choice.