Why we’ve never held tech – and have long owned energy

Overview

According to Jeremy Siegel (Stocks for the Long Run, 6^th ed., McGraw-Hill, 2023), if you want to outperform over the short term – that is, if you’re a speculator rather than an investor – you have two options: either (1) buy the stocks of companies whose “growth metrics” during the next 12 months will exceed “consensus expectations” or (2) those which you think other speculators, for whatever reason, will purchase.

Yet such tactics are futile. They presuppose something which nobody can possess: the ability to foresee stocks’ short-term fluctuations. Speculators are unwittingly attempting to heed the advice which the American humorist, Will Rogers, offered on 31 October 1929. “Don’t gamble,” he urged. Instead, “take all your savings and buy some good stock and hold it ‘til it goes up, then sell it. If it don’t go up, don’t buy it.”

The fundamental, unavoidable and insuperable problem is that companies’ short-term returns are random and thus unpredictable: hence nobody can dependably pick stocks which will outperform over the next 12 months (for details, see Stop kidding yourself: Nobody can “time the market,” 29 June 2025).

About one key thing, however, speculators – and today’s conventional wisdom – are generally correct: revenues and earnings tend to grow more quickly in the technology sector than in other sectors and the overall market. In these senses, “tech” stocks are typically “growth” stocks. The consensus assumes that techs’ relatively rapid growth produces higher returns, and that comparatively slow growth, as occurs in the energy sector, begets lower returns.

This assumption, however, is false: over short-term, medium-term and long-term intervals over the past century, and taking into account their returns’ volatility, tech stocks haven’t reliably outperformed producers of coal, gas and oil – and over very long intervals, tech has dependably underperformed energy. Moreover, energy stocks’ prospective returns are presently attractive and techs’ are poor.

What, then, to do? If you’re an investor rather than a speculator, says Siegel, “you must pursue a very different strategy. Surprisingly, earnings growth is not the primary consideration for better long-run returns” (see also Everything the mainstream says about earnings is wrong, 12 March 2024). Indeed, except for dividends, nor is “growth” of any kind. “In fact,” he finds, “faster growth (of revenues or earnings) often leads to overvaluation that results in long-term underperformance.”

Siegel’s conclusion, which advocates of “growth” ignore or deny, is crucial: “history shows that the most important criterion for long-term investors is to buy and hold stocks that stay reasonably priced relative to their fundamentals, (and) not to pursue those firms that grow the fastest.”

That’s why Leithner & Company has always shunned “growth” stocks. In particular, we’ve never purchased shares of tech firm companies but have long owned those of energy producers. Many people search relentlessly for the “next Microsoft.” Few, however, seek to buy globally significant producers of energy at a discount.

Yet the latter opportunities are far more numerous than the former. Hence undervalued or reasonably priced energy companies have over the past century provided a more reliable route to better long term returns than overvalued techs.

What Are “Technology Companies”?

Technology is applied science. It enables us to produce new, more and better products with fewer inputs of capital, labour and land. Over time, this makes us richer. Greg Ip (“Tech Has Never Caused a Job Apocalypse. Don’t Bet on It Now.” The Wall Street Journal, 27 February) adds that “it’s why we produce many times more food with far fewer farmers than 150 years ago, and our factories crank out more products with a smaller workforce than in 1979.” 

“The technology sector,” says Investopedia, “is the category of stocks relating to the research, development or distribution of technologically-based goods and services. (It) contains businesses revolving around the … creation of software, computers, or products and services relating to information technology ...”

Morningstar (“What Is a Technology Stock?” 1 May 2018) adds a crucial point. The tech sector “has been broadened many times … (It) was initially (in the 1950s) anchored in semiconductors, computing hardware and communications equipment … The addition of software companies (in the 1980s) expanded the perceived tech sector to include anything based on coding. (In the 1990s) more room had to be made for internet companies, (some of which) were media and content companies … Still others were launching features that (became) e-commerce, social media, the sharing economy and even cloud-based computing …” Today, of course, AI epitomises tech.

Given its definition, Morningstar concludes: “Amazon is still very much a tech stock rather than a retail stock because of the … way they are using technology to create … disruption and to disrupt new areas like healthcare, the cloud and now increasingly financial services …”

It’s a common misconception: major advances of technology didn’t suddenly erupt during the late-20^th century.

The pace of technological change has greatly accelerated since the Industrial Revolution. The major innovations of the first half of the 19^th century included electric light, photography, rail transport, refrigeration, steam power and the telegraph; those in its second half included motor cars, radio, telephonic and trans-Atlantic communications. The first half of the 20^th century added airplanes (including jet engines), air conditioning, mechanical and electronic computers, television, nuclear fission, radar and sonar.

During the second half of the 19^th century, railways were leading “tech” companies; during the first half of the 20^th century, so were electrical utilities as well as developers and manufacturers of automobiles, airplanes, radios and TVs.

Two points are thus indisputable: firstly, technology has long entailed much more than IT; secondly, over the past half-century IT has advanced more quickly than any other technology – and thereby dominated the “tech” sector. A third one is also evident: with the possible exception of EV developers and manufacturers, the automotive sector isn’t presently regarded as part of the “tech” sector. Hence a fourth key point:

Precisely because technology advances, yesterday’s exciting and pioneering technology becomes today’s outré and obsolete tech. Among the numerous examples: in the 19^th century, rail displaced canal-based transport – and in the 20^th, automotive and air transport replaced long distance passenger rail.

Somewhat arbitrarily but for simplicity, I’ll assume that the basis of most technological advances since the 1970s – that is, IT – dates from the 1920s. During that decade, International Business Machines became what it remained for much of the next century: one of America’s leading tech companies. Named the Computing-Tabulating-Recording Company when it formed in 1911 via the merger of several firms, it became International Business Machines (IBM) in 1924.

From then until shortly after the Second World War, it was best known for its punch card systems and tabulating machines; until the invention and development of semiconductors and integrated circuits, these were essential for large-scale data processing in business and government. The IT industry’s undisputed giant unveiled its first electronic computer in March 1953. In July of that year, it released the first mass-produced computer for business and education. These developments established IBM’s dominance of the “mainframe” market – which, almost 75 years later, it retains.

Indeed, according to Siegel from the 1950s to the 1980s IBM dominated computer technology “like no other firm before or since.”

Why, More Than Most Sectors, Does Tech Attract Speculators?

Clearly, tech companies often grow rapidly – and, speculators mistakenly believe, growth reliably produces outsized returns. Similarly, many speculators obsess about tech companies because they exemplify rapid innovation. Commercialisation is exciting – and thus, like moths to light, attracts speculators’ attention.

Tech stocks, in short, offer “growth” – which, in turn, creates expectations of massive opportunities for substantial returns.

Technology is the practical result of science; it applies advances of scientific knowledge to agriculture, commerce, finance, industry, medicine, transport and myriad other fields. Like their predecessors, over the past century tech companies have commercialised major advances of science: these included tabulating machines in the 1920s, mainframe computers in the 1950s, transistors and integrated circuits in the 1960s, PCs and associated software in the 1980s, the internet and e-commerce in the 1990s, “smart” phones and “cloud” computing in the 2010s and AI today.

Over time, these developments have radically changed the business, economic, financial and social landscape. They can also generate immense returns to the initial owners of the companies which lead these innovations.

How will a major technology evolve? Will it advance further or quickly become obsolete? Which companies – and which shareholders – will benefit? Which will lose? It’s undoubtedly very difficult, and probably impossible, to say; accordingly, opinions fluctuate constantly and sometimes dramatically. Hence a second factor: tech stocks, particularly those of “start-ups,” tend to be highly volatile.

They thereby provide an irresistible target to those – namely speculators – who naïvely believe they can ride booms and dodge busts.

It’s not unreasonable to describe major advances of technology as “revolutions.” Yet market participants and the mass media typically – and often grossly – overstate their effects. Most notably, it’s hard to demonstrate that IT has enhanced the economy’s overall productivity – and to the extent that it has, it’s easy to overstate its impact. As Robert Solow, the winner in 1987 of the Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel, aptly quipped: “you can see the computer age everywhere but in the productivity statistics” (see, for example, Never mind DeepSeek: here’s why the AI mania won’t last, 2 February 2025).

Never mind facts and history: business and financial media typically entertain little doubt, and occasionally their certainty hardens into messianic advocacy.

During the late-1990s and early-2000s, they were utterly certain that “Dot Coms” – many of which lacked revenues (never mind profits) – would become immensely profitable. The consensus was therefore unshakeable: the purchase of these shares, regardless of their prices and valuations, would continue to generate hefty gains.

Hence a third factor: speculators seek to profit from the media hype and popular excitement which boost tech stocks’ prices.

A fourth factor also relies upon hype and appeals to greed: in order to access new technology, large and established tech companies often acquire smaller firms. A takeover offer often causes the price of the target company’s stock to skyrocket. Investment bankers encourage these buyouts, and speculators try to predict and profit from them.

A fifth factor is more general: the rise of online trading platforms and mobile apps – which reflect the ubiquity of the IT revolution – has enabled small-scale speculators to undertake high-risk trading.

As Robert Shiller observed in Irrational Exuberance (Princeton University Press, 3^rd ed., 2015), this easy access magnifies the “feedback loop” which boosts tech stocks’ volatility.

In conclusion, the perceptions of massive, quick and easy profits attract speculators; innovative products or services, they believe, are particularly likely to generate such profits; and the tech sector, they also reckon, offers these prospects, products and services.

The tech sector thus offers the potential for immense gains; accompanying this potential, however, is the significant risk of considerable loss. Net of the risk, however, and as we’ll see, past returns have been far lower than tech speculators expect in the future.

Data

Although they’re no longer unchallenged, for a century American companies have dominated the tech sector. Moreover, stock market data are longer and more valid, reliable and detailed in the U.S. than elsewhere. For these reasons, I’ve analysed American data compiled by Kenneth French and his colleagues. For each month beginning in July 1926, they assigned each stock listed on the AMEX, NASDAQ and NYSE to one of ten portfolios based on its Standard Industrial Classification at the end of June. Table 1 summarises them. (The SIC is a system of numerical identifiers which is used by the U.S. government to classify businesses by their primary activity.)

Of these ten sectors, tech stocks reside primarily in the “high tech” (hereafter “tech”) and secondarily (in the form of “bio-techs,” drug and medical equipment manufacturers) in the “health” sector. I’ve therefore analysed these portfolios’ returns separately.

Table 1: French’s Ten Sector Portfolios

Figure 1 plots the number of companies in the energy, health and tech portfolios, as percentages of the total number of firms in all ten portfolios, monthly since July 1926. In that month the 51 companies in the energy portfolio comprised 10% of the 511 across the ten sectors. The 18 firms in the tech portfolio comprised 4% of the total, and the seven in the health portfolio comprised 1.4%. From the mid-1920s to early-1970s, energy’s share plummeted to 3%; by 1982, however, it zoomed to 8% – and then resumed its long slide. From 1995 to 2006, and since 2020, it’s been 3%.

Figure 1: Percentage of Firms by Sector, July 1926-November 2025

In contrast, from the 1920s to 1980 tech’s percentage share grew slowly; and from then until 1987, it lurched upwards (to 17%). After receding to 15% in 1998, it again rose rapidly (to its all-time high of 24% in 2001); since then, it’s sagged to 16% in 2021 and rebounded to 18% in 2025. Similarly, from the 1920s to the 1980s the health portfolio’s share grew slowly; from then to ca. 2015, it rose more rapidly; and from then until 2023 it increased to its all-time high (23%); since then, and for the first time, it’s receded marginally (to 22% in November 2025).

Energy’s share of the total has thus plunged by almost three-quarters, from 10% in 1926 to 3% in 2025. In contrast, the combined share of the two tech-heavy sectors, tech and health, has thus risen eight-fold, from almost 5% in 1926 to 40% or more in 2023 and 2024; it’s currently 40%.

Energy versus Tech: Siegel’s Telling Comparison

Given the halo and hype that’s long surrounded IT, and the animus and hysteria that coal, gas and oil (“fossil fuels”) have attracted over the past 10-20 years, the choice between companies which develop and commercialise cutting-edge IT, such as those in the tech portfolio, and “old economy” explorers for and producers of coal, gas and oil, is seemingly obvious: most people probably wouldn’t hesitate to choose the former.

Siegel provides a specific comparison which superficially justifies this popular – indeed, enduring and fundamental – bias.

From 1950 to 2021, America’s leading oil company, The Standard Oil Company (New Jersey), consistently fell short of IBM, a leading “tech” company since the 1920s, in every “growth category” he examined. These categories include share price appreciation, growth of revenue and earnings per share, etc.

As a first aside, in 1973 Standard changed its name to Exxon; and in 1999, Exxon and Mobil merged. The $73.7 billion deal made it the world’s largest non-state integrated oil company at that time. Ironically, ExxonMobil resulted from the recombination of the two major descendants of the original Standard Oil Trust. In 1911, the U.S. Supreme Court ruled that it was a monopoly which violated the Sherman Antitrust Act; it therefore ordered that Standard be separated into 34 separate and competing companies.

These successor companies included Amoco (formerly Standard Oil of Indiana, which merged with BP in 1998), Chevron (formerly Standard Oil of California), ConocoPhillips (formerly the Continental Oil and Transportation Company, which the Trust acquired in 1885), Exxon (formerly Standard Oil of New Jersey), Mobil (formerly Standard Oil of New York) and Sohio (formerly Standard Oil of Ohio, which BP acquired in 1987).

As a second aside, today IBM is no longer a top-tier, consumer-focused tech giant like Apple, Amazon, Google, Meta and Microsoft. Yet it remains formidable: it operates in over 175 countries, is pioneering fields like quantum computing and remains a leader in mainframe technology (which provides secure, scalable and high-volume transaction processing for most large organisations in the banking, finance, insurance, healthcare, government and transport sectors, including ca. 70% of the companies in the S&P 500), hybrid cloud platforms, enterprise software, consulting services and aspects of AI.

As evidence of its unrivalled ability to innovate and commercialise scientific research, “Big Blue” holds the record (29 consecutive years from 1993 to 2021) for the most U.S. patents per year generated by a single business.

Siegel’s specific point is also generally true: over the past century, the revenues and earnings of the companies in the tech portfolio have generally grown more quickly than those in the energy portfolio. “With all this information,” Siegel reckons, you’d favour tech and shun energy shares. “The decision,” he adds, “looks like a slam dunk.”

Yet this has often been the incorrect choice; moreover, if past is prologue then it’s the wrong choice today – and will be over the next decade.

Results

Very Long-Term Compounding of Returns

Let’s start with two results which at first glance flatter the health and tech portfolios. Figure 2 plots the total (including dividends but excluding taxes and costs of transactions), CPI-adjusted value per $1 invested in the three portfolios in July 1926. Each $1 invested in energy grew to $883 in November 2025. That’s a compound annual growth rate (CAGR) of 7.4% per year.

Figure 2: Total, CPI-Adjusted Value per $1 Invested, Three Portfolios, July 1926-November 2025

Each $1 invested in the health portfolio in July 1926 grew to $2,873 in November 2025. That’s a CAGR of 8.3% per year. Each $1 invested in tech grew to $2,615. That’s a CAGR of 8.2% per year. Over this 99-year interval, it appears, health has outperformed tech, and both health and tech have outperformed energy.

Equally clearly, their outperformance is relatively recent. Until April 2017, for example, energy’s total return (proceeds of $603 and CAGR of 7.2% per year) matched tech’s (proceeds of $598 and CAGR of 7.2% per year).

Using July 1926 as a starting point, I’ve computed the CPI-adjusted total returns (expressed as CAGRs) of the three portfolios, as well as the average of all 10 sectors described in Table 1, over increasingly long intervals; Table 2 summarises the results.

Three are paramount. Firstly, over all intervals, health outperforms energy, and over intervals of 60 years and longer, health also outperforms tech. Secondly, over intervals from 10 to 60 years, tech outperforms both energy and the 10-sector average.

Thirdly, however, in two of the three intervals of 70-99 years energy outperforms tech and the ten-sector mean.

Table 2: CPI-Adjusted, Total Returns (CAGRs), Four Portfolios, 1926-2025

Short-Term, Medium-Term and Long-Term Returns

Table 2 contains three crucial shortcomings. Firstly, people seldom hold portfolios for periods of more than 50 years. Secondly, both the energy and tech portfolios’ returns since the GFC have been grossly unrepresentative of their long-term averages: as we’ll shortly see, energy’s recent returns have been well below its average, and tech’s have been well above. Finally, the CAGRs in Table 2 quantify results for single intervals (e.g., the ten years from July 1926 to July 1936, 20 years from July 1926 to July 1946, etc.) rather than for ALL intervals of 10 years, etc. since July 1926.

Table 3 corrects these defects – and thereby reveals two problems for enthusiasts of tech stocks. Table 2 examined just one 10-year interval; Table 3 examines all (1,072) of them since 1926, and so on for all other periods.

Table 3: CPI-Adjusted, Total Returns (CAGRs), Four Portfolios over Seven Intervals, July 1926-November 2025

Over rolling short term (12-month), medium-term (60-month), long-term (120-month) and very long-term (20, 30, 40 and 50 years) intervals over the past century, the average total (including dividends), CPI-adjusted returns of American energy stocks have exceeded the 10-sector portfolio’s return.

Here’s the first problem for enthusiasts of tech stocks: over all intervals, the volatility of their returns (measured by the standard deviations of their CAGRs) exceed the fluctuations energy stocks’ returns. This problem is fundamental, and I’ll detail its significance below. The second problem is that, although their average short- and medium-term CAGRs are higher than energy stocks’; the two sectors’ average long-term returns are approximately equal – and energy’s very long-term average returns exceed tech’s.

The third problem: over all intervals, health stocks’ CAGRs exceed the other portfolios’. Hence health-related tech consistently outperforms IT-related tech.

Finally, the energy portfolio’s most recent CAGRs are lower than their averages; in contrast, over all intervals the tech portfolio’s most recent CAGRs greatly exceed their averages. Although I’ve omitted the details (for examples, see Stop kidding yourself: Nobody can “time the market,” 29 June 2025), the returns in Table 3 are approximately (that is, apart from “fat tails”) normally distributed.

Hence outsized returns, like the tech portfolio’s most recent ones, are likely to fall (that is, regress to their long-term means); similarly, the energy portfolio’s recent very low returns are likely to rise.

The Cyclicality of Long-Term Returns

Figure 2, which plots these portfolios’ long-term (120-month, i.e., ten-year) rolling total CPI-adjusted returns, elaborates the results in Table 3. Each of the series is highly cyclical. Moreover, during the 20^th century each series’ fluctuations were highly synchronised with the others. Since the turn of the century, however, and particularly since the Global Financial Crisis, they’ve become less correlated.

The tech portfolio’s most recent CAGRs are among its highest on record (albeit not as high as those during the 1950s (when the transistor and integrated circuit, the building blocks of the IT revolution, were commercialised) and late 1990s-early 2000s (which marked the apex of the Dot Com bubble). In contrast, the energy portfolio’s recent CAGRs are among its lowest since the early-1980s.

Figure 2: CPI-Adjusted Total Returns (CAGRs), Rolling 120-Month Intervals, Three Portfolios, July 1936-November 2025

Since the 1920s, these series have regressed to their overall means. It’s therefore reasonable to expect that the tech portfolio’s subsequent long-term CAGRs will fall, and that the energy portfolio’s will rise.

For each month since July 1936, I’ve (1) calculated the energy portfolio’s total, CPI-adjusted total return (CAGR) over the previous 120 months, as well as its CAGR over the subsequent 120 months; (2) rank-ordered the data by the return over the previous 120 months; (3) divided the dataset into five equal (by numbers of observation segments (“quintiles”); (4) calculated the energy portfolio’s average CAGR within each quintile; and (5) repeated steps 1-4 for the health, tech and 10-sector portfolios.

Table 4: Energy Portfolio’s CPI-Adjusted Total 10-Year Returns (CAGRs) by Quintile of Long-Term Past Returns, July 1936-November 2025

Table 4 summarises the results. It confirms that each portfolio’s long-term CAGRs regress to their means. The lower was a portfolio’s return during the previous 10 years, the more strongly it rises during the subsequent ten. Conversely, the higher were CAGRs during the previous ten years, the lower they fall during the subsequent ten.

Today’s returns, in other words, are where yesterday’s weren’t; and tomorrow’s will be where today’s aren’t. This cyclicality implies very different returns for the energy and tech portfolios over the next 10 years.

Recall from Table 3 that during the most recent 120 months (that is, to November 2025) the energy portfolio’s CAGR was 0.8% per year. That ranks among its lowest 20% (i.e., Quintile #1) returns since 1936. Table 4 indicates that returns in this quintile they subsequently rise strongly – by an average of 10.0% per year over the next ten years.

During the most recent 120 months, techs’ CAGR was 18.2% per year. That ranks among the portfolio’s top (Quintile #5) returns since 1936. Returns have risen to this quintile subsequently sag drastically – to an average of just 3.6% per year over the subsequent ten years. Finally, recall from Table 3 that during the most recent 120 months the health portfolio’s CAGR was 8.7% per year. That approximates the portfolio’s median (Quintile #3) return since 1936. Median returns during the previous ten years remained much the same (average of 9.2% per year) over the subsequent ten years.

Over the next decade, it’s thus reasonable to expect that energy (expected CAGR of 10.0% per year) will outperform tech (3.6%) and health (9.2%).

Variability of Returns and Monte Carlo Experiments

Until now, I’ve omitted what’s perhaps this article’s most important point: energy stocks’ returns have almost invariably fluctuated less – and thus by orthodox standards been less risky – than health and tech stocks’ returns.

Over all intervals, the standard deviations of the energy’s CAGRs are lower than health stocks’ and tech stocks’ (Table 3). What happens when we incorporate this volatility – which the mainstream regards as “risk” – into the analysis?

To answer this question, I’ve conducted a series of Monte Carlo simulations. These methods, sometimes called experiments, are a form of algorithmic analysis. They undertake large numbers of repeated random samples from probability distributions with given parameters; their outputs enable us to assess risks and opportunities under different assumptions and scenarios. These simulations are well-suited to situations in which we must consider outcomes whose likelihoods vary.

Monte Carlo simulations therefore provide a valuable but seldom-utilised tool; in particular, they’re an ideal way to quantify and assess stock markets’ possible and likely returns (see, for example, Australian versus American equities: past, present and future, 24 November 2025).

In essence, I’ve 

1. created a simple mathematical model of the processes I wish to analyse;
2. represented them not as single values (that is, a mean CAGR and its standard deviation) but as a probability distribution which has a limitless number of possible observations; and
3. run a very large number (10,000) of simulations of each distribution.

These simulations enable us, over various timeframes, to estimate the probability that the energy portfolio outperforms the others. On this basis, investors can better understand risks and make more informed decisions.

As a first step, I produced 10,000 simulated observations from a probability distribution whose mean is 9.5% and whose standard deviation is 22.9% (i.e., is identical to the energy portfolio’s total, CPI-adjusted rolling 12-month measures of central tendency and dispersion since July 1926). I then did the same for a distribution whose mean is 11.8% and whose standard deviation is 28.3% (in other words, the parameters of the tech portfolio’s 12-month return). 

For each of these pairs of 10,000 simulated observations, I then (1) subtracted the energy portfolio’s return from the tech portfolio’s; (2) calculated the mean and standard deviation of this series of 10,000 simulated relative performances, and (3) calculated the percentage of simulations whose value is greater than 0.0% (i.e., in which energy outperformed tech). I then repeated this exercise for the other intervals, and for comparisons with the health and 10-sector portfolios.

Table 4: Results of Monte Carlo Experiments, Energy versus Three Portfolios, July 1927-November 2025

Table 3 summarised the 12-month, 60-month, … and 50-year returns which have occurred; based upon these actual results’ parameters, each Monte Carlo simulation generated 10,000 results which could have occurred. That’s 10,000 simulations per interval, and six sets of comparisons; that’s thus 10,000 × 7 × 6 = 420,000 simulated returns. Table 4 summarises the results. They incorporate the volatility of returns into the analysis – and disconfirm speculators’ – particularly tech speculators’ – groundless but unshakeable opinion:

Over all of the rolling short (12-month), medium (five-year) and long (ten-year) terms since 1927, neither the tech nor the health portfolio has significantly outperformed energy.

Over these intervals, energy’s underperformance of health and tech (and outperformance of the 10-sector portfolio) is insignificant in practical terms: vis-à-vis each portfolio, energy’s probability of under- and outperformance is equivalent to the toss of a fair coin (i.e., close to 50%).

For very long-term (20-year to 50-year) intervals, however, the relative returns are substantively significant: the probability that energy outperforms the tech portfolio approaches 70%; the probability that it outperforms the ten-sector portfolio exceeds 70%; and the probability that it underperforms health exceeds 60%.

Conclusion

Considered as a whole, and as their many boosters rightly insist, tech stocks are “growth” stocks. Yet taking into consideration the volatility of returns, over periods up to 10 years high-growth tech doesn’t reliably outperform comparatively low-growth energy; and over periods of more than 10 years it usually underperforms. Moreover, energy’s prospects are currently comparatively bright and tech’s are relatively dim.

That’s why Leithner & Company buys the shares of globally significant Australian energy companies when short-term vicissitudes depress their values well below our conservative estimates of their fair value. It’s also why we’ve always avoided tech stocks.

Implications: Value Investors and “Tech” Companies

Leithner & Co’s behaviour resembles Berkshire Hathaway’s and other strict value investors’: they’ve seldom acquired, or even seriously considered, techs. That’s because they focus upon relatively stable and thus roughly predictable businesses whose valuations are reasonable or compelling. Technology’s overall advance is highly foreseeable; however, the exact direction and speed of its progress are anything but predictable.

The only thing that’s certain about technology is that it will advance; everything else is sheer guesswork. How, over the next decade, will it advance? Which companies will benefit most? Will today’s tech leaders remain so? Nobody can credibly answer such questions.

Tech Companies Generally Lack “Moats”

My results justify value investors’ actions. Whether in part (via the share market) or in whole (via takeover), Berkshire Hathaway, for example, seeks to acquire companies which possess strong and durable competitive advantages (“economic moats”). The fast-changing nature of the tech industry, wherein innovation is rapid and erratic and whose products’ lifespans are often short, make it particularly and perhaps insuperably difficult to assess – never mind predict – tech companies’ long-term cash flows.

Techs’ moats aren’t nearly as broad and deep as their enthusiasts commonly assert: major players can and do fall by the wayside.

From the 1950s to the 1980s, for example, IBM led the computer revolution and was by far America’s – and thus the world’s – biggest tech company. Today, however, its market cap is just one-twentieth of the average “Magnificent 7” company (all of whom qualify as techs). The same point applies much more forcefully to Texas Instruments and Fairchild Semiconductor: during the late-1950s they led the semiconductor revolution; today they’re also-rans (Fairchild still exists; have you even heard of it?). The market cap of Hewlett Packard, the model of numerous Silicon Valley startups including Apple and Google, is $20 billion. That’s just 1% of the average Mag 7 stock’s market cap.

Indeed, the unpredictable course of technology’s advance has rendered once-major players, such as Yahoo, largely obsolete. Who can credibly say that Apple, Google and Meta can’t and won’t follow the same path as IBM – or Research in Motion (now known as BlackBerry) or even Nortel?

Predictability and Valuation

Value investors seek relatively stable businesses – that is, those whose cashflows are reasonably stable and whose shares can therefore be plausibly valued. As I’ve demonstrated (see Forget next year’s commodity prices: focus on 2075’s, 31 August 2025), the very long-term prices of essential commodities are roughly foreseeable; for this and other reasons, so too are the cashflows of producers of coal, natural gas and oil.

In contrast and by value investors’ standards, many tech firms’ future is unpredictable. Hence their valuations are largely speculative.

Shift of Perspective

Warren Buffett’s attitude towards technology has evolved over the years. Exemplifying its evolution are Berkshire’s purchases of IBM’s and, more recently and successfully, Apple’s shares. Berkshire began to buy IBM in 2011. Underpinning its purchase was the belief that institutional “stickiness” (large corporations couldn’t easily switch from Big Blue’s legacy IT systems) provided a durable competitive advantage. As IBM struggled to adapt to new competitive dynamics, Buffett admitted that this premise was flawed; Berkshire exited the position in 2018 at a loss (see also Warren Buffett’s 25 biggest mistakes – and 4 lessons they teach, 21 November 2024).

Beginning in 2016, Berkshire began to accumulate a massive stake in Apple, which over the next several years became one of its biggest investments and most profitable holdings. More recently, Berkshire has sold much of this stake, yet Buffett continues to view Apple favourably (albeit less as a tech company and more as a developer of innovative consumer goods whose brand and customer loyalty are very strong).

In summary, Berkshire applies the same criteria to the tech sector that it applies to all sectors: it’s invested only when Buffett could establish, and reasonably confidently foresee, a business’s long-term competitive advantage(s) (see also How Warren Buffett has trounced “the world’s greatest hedge fund manager,” 11 August 2025 and Naysayers are wrong: You CAN emulate Warren Buffett, 10 June 2025).

The Bottom Line

It’s boring, and perhaps for that reason most speculators blithely ignore or strenuously deny it: relatively low multiples (of price to earnings, price to book value, etc.), and high and growing dividends, underpin long-term outperformance. From 1950 to 2021, as Jeremy Siegel has demonstrated, America’s leading oil company, The Standard Oil Co. (New Jersey), consistently underperformed IBM, among the leading tech companies from the 1950s to the 1980s, in every growth category he examined: price appreciation, growth of revenue per share, earnings per share, etc.

Yet over these decades, the relatively slow-growing Standard Oil, which later became ExxonMobil, generated a much higher total return than the faster-growing IBM. The tortoise, in other words, beat the hare. How? Is this a specific instance of a general phenomenon?

Siegel explains: “the excitement caused by IBM’s new products caused its valuation to rise markedly, and although it continued to churn out superior growth, the average price investors paid for IBM’s shares during those six decades was just too high.” Over these years “the average P/E ratio of Standard Oil was about half of IBM’s, and the oil company’s dividend yield was almost three percentage points higher.” For these two reasons, “those who bought (Standard Oil’s) stock and reinvested (its) dividends would have accumulated 11.87 times the number of shares they started out with, while investors in IBM accumulated only 3.63 times their original shares” (see also Do tech stocks really outperform their value counterparts? 29 August 2022).

“Although the price of Standard Oil’s stock appreciated at a rate that was almost two percentage points below IBM’s,” Siegel concludes, “its higher dividend yield made the oil giant the winner for long-term investors” (for details, see Dividends aren’t a bane – they’re a boon (19 November 2023).

Historically, technology stocks’ dividends have at best been poor and often non-existent. Yet that’s changing: currently, ca. two-thirds of the companies in this sector pay dividends. Indeed, on an absolute basis tech is becoming a major source of payments. On the other hand, techs’ yields remain very low (often below 1%) compared to sectors like energy.

Well-established, conservatively financed and dividend-paying companies which sell at discounts to their fundamentals typically offer less volatile yet higher long-term total returns. Energy companies are relatively likely to meet this criterion, health stocks less so, and techs least so.

That’s why Leithner & Company has long held the shares of leading providers of energy, and is prepared to hold attractively-priced major pharma manufacturers or medical equipment makers, but has never held techs and has no plans to do so.

Click LIKE so that Livewire knows that you want more of this type of content. Click FOLLOW for notification when my next wire appears.