EP97 · Economy · first published 2021-09-03

Winner Stocks in the Portfolio | Ernst Grönblom | Negotiator 97

This is a summary on Neuvottelija — Articles. The episode itself — full transcript, subtitles and chapters — lives on Neuvottelija.com, which is its canonical home.

Ernst Grönblom, a portfolio manager at United Bankers, builds portfolios of fewer than 20 stocks — and the episode's premise makes that look hopeless from the start: Hendrik Bessembinder's dataset of 26,000 companies shows that about four percent produced all of the return above the risk-free rate over 90 years, and under half a percent produced half of it. Grönblom's answer begins with what he does not do: closet indexing can only be avoided with a sufficiently concentrated portfolio. The factors — network effects, a founder-led enlightened dictator, a strong brand, America — are not separate criteria but components of Charlie Munger's Lollapalooza effect, and it is their combination that he argues the market misprices. The most analytical stretch unpacks the decade-long growth-versus-value anomaly: he offers rising investor sophistication, real-time information, the structuredness of valuation, and quantitative investing — and finally the point that growth stocks cannot be treated as a monolith, because the same superstar phenomenon explains this gap too. Grönblom makes the counter-argument himself: a concentrated portfolio's good track record is precisely the kind a statistician cannot distinguish from luck. The article extends the episode into a sourced analysis of growth, value and quality investing: Bessembinder's own conclusion favours diversification, Grönblom's criteria read more precisely as quality criteria, and Buffett's alpha has been decomposed in the research into cheap leverage and quality exposure.

Sami Miettinen · Sections: AI and the Economy

Winner Stocks in the Portfolio | Ernst Grönblom | Negotiator 97

Summary: Ernst Grönblom runs the UB Thales Argo equity fund with a concentrated portfolio of fewer than 20 stocks. The episode opens with him making the most thorough case against his own approach.

Its hard centre is Hendrik Bessembinder’s dataset: 26,000 US listed companies from 1926 to 2016, of which about four percent produced all of the return above the risk-free rate — and just over 90 companies, under half a percent of the population, produced half of that excess return. A randomly picked 20-stock portfolio is therefore a losing bet from the outset.

A note on reading this

Grönblom is a fund manager whose fund the interviewer says on air that he has personally invested in. Both of them state this. Views are attributed by speaker, and this article takes no position on whether the method described produces excess return — that question is exactly the one the episode itself treats as unresolved.

Bessembinder’s numbers

Sami Miettinen opens with the counter-intuitive part: the ordinary intuition is that roughly half of active managers beat the market return and half lose, with costs shifting the distribution slightly below the line. Bessembinder shows the reality is “an awful lot harsher”.

Grönblom says the study started from something banal: one of Bessembinder’s students asked half-jokingly whether stocks actually return more than the risk-free rate. In finance that is one of the self-evident truths — and the data did not support it as plainly as assumed.

Miettinen draws three consecutive conclusions from this, each of which dismantles a piece of standard advice. First: 20 holdings are enough to diversify away idiosyncratic risk but not enough to capture the return — that works only if you pick those 20 exactly right. Second: a broad ETF works precisely because the net catches the outliers. Third: a narrow 25-stock ETF may leave out exactly the return generators.

Grönblom adds a second concentration that was not Bessembinder’s topic but that others have studied: returns and losses are extremely concentrated in particular months, weeks and even days. His archive holds reports on how dramatically a ten-year total return falls if you are out of the market for the best ten or twenty days.

And he closes the door on timing without hedging:

“Predicting either one in advance is, if you ask me, completely impossible. Of course there are those who claim otherwise, and if they have that crystal ball, then good for them.”

What Bessembinder himself concludes — and why it matters

This section is not from the episode but an addition, because it is decisive for how the study should be used.

Bessembinder’s result is often quoted in defence of concentrated stock picking: because the return sits in a few companies, go and find exactly those. The study’s own conclusion runs the other way.

Positive skewness means the median stock underperforms. In the original sample, 57.8 per cent of stocks returned less over their lifetime than a one-month Treasury bill. When a small tail lifts the distribution’s mean, an undiversified portfolio is not a neutral bet but a systematically unfavourable one: the most likely outcome is underperformance even where the expected value is the market’s. Bessembinder puts this himself in the form that the results help explain why active strategies, which tend to be poorly diversified, most often underperform their benchmark.

Put differently: the same study that makes stock picking attractive also makes it unlikely. The prize is large and the probability is small, and these are two sides of one skew.

What makes this episode unusual is that Grönblom does not misuse the study. He never claims Bessembinder argues for picking. He says outright that the odds are against him and that his own record may be chance. That is rare, and worth noting before reading on.

The disclaimer the guest raises himself

Miettinen says he rewarded Grönblom’s fund with a 20,000-euro investment — as he earlier did with Petri Deryng’s PYN Elite — and justifies it by saying he simply thinks “it’s great that someone tries against the odds”. He keeps his own score that the odds are against his guest.

Grönblom’s first response is not a sales pitch but its opposite:

“Even though I happen to have a decent track record, if you look at it wearing a statistician’s hat there is no getting around the fact that it cannot be ruled out that it is pure chance. And the argument is even weightier because my portfolio is so concentrated — it may well be that I have simply been lucky.”

Concentration cuts both ways, and he says so first. It is the most methodologically honest moment of the episode: the same property he uses to argue that excess return is possible is also what makes excess return hardest to distinguish from luck.

How a lawyer became a fund manager

Grönblom’s career is a detour that began with the efficient market hypothesis. As a teenager he read the finance classics — Brealey & Myers and others — and concluded that if Nobel laureates say markets are efficient, then active investing is a fool’s errand. He studied law and worked for a long time as a corporate lawyer on M&A and listings.

That is where the observation came that turned it back around:

“A pretty significant share of even very large financial decisions — we’re talking hundreds of millions or billions, in acquisitions for instance — ultimately rest on intuition.”

He does not claim the observation is valid. He says it was “valid enough to spark curiosity” — reason enough to reopen the case. And on reopening it he ran into the value investing school, which contains names whose alpha is very hard to explain by chance.

Closet indexing: what he does not do

The first building block of Grönblom’s philosophy is a prohibition, not an instruction. About 20 years ago he encountered the concept of closet indexing, which Antti Petäjistö has studied and for which he developed a measure, active share.

The number is stark: depending on the market and the criteria, a quarter and at worst as much as half of funds marketed as active are de facto closet index funds.

And the mechanism, Grönblom says, is rarely deliberate copying. It is diversification:

“The diversification is so extremely broad that it starts to resemble the market portfolio more and more. And if it resembles the market portfolio in composition, then of course it also resembles it in return. Once you add heavy fees on top, creating alpha becomes impossible.”

From this follows his first criterion, and it is purely negative: the easiest way to avoid closet indexing is to build a sufficiently concentrated portfolio. Concentration is therefore not an expression of conviction but a necessary condition for conviction to show up in the return at all.

Growth and value: is the split meaningful

Miettinen offers the classic split as a weighting of cash flows in time: growth investing weights distant but fast-growing cash flows, value investing weights high near-term cash flows at a lower risk profile.

Grönblom accepts the taxonomy but not its explanatory power. He is not alone: Warren Buffett has said that growth and value are joined at the hip, and Grönblom names the reason that makes the split partly illusory:

“For all the more systematic valuation processes, one of the most important inputs is future growth expectations.”

In other words, growth is an input to valuation too — the split does not separate two methods but two weightings within the same method.

Lollapalooza: factors are not a list

Here is the episode’s central concept, and it is not Grönblom’s but Charlie Munger’s. Miettinen half-remembers the term, and Grönblom corrects it and defines it:

“By the Lollapalooza effect Munger means a situation where two, or preferably even more, trends or phenomena push a company in the same direction.”

And what makes it an investment criterion rather than a description:

“If there are several such supporting trends, and they additionally have some synergy with each other, you easily get situations where the average investor and the market blatantly underestimate that company’s potential.”

The claim is therefore not that factors produce returns. It is that the combination of factors goes unpriced, because assessing it is not addition. Grönblom, quoting Jack Treynor, calls it a slow-moving idea.

Miettinen’s brother Topi Miettinen raises the sharpest counter-argument in the episode, and it is left open: are the right factors not already in the price? Miettinen offers a possible answer himself — that some of them are lagged, and the synergy effects combine into a “typhoon” the market has not yet discounted.

Shopify: the factors unpacked

Grönblom uses Shopify as a worked example. The Canadian company gives small and mid-sized firms a platform on which an online store can be opened turnkey.

The factors landing on one company:

1. Digitalisation — the growth of e-commerce, the most obvious trend.
2. SaaS — a cloud-based subscription model with annual recurring revenue.
3. Capex into opex. This is his most analytical point. Before such services, an online store had to be built in-house: servers, software, consultants, more IT staff. That was a large front-loaded investment carrying real risk. Shopify converts the investment into an operating cost — at the cheapest, twenty or thirty euros a month. “And if it doesn’t fly, it isn’t a big loss.”
4. Supply-side economies of scale. Miettinen adds his own observation: the platform is standardised, customisation cost is small, margins can exceed 90 percent, and cloud capacity costs almost nothing these days.

Miettinen also makes a qualification Grönblom accepts: Shopify is not the best example of network effects. It has some, but they are not massive compared with social media.

Network effects: demand-side economies of scale

Grönblom builds the concept carefully through microeconomics. Ordinary economies of scale are supply-side: as output rises, unit cost falls. That is a cornerstone of all industrial logic.

A demand-side economy of scale is a different thing:

“There are products or services — usually a service — that have this curious property that the larger the user base, the more useful the service is to every user.”

The examples are deliberately old. The telephone: with two devices it was a toy; every new node raised the utility for all users. The marketplace: sellers go where the customers are, and customers go where the sellers are. Network effects are not a new phenomenon — they have existed as long as economic activity.

Miettinen states the investor-relevant part: on the supply side the cost curve falls, on the demand side the customer’s utility curve rises — and when utility rises, pricing power can grow with volume. “Which leaves you an expanding margin from both directions.”

The best example is Facebook, and Grönblom’s evidence is negative:

“Do you remember, about five years ago, that Google tried to launch a competing social media platform, Google+? Google, which presumably has bottomless resources, presumably the world’s best engineers and cleverest strategists on the job, could not make the slightest dent in Facebook’s position.”

Miettinen offers a qualification Grönblom does not dispute: Alphabet owns YouTube, the world’s second largest search engine. But they operate in different markets on a different model — and Spotify has not got there either, despite signing Joe Rogan for a hundred million.

America as a factor — and China as its opposite

Miettinen’s own factor is geographic, and he says he learned it the hard way: a US stock gives you tailwind by itself, and Asia is worth being careful with.

The example is Alipay. Jack Ma built a micropayments infrastructure in China whose network could have expanded worldwide — and the expansion was blocked arbitrarily: no entry into America, and the domestic payment system probably has to be subordinated to China’s central bank money system even though that adds no value.

Grönblom’s answer is affirmative and rueful:

“As a Finn, a patriot and a European, it is a little sad to have to admit how incredibly dominant America is, at least on the economic side.”

The reasons are structural rather than slogans: the breadth and depth of the capital markets, and at the other end, how entrepreneurship is supported and admired in American culture.

The enlightened dictator

This is the episode’s most surprising factor, and it emerges during the episode itself — Grönblom says it was not on the pre-agreed list.

Miettinen asks whether a charismatic founder-CEO is a factor. Grönblom starts from Churchill: democracy is the worst form of government except for all the others. His reading is precise: democracy has flaws, but in non-democratic forms of government the flaws are catastrophic — because an enlightened dictator can neither be found nor removed, and “sooner or later they all go mad”.

A company, he thinks, is a different case:

“In a company I believe the most effective form is precisely dictatorship. If you ask me, one criterion of the ideal investment target is that power is held by an enlightened dictator, preferably a founder-owner.”

The examples: Jeff Bezos, Steve Jobs, Elon Musk. The reason is twofold — they hold both the power that comes from ownership and the authority that comes from their own vision — and it connects to the pace of change: decisions must be made fast, and they must be right.

Miettinen notes that Ronald Coase argued the same thing in his theory of the firm: the firm exists precisely because hierarchy is a cheaper way to coordinate than the market mechanism.

The brand and its limits

Miettinen asks about Buffett’s brand investments — Coca-Cola, Nike — and Grönblom concedes that a strong brand is one of the few sources of durable competitive advantage.

Coca-Cola is his perfect example, and he puts it bluntly: the company will be able to sell, until the end of time, the same “brown liquid that almost nobody can tell from generic cola in a blind test” — at roughly five times the margin of the generic.

But he makes a distinction, and flags it himself as partly a matter of taste. A consumer brand is usually in a mature market, where explosive growth is hard to sustain decade after decade. The contrast is Facebook, which is gigantic and still grows 25–35 percent a year — “completely incomprehensible that a company that large can still grow at numbers like that”.

Miettinen names the mechanism that separates them: upsell. Coca-Cola Zero and regular Coca-Cola are alternatives to each other, whereas a platform company can sell the same customer base more services and grow revenue per customer.

There have still been consumer brands in the portfolio. Inditex, the Spanish owner of the Zara brand, was held for a long time while it was still growing briskly. And Fever-Tree, which Grönblom calls “the cleverest modern branding project I have come across”: two Britons noticed that the premium gin boom kept producing new artisanal gins that were still being mixed with industrial bulk tonic — and asked why fine gins had no mixer worthy of them.

When a factor flips sign

Miettinen’s hard question is the turning point: when a factor changes sign — rates rise, inflation arrives, value makes a comeback — can those inflections be exploited?

Grönblom answers first at a general level, and the answer is unusually blunt about his own profession:

“All active portfolio management ultimately rests on the rather arrogant belief that I know better than most other investors how the world works and where it is going. That is, in the end, a fairly arrogant assumption — I’d almost say psychopathic.”

Why growth won for a decade

This is the longest analytical stretch, and it is built as hypotheses rather than claims. The starting point: over the long run value stocks have returned better, so a decade of growth dominance is itself an anomaly.

Grönblom says he has drawn up a list of about 20 hypothetical explanations for his own use. Four are covered in the episode.

1. Low rates and central bank funding. Obvious and significant, but in his view not sufficient.

2. Investor sophistication has risen. The average investor today is considerably more sophisticated. Concretely: “even the average investor today knows perfectly well what a P/E ratio is. Twenty years ago they might not have.” From which it follows that the value investing field is more competitive than it used to be.

3. Information is real-time. Before the internet, a retail investor read the share tables in the newspaper, and only the Helsinki exchange at that; in the 1990s professionals had annual reports burned onto CD-ROM. Now anyone with a smartphone can see the ratios of every listed company in the world in real time. And here is the link that makes the observation asymmetric: near-term information is precisely what value investing needs.

4. The mirror-image argument. Grönblom’s own favourite, and the subtlest point in the episode: value and growth stocks cannot be examined in a vacuum. Amazon’s rise did not happen in a vacuum — as Amazon succeeded, many traditional retailers suffered or went to the wall. And those were exactly the companies sitting in the value bucket ten years ago. “That has partly pushed down value stock returns.”

Miettinen takes the same observation one layer deeper: as a by-product Amazon created AWS, which has itself transformed the startup scene while destroying traditional data centre and database players.

Structured and unstructured problems

This is the episode’s most original single idea, and Grönblom says he picked it up from his wife’s psychology exam.

A structured problem: one and only one correct answer, and one or at most a few paths to it. The classic example is a mathematical problem.

An unstructured problem: no single correct answer, context-dependent and partly subjective — in English, judgment. Example: what is the optimal strategy for company X.

The link to investing: if investing is a valuation problem, different targets can be placed on a spectrum by how structured they are.

And from this follows the explanation that ties the section together: structured problems can be put into algorithmic form. As computing capacity and data quality rise, quantitative investing screens every listed company and identifies the objectively cheap ones.

“When enough participants do that, the meat gets eaten off the bone.”

Grönblom bounds his own claim: he does not argue that value investing is easier than growth investing — quite the opposite.

Old and new

Grönblom borrows an alternative taxonomy from the Scottish asset manager Baillie Gifford: instead of talking about growth and value companies, talk about companies representing the old and the new.

He stresses immediately that there is no value judgement in the split — neither is better — and that it does not map one-to-one onto growth and value.

As support he brings a measurable trend: the breakthrough of new technologies is accelerating. From the launch of the telephone it took about 75 years before half of US households had one; for the internet the same took 10 years. Order the technologies chronologically and the trend is unmistakable.

Amara’s law

Roy Amara, the American futurist, formulated a rule Bill Gates has restated in his own words:

“People on average overestimate the effects of a new technology in the short run and underestimate its significance in the long run.”

Grönblom’s example is the smartphone: travel 20 years back to the peak of the tech boom and “not even the wildest visionaries could have imagined what a device a modern iPhone is”.

Miettinen offers a counter-example that keeps the discussion honest: self-driving cars were expected five years ago and did not arrive. Grönblom concedes it directly — and generalises: over long horizons people have overestimated how fast the future comes at them, which has led to overvaluations in growth stocks.

The closing conclusion: growth stocks are not a monolith

Asked whether growth’s winning streak continues, Grönblom refuses to predict but says an automatic reversion to the mean is not self-evident to him — “at least at the healthier end”, as Miettinen qualifies and Grönblom accepts.

And he ends the episode by returning everything to Bessembinder, which is its most elegant structural move:

“It is of course also misleading to talk about growth stocks as one monolith. They are nothing of the kind. This phenomenon is largely explained by the superstar company effect that Bessembinder brought out.”

In other words: it is not true that growth stocks as a group returned better. A tiny handful of companies lifted the whole group — and he throws out a figure from the hip: about 95 percent of them were dreadful investments, some of them going to zero.

A long analysis: growth, value and quality

From here the article stops summarising the episode and sets it against the literature. Claims made in the episode are marked as such; research findings carry sources. The full source list is at the end.

Three styles, of which the episode names only two

The episode works through the growth-versus-value opposition carefully, and Grönblom shows its limits. But something happens in the episode that neither speaker follows to its end. Asking about Facebook and Shopify, Miettinen makes an aside:

“I have to make a small detour into your philosophy — this is quality, after all: fairly large firms.”

It is an accurate observation, and it is left hanging. Grönblom’s stated criteria are not growth criteria. They are quality criteria. Network effects, a strong brand, founder-owner leadership, high gross margins, switching costs — these describe the durability of a competitive advantage, not a rate of growth. Growth is their consequence, not their cause.

A three-way split is therefore more useful than a two-way one:

Academically, quality is defined more precisely than in ordinary usage. Novy-Marx showed in 2013 that gross profitability relative to assets predicts returns about as well as classic value measures and is negatively correlated with them — so quality and value together work better than either alone. Asness, Frazzini and Pedersen assembled this into the QMJ factor (Quality Minus Junk, 2019): profitability, growth, safety and payout. Fama and French added profitability (RMW) and investment (CMA) to their five-factor model in 2015 — in practice, a quality dimension.

Buffett at the intersection of the three

Grönblom grounds the possibility of active investing in Buffett: there are names whose alpha is “terribly hard to explain by chance”. This is the episode’s strongest appeal to authority, and there is a direct academic counterpart to it that the episode does not mention.

Frazzini, Kabiller and Pedersen decomposed Buffett’s return in Buffett’s Alpha (2018). The result has two halves and is worth reading whole:

1. Buffett’s return is statistically real. It is not a sampling artefact and is not explained by chance; the risk-adjusted return has been exceptional across five decades.
2. It is nonetheless largely explained by systematic exposures. Buffett has bought cheap, safe, high-quality companies — low beta, high profitability, stable earnings — and applied roughly 1.6-to-1 leverage whose funding cost has been below the market’s, thanks to insurance float.

Which reading is right depends on the question. If the question is “does skill exist”, the answer is yes: identifying factors before they were factors, and holding them for decades, is skill. If the question is “can anyone reproduce it by picking stocks”, the answer is no: a large share of the return came from cheap and permanent leverage that most investors do not have.

This bears directly on the episode’s argument. Buffett is not evidence that concentrated picking produces alpha. He is evidence that quality bought cheaply, levered, and held for a long time produces alpha. That is a narrower claim, and a better supported one.

An aside that also narrows the authority: Buffett has repeatedly directed his heirs and the general public to an index fund, and won a public 2008–2017 wager with an S&P 500 index fund against a basket of hedge funds. Buffett can therefore be quoted in support of either position — which is a good reason not to quote him as the resolution of either.

Why value lost for a decade — four explanations from the episode and two from the literature

Grönblom’s four explanations are low rates, rising investor sophistication, the real-time availability of near-term information, and quantitative investing. The last three form a tight argument together: valuation on the value side is a more structured problem, hence more amenable to algorithms, hence more competed. It is the episode’s most original idea and needs no improvement.

The literature offers two explanations the episode does not reach, and they point in different directions.

First: intangible capital breaks book value. The classic value measure is P/B, price relative to book equity. In accounting, research and development, brand building and software are largely expensed, not capitalised. As the economy tilts towards the intangible, a firm’s economic capital grows without its book equity growing — and P/B keeps classifying it as an ever more expensive “growth company” regardless of what it costs relative to its real capital. Lev and Srivastava present this as a central cause of value investing’s recent failure: the measure broke, not necessarily the phenomenon. This supports Grönblom’s conclusion but on different grounds — his case is that the value side has been competed away; this one says the value side was never measured correctly.

Second: the scale illusion, and the fact that value got cheaper. Arnott, Harvey, Kalesnik and Linnainmaa argued in 2021 that much of value’s underperformance was revaluation rather than deteriorating fundamentals: value companies did not do worse as businesses, people simply paid less for them. Revaluation cannot continue indefinitely, because the multiple cannot reach zero. Fama and French, for their part, showed that the shortfall in the later period sits within the dispersion the premium’s own volatility creates — a decade of underperformance does not statistically establish the premium’s death. This pulls against Grönblom’s conclusion: it says we cannot know.

Two Finnish notes in passing. The active share measure was developed by Antti Petäjistö (with Cremers), and the group questioning the value premium’s death includes Juhani Linnainmaa. Both sides of the episode’s argument rest partly on Finnish researchers.

What Bessembinder means for each style

This is the core of the analysis, and it is simpler than it looks.

Value investing does not need the tail. The value premium is a statistical property realised through the average, and therefore through diversification: you buy a hundred cheap companies, some are traps, the whole returns. Bessembinder’s skew does not overturn this, because a value strategy is not trying to hit superstars.

Growth investing needs the tail. If the return to growth stocks comes, as Grönblom himself says at the end of the episode, from a small group of superstars while roughly 95 per cent of the rest were dreadful investments, then a concentrated growth portfolio is a direct bet on hitting the tail. That is the honest description of what a concentrated growth portfolio does — and Grönblom gives it himself.

Quality investing sits between the two, which is what makes it interesting. A durable competitive advantage is precisely the property that lengthens the lifespan — and the harshest single figure in Bessembinder’s data is not the four per cent but the fact that the typical listed company disappeared from the register within a few years. A company that does not die lets compounding work on its behalf. A quality criterion is therefore not an attempt to forecast the winner; it is an attempt to remove losers, which is a different and easier task.

Read through this frame, Grönblom’s portfolio is a quality portfolio priced as growth. That explains both its strength and its vulnerability: if the quality is real, durability carries it; if a growth price was paid for that quality, a change in the rate environment cuts the valuation regardless of whether the quality was real. That is exactly what happened in 2022.

One question the episode does not ask

The episode never asks what happens when enough people believe in the superstar phenomenon. If Bessembinder’s finding is widely known — and it is, being among the most cited finance results of the 2010s — then the desire to own possible superstars raises their price in advance. The return to future superstars can then be lower than to past ones even if the phenomenon is unchanged. This is the serious form of the question Topi Miettinen asks in the episode — “aren’t they already in the price?” — and the episode does not answer it. That is not a criticism; it is an open question nobody has an answer to.

What happened after the episode

An addition to the article. The episode is from September 2021, and its central open question was whether growth’s winning streak would continue.

2022 gave the first answer, and it went against Grönblom. The rapid rise in rates cut the present value of distant cash flows exactly as theory says it should, growth indices fell markedly more than value indices, and concentrated growth portfolios suffered badly. The factor flipped sign in precisely the place Miettinen named in the episode.

2023–2025 gave a second answer, and it went Grönblom’s way — but in the shape of his closing argument, not the shape of his portfolio. The return of the US market concentrated unusually heavily in a handful of large companies around which the AI cycle was built. That is the superstar phenomenon itself, the thing he used at the end of the episode to explain the entire growth-value gap. Index concentration rose to levels that themselves drew concern.

Neither phase settles the argument, and saying so matters more than describing either phase. Four years is a short sample for a phenomenon Bessembinder needed ninety to measure. The episode’s own methodological caveat — that good returns cannot be separated from luck on a short record — applies to this postscript too.

What to take away

1. Bessembinder’s number argues both for the index and for a concentrated portfolio — each works only if the superstars are included; the difference is whether you catch them with a net or by picking.
2. Concentration is the condition for avoiding closet indexing, not a measure of conviction — and it is simultaneously why a good track record cannot be told apart from luck.
3. The Lollapalooza claim is not about individual factors but about their combination, which cannot be assessed by addition and therefore goes unpriced.
4. Four structural explanations are offered for the growth-value anomaly, of which real-time information and quantitative investing hit the value side asymmetrically.
5. The same superstar phenomenon also explains growth’s dominance — so growth’s victory was never the group’s victory.
6. Bessembinder’s own conclusion favours diversification. The skew that makes picking attractive also makes it unlikely; they are two sides of the same thing.
7. Grönblom’s criteria are quality criteria, not growth criteria. Network effects, brand and a founder-owner describe the durability of an advantage; growth is the consequence.
8. Buffett does not prove the case for picking but a narrower case: quality bought cheaply, levered, and held for a long time.
9. Value needs the average, growth needs the tail, quality buys lifespan. Three styles meet Bessembinder’s skew in three different ways.

Episode timeline

• 00:00 — Superstar companies and the highlights of the episode
• 01:00 — Guest Ernst Grönblom, a manager who has clearly beaten the index
• 02:53 — Bessembinder’s study: 26,000 companies over 90 years
• 06:01 — Four percent of companies produced all the excess return
• 08:35 — Returns cluster in a few days, timing is impossible
• 09:47 — Miettinen invests in Grönblom’s fund despite the odds
• 10:58 — Disclaimer: a good track record can be luck
• 11:52 — The efficient market hypothesis sent Grönblom into corporate law
• 14:38 — Buffett and the value investors: alpha hard to explain by chance
• 17:05 — Closet indexing and Antti Petäjistö’s active share
• 19:55 — A concentrated portfolio is the only way to avoid closet indexing
• 20:53 — Growth or value: is the split meaningful
• 22:41 — Buffett: growth and value joined at the hip
• 24:26 — Munger’s Lollapalooza effect: several trends in one direction
• 27:09 — Topi Miettinen’s question: are the factors already in the price
• 27:35 — Shopify: an e-commerce platform for small and mid-sized firms
• 29:52 — SaaS turns capex into opex and lowers the threshold
• 33:06 — Supply-side economies of scale in the cloud
• 34:52 — Network effects, or demand-side economies of scale
• 37:03 — The telephone and the marketplace as network-effect examples
• 39:42 — Facebook’s de facto monopoly and the failure of Google+
• 42:56 — America as a factor, Alipay and Chinese risk
• 45:40 — The enlightened dictator: a founder-owner running the company
• 50:22 — ESG and the sustainable development goals as factors
• 53:21 — A strong brand as a durable competitive advantage, Coca-Cola
• 55:00 — The weakness of consumer brands versus network effects
• 58:22 — Inditex and Fever-Tree in Grönblom’s portfolio
• 1:02:24 — When a factor flips sign: the arrogance of the active investor
• 1:04:17 — The rotation from growth to value and a ten-year anomaly
• 1:08:08 — Why growth won: rates, skill and information
• 1:11:59 — Amazon and value stocks as mirror images of each other
• 1:15:56 — Structured and unstructured problems in valuation
• 1:20:45 — From a bond to a startup: the valuation spectrum
• 1:23:16 — Quantitative investing eats the value premium
• 1:26:04 — Baillie Gifford: old and new instead of value and growth
• 1:29:58 — The pace of change accelerates: telephone 75 years, internet 10
• 1:32:02 — Amara’s law: technology is over- and underestimated
• 1:35:35 — Will the winning streak of growth stocks continue
• 1:37:21 — Growth stocks are not a monolith, superstars explain it
• 1:39:35 — Closing words and an invitation to alpha investors

Sources

Claims made in the episode are attributed to their speaker in the text. Below are the published studies and sources the article rests on — both those named in the episode and those the article brings in.

Bessembinder’s studies

• Bessembinder, H. (2018). Do stocks outperform Treasury bills? Journal of Financial Economics 129(3), 440–457. CRSP data 1926–2016, about 26,000 US stocks. Key figures: 57.8% of stocks returned less over their lifetime than a one-month Treasury bill; the entire net wealth creation of the market came from the best ~4% of firms; the top ~90 firms produced over half of that.
• Bessembinder, H., Chen, T.-F., Choi, G. & Wei, K. C. J. (2023). Long-Term Shareholder Returns: Evidence from 64,000 Global Stocks. The same skew repeats in global data: most firms underperform bills, and net wealth creation comes from a few per cent of companies.

Activity and closet indexing (named in the episode)

• Cremers, M. & Petäjistö, A. (2009). How Active Is Your Fund Manager? A New Measure That Predicts Performance. Review of Financial Studies 22(9). The original presentation of active share.
• Petäjistö, A. (2013). Active Share and Mutual Fund Performance. Financial Analysts Journal 69(4).

Value, growth and quality

• Fama, E. F. & French, K. R. (1992). The Cross-Section of Expected Stock Returns. Journal of Finance 47(2). The classic documentation of the value premium.
• Fama, E. F. & French, K. R. (2015). A five-factor asset pricing model. Journal of Financial Economics 116(1). Profitability (RMW) and investment (CMA) added to the model.
• Fama, E. F. & French, K. R. (2021). The Value Premium. The later-period shortfall sits within the dispersion created by the premium’s own volatility.
• Novy-Marx, R. (2013). The Other Side of Value: The Gross Profitability Premium. Journal of Financial Economics 108(1).
• Asness, C., Frazzini, A. & Pedersen, L. H. (2019). Quality Minus Junk. Review of Accounting Studies 24. The quality factor defined: profitability, growth, safety and payout.
• Arnott, R., Harvey, C. R., Kalesnik, V. & Linnainmaa, J. (2021). Reports of Value’s Death May Be Greatly Exaggerated. Financial Analysts Journal 77(1).
• Lev, B. & Srivastava, A. Explaining the Recent Failure of Value Investing. The accounting treatment of intangible capital and the breakdown of book value as a value measure.

Buffett

• Frazzini, A., Kabiller, D. & Pedersen, L. H. (2018). Buffett’s Alpha. Financial Analysts Journal 74(4). Buffett’s return decomposed: cheap, safe and high-quality companies plus roughly 1.6-to-1 leverage funded below market cost.
• Buffett, W. (1992). Berkshire Hathaway, letter to shareholders. The source of the line Grönblom quotes from memory in the episode: separating growth and value displays ignorance rather than sophistication, because growth is always a component of the calculation of value.

Concepts referred to in the episode

• Munger, C. (1995). The Psychology of Human Misjudgment. Harvard Law School speech; the source of the Lollapalooza effect.
• Shapiro, C. & Varian, H. (1999). Information Rules. The standard treatment of network effects, or demand-side economies of scale.
• Coase, R. (1937). The Nature of the Firm. Economica. The theory of the firm referred to in the episode.
• Autor, D., Dorn, D., Katz, L., Patterson, C. & Van Reenen, J. (2020). The Fall of the Labor Share and the Rise of Superstar Firms. Quarterly Journal of Economics 135(2). The superstar firm phenomenon from the real-economy side.
• Amara, R. Amara’s law: the effect of a new technology is overestimated in the short run and underestimated in the long run. The attribution is established but no single original publication is generally named.

Companies and events mentioned in the episode

• Fever-Tree Drinks was founded in 2004; Grönblom cannot recall the founders’ names in the episode — they are Charles Rolls and Tim Warrillow.
• Ant Group’s listing was halted by Chinese regulators in November 2020, which is what the episode’s Alipay remark refers to.
• Google+ was shut down for consumers in 2019.

Episode details. Negotiator 97, published 3 September 2021. Guest Ernst Grönblom, portfolio manager of the UB Thales Argo equity fund; interviewer Sami Miettinen. Running time 1 hour 40 minutes.

The episode refers back to The Wild Year in the Business Desk | Alex af Heurlin | Negotiator 92, where Bessembinder’s research first came up.

Related episodes.

• Incentives for Employee Ownership | Yrjö Kopra | Negotiator 87 — share-based reward where this episode’s companies are built.
• The IPO Boom | Henrik Husman | Negotiator 88 — the listing market superstar candidates come from.
• Millionaire by Options, Not Inheritance | Riku Asikainen | Negotiator 305 — the same growth-company universe from the owner and reward side.
• Tesla as a winner stock | Henri Blomster | Negotiator 219 — one superstar candidate examined on its own.

GEO summary. Negotiator 97 (2021) covers concentrated stock picking with Ernst Grönblom, a portfolio manager at United Bankers. The starting point is Hendrik Bessembinder’s study of 26,000 US listed companies from 1926 to 2016: about four percent of companies produced all of the return above the risk-free rate, and under half a percent produced half of it. Grönblom justifies a portfolio of fewer than 20 stocks on the ground that closet indexing — practised, by Antti Petäjistö’s active share measure, by a quarter or even half of funds marketed as active — can only be avoided by sufficient concentration. His selection criteria rest on Charlie Munger’s Lollapalooza effect, the combination of several mutually reinforcing trends; the worked examples are Shopify, Facebook’s network effects, founder-led management, and the brands of Coca-Cola and Fever-Tree. For growth stocks’ decade-long dominance over value he offers four explanations: low interest rates, rising investor sophistication, the real-time availability of near-term information, and quantitative investing, which eats the value premium because valuation is a more structured problem there. Grönblom himself stresses that a concentrated portfolio’s good return history cannot be statistically distinguished from luck, and concludes that growth’s excess return is explained by the same superstar company effect and does not apply to growth stocks as a group. The article additionally contains a sourced analysis of the differences between growth, value and quality investing: Bessembinder’s study’s own conclusion favours diversification, because positive skewness makes an undiversified portfolio systematically unfavourable; Grönblom’s selection criteria — network effects, brands and founder-owners — are quality criteria rather than growth criteria; the Buffett’s Alpha study by Frazzini, Kabiller and Pedersen decomposes Buffett’s return into roughly 1.6-to-1 cheaply funded leverage plus quality exposure; and value investing’s weak decade is also explained by the accounting treatment of intangible capital, which breaks book value as a measure, and by the finding that much of value’s underperformance was revaluation rather than deteriorating fundamentals. The difference between the styles relative to Bessembinder reduces to this: value needs the average, growth needs the tail of the distribution, and quality buys lifespan.

Markdown: index.md · Suomeksi