Neuvottelija · Articles

EP236 · Economy · first published 2024-02-07 · updated 2026-11-05

Quality Investing with Sifter | Santeri Korpinen | Neuvottelija 236

This article is based on a Neuvottelija episode. Find the episode details and recording on Neuvottelija.com.

Santeri Korpinen, CEO of Sifter Capital, explains how a single-fund equity house screens about 65,000 listed companies down to a portfolio of about 30 quality businesses, and why it holds no Finnish shares. A quality investor buys neither cheap value stocks nor loss-making growth stories, but companies whose business model, competitive advantage and returns on capital should carry them at least five years ahead. Examples include Novo Nordisk, Microsoft, Intel, Atlas Copco and Costco. The conversation also covers key-person risk in visionary CEOs, why Sifter left China, the temptations of Japan, the 2022 drawdown, alpha and the Sharpe ratio, investor psychology, Hannes Kulvik's legacy and the fund's daily liquidity. Published on 7 February 2024 in commercial collaboration with Sifter Capital. Not investment advice, and past returns do not guarantee future returns.

Sami Miettinen · Sections: AI and the Economy

Commercial collaboration: Sifter Capital. This article is based on Neuvottelija 236, published on 7 February 2024 and produced in commercial collaboration with Sifter Capital, as the host states at the start of the episode (00:08). Assessments of companies, portfolio holdings and markets reflect the interview at that time. This article is not investment advice, and past returns do not guarantee future returns.

Summary: Santeri Korpinen, CEO of Sifter Capital, tells Sami Miettinen how a house that runs a single global equity fund narrows roughly 65,000 listed companies down to a portfolio of about 30 quality businesses. In Korpinen’s account, quality investing is neither value nor growth investing. Sifter looks for companies that have already shown they can make money and whose business model, competitive advantage and returns on capital should hold up at least five years ahead. He estimates that numerical analysis is less than a fifth of the work. The rest is understanding why a business works. The examples are Novo Nordisk (bought for diabetes, with the obesity drug as a bonus), Microsoft (almost a value stock before the cloud), Intel (sold as its competitive position weakened), Atlas Copco and Costco, whose membership-fee model is the interview’s textbook case. The conversation ends with discipline, the team’s collective intelligence, the lessons of Hannes Kulvik and the fund’s daily liquidity. Published 7 February 2024, in commercial collaboration with Sifter Capital.

A note on reading this. This article is compiled from the episode’s Finnish machine captions and the publisher’s own chapter list. Views are attributed by speaker. Korpinen speaks as Sifter’s CEO, and Miettinen speaks as interviewer and from his own investing experience. The structure is thematic, so the timestamps do not run in order. Korpinen’s figures are his own statements. Where a figure has been checked against public sources, the article says whether it matches, and any difference is flagged rather than quietly corrected. Passages the captions left unclear are listed at the end. Quotations are translated from Finnish.

The interview opens with returns for 2023. Miettinen says the Helsinki stock exchange returned about –2% that year. Korpinen reports a 31% return for Sifter, which by Miettinen’s arithmetic is about 33 percentage points better. The figure is consistent with the fund’s own annual report for its R/I share class. The original YouTube chapter list incorrectly labels this comparison as 2022: the speakers are discussing the previous year in a February 2024 interview. Their opening comparison with the Helsinki stock market does not establish the global fund’s risk-adjusted outperformance. Sifter’s own annual report benchmarks the fund against the global MSCI ACWI index. (00:42, Sifter’s 2023 annual report)

1. One fund, 65,000 companies, no Finnish shares

Korpinen begins by correcting the introduction. Sifter is not an asset manager in the broad sense: it has one fund, a single equity fund, and the whole organisation is focused on it (00:14).

Miettinen asks directly whether the outperformance simply comes from holding no Finnish shares. Korpinen answers with probabilities. The Helsinki exchange lists about 170 companies, and there are about 65,000 listed companies worldwide. The chance that Sifter’s portfolio would include a Finnish company is therefore, in his words, “a few per mille”. The ratio is about 2.6 per mille, so his estimate holds. Over its history the fund has held a single Finnish company, which comes up later. (01:06)

The investment universe is global, liquid and tilted towards large companies. According to Korpinen, the biggest holdings have market capitalisations above ten billion (01:33). Miettinen compares his own portfolio, which he says returned about 25% in 2023, and notes that the Nasdaq and the so-called Magnificent 7 did even better. He suggests that international allocation and US companies explain part of Sifter’s return, and asks whether the rest comes from quality. Korpinen’s answer is that Sifter believes specifically in the quality of the business. (01:48, 02:19)

2. Quality is neither value nor growth

The classic split is value stocks versus growth stocks. Korpinen describes quality investing as a third, rarer style that gets little attention in Finland. Internationally, he says, there is a literature on “quality investing”, and an index of quality companies has performed very well for decades. Miettinen’s comparison becomes a running joke: if there are alpha males and beta males, quality investors are “sigma males”, who are neither. (02:43, 03:02)

The styles differ in what the investor pays for. Later in the episode the difference is put this way. A value investor buys stable companies with healthy balance sheets and good dividend yields, which are often inexpensive. A growth investor pays heavily for future growth in today’s valuation and often accepts losses. Korpinen says Sifter wants companies that have already made money and are accelerating healthy gross-margin growth. (10:21, 12:46)

Expensive now, cheap in five years

Korpinen says quality shows only over the long run. In the short term, quality businesses are usually priced fairly, perhaps even a little expensively. Given enough time, however, the purchase price often looks cheap five years later, because earnings have grown. In his view the market does not look very far ahead. (03:47)

He illustrates the point with Apple, one of the most closely followed companies in the world. According to Korpinen, about 40 approved analysts on Bloomberg publish forecasts for Apple’s current and next year. For years three, four and five there are only about four forecasts, a tenth as many. Miettinen had guessed twenty. (04:14) These numbers have not been checked, because analyst coverage varies over time. The point is structural: the market rarely tracks long-term earnings.

The five-year view is still an investment hypothesis that has to be monitored. Expected growth and the future share price remain uncertain.

3. Your own hypothesis before anyone else’s

Miettinen calls equity research the fund’s fuel and asks how much Sifter relies on sell-side research from banks. Korpinen’s answer describes the firm’s DNA: “We are a research house – we research, research and research.” Most of the time, the research ends with the business being rejected. (05:09, 05:29)

As he describes it, the process runs like this (05:52):

  1. About 65,000 companies are screened on numerical criteria, meaning historical, quantitative measures.
  2. When an interesting business turns up, the team forms its own hypothesis about why the company could keep doing even better for the next five or ten years.
  3. Only then does the team look at what others say.

The order is deliberate. Korpinen calls reading other people’s views first “a big bias”, because it leads your own thinking astray. (06:17)

The business model before the spreadsheet

Miettinen asks about a DCF model or some other valuation engine. Korpinen says the firm avoids the word “value” altogether, because “it’s a bit of a guess” (06:40). The first questions are about history (06:48):

The answers are worked out from the business model before “any number-crunching” begins. Miettinen asks whether the team uses unlevered measures, such as return on invested capital rather than return on equity, or EBITDA margin rather than net margin. Korpinen agrees and adds gross margin as an example. If a company keeps 60–65% after direct manufacturing and selling costs, it must have very strong pricing power. In SaaS companies, gross margin can reach 90%. (07:18, 07:35)

Next comes perhaps the most important statement in the episode about how the work is divided. Korpinen estimates that numerical analysis accounts for less than 20% of the work. You can run numbers in Bloomberg, Excel or your own model, but “it’s all just paper”. What matters more is understanding why the business works the way it does, where its competitive advantage lies and why it could keep going at an even faster pace. (07:58)

A private-equity horizon on a listed share

Miettinen offers an analogy: private equity with liquid listed shares. The allocation decision is made as if for a private company, with a horizon of about five years, but the investor can get out because the share is liquid. Korpinen says this thinking is built in and probably came from the firm’s founder: buy good businesses and stay with them as long as necessary. (08:33, 09:04)

No activism, no management meetings

From his banking career, Miettinen remembers activist investors such as Cevian Capital. They would build a large stake and then demand that a company sell itself at a premium or break itself up. Korpinen replies that Sifter invests in large companies, where it is a small shareholder. He is equally blunt about management meetings. If Novo Nordisk’s chief executive, or one of its C-suite executives, gave Sifter an hour, the team would hardly learn anything it had not already found in other sources. That time, and the travel time, is better spent elsewhere. (09:15, 09:42)

4. Novo Nordisk: bought for diabetes, obesity as a bonus

Novo Nordisk shows how an established business can contain new growth opportunities. Korpinen describes Novo Nordisk at the time of the interview as a growth company, with revenue growing “around 15%” a year. When Sifter invested, he says, it was not a growth company. The investment was made in the market leader in type 2 diabetes. He mentions Eli Lilly as a competitor, but Novo was seen as the clear number one. (10:45, 10:57)

The obesity treatment known in Finland as Ozempic and Wegovy was recognised as a growth engine. It was given no value in the models, however, because at the time it was still in the FDA approval process. “We avoid risks, if at all possible, by every means.” (11:12, 11:20)

Checked. According to Novo Nordisk’s annual results release, sales grew by 31% in Danish kroner in 2023 (36% at constant exchange rates), and obesity care sales grew by 154%. (Novo Nordisk, Form 6-K, full year 2023)

A note on the figure given in the episode. Korpinen’s “around 15%” is well below the growth Novo actually reported for 2023. If he meant a longer-run pace, the difference is understandable. If he meant 2023, the public figure is about twice as high. The gap supports his own point: the obesity drug was an option left out of the models, and it paid off more than expected. According to Sifter’s annual report, Novo was the fund’s largest holding at a weight of about 8%, and the position was reduced during 2023 (Sifter’s 2023 annual report).

5. Addressable market, market leadership and the price of growth

Miettinen asks what kind of market position a quality company needs. Korpinen says the total addressable market (TAM) has to be healthy and large enough, and the company needs “the ability to fight” in it (11:33). Whether it is number one or number two matters little:

Miettinen asks whether an ordinary start-up on the early part of the S-curve could end up in the portfolio. Korpinen says plainly that Sifter avoids them. It would be nice to back early-stage scale-ups, but many of them never become Nvidias. At the other end of the S-curve are mature businesses, such as General Electric at one point. When growth in their own market runs out, they try to keep themselves alive by buying start-ups, and many fail. (13:09, 13:22)

Capital allocation and the core business

Following the same logic, Korpinen is sceptical of acquisitions that take a slowing business far from its own expertise. His example is Cisco. When cash piles up and growth stalls, management comes under “terrible pressure to buy something” and build an empire, and the result often goes badly wrong. He thinks many Finnish companies, too, should have taken their core business deeper and global instead of buying things they did not know how to run. (14:08, 14:29)

Miettinen sums up that capital allocation needs discipline: return the cash to shareholders, split the company, or build the strategy sustainably. Korpinen prefers investment that strengthens the existing business: growing the market, building value-added services around the core and acquiring distribution channels at high returns on capital. “We like a focused business.” (14:42, 14:58)

6. Leaders, owners and countries

Key-person risk in a visionary CEO

What if the CEO is a visionary in the mould of Jeff Bezos or Elon Musk? “No exceptions allowed.” Korpinen says such “demigod CEOs” would be tempting, and sometimes they succeed, as Steve Jobs did. But they carry a large key-person risk, and the culture can turn into the rule of one person, whether a good dictator or a bad one. (15:18, 15:44)

The conversation notes that at Microsoft the transition worked under Satya Nadella, even though Steve Ballmer was tried in between. Miettinen brings up Jim Collins’s Good to Great (Finnish edition Hyvästä paras) and its levels of leadership. Korpinen says many of the executives on magazine covers are Level 4 leaders: they make a lot of noise about themselves and sometimes push the share price up too. Sifter looks for Level 5 leaders. They are quiet, work hard behind the scenes and build a system that works without them and can be handed on to “boring” successors. (15:55, 16:14, 16:36)

Owners as a hygiene factor, and leaving China

Sifter looks closely at a company’s ten largest shareholders, which Korpinen calls a hygiene factor. Norway’s oil fund, for instance, is such a common shareholder that its presence no longer tells you much. In Japan, the kind of owners a company has needs particular attention. (17:15)

On China, Korpinen describes a concrete decision. When the team studied Chinese companies, it always found some state proxy company among the owners. If Sifter put ten million into a Chinese company, it would be “a ridiculously small investor” with no voting power. So China was dropped entirely, and Sifter has not invested there for five years. Miettinen recalls what happened to Alibaba founder Jack Ma. (17:40, 17:50)

Japan: cheap, cash-rich and opaque

Miettinen raises Japan’s keiretsu and South Korea’s similar conglomerates, comparing them with Finland’s industrial groups of the 1980s. Korpinen calls Japanese corporate culture “quite peculiar”. Companies leave money on the balance sheet and hold a great deal of cash. (18:12, 18:41)

According to Korpinen, Sifter held three Japanese companies at the time of the interview, down from five the previous year. One is highly specialised. Another is Sony, owned for a couple of years. Sony has six business areas that he considers too far apart from one another, so Sifter built its hypothesis on three strong ones. (18:46, 19:01)

Miettinen sets out the activist’s logic. Idle cash drags down the valuation multiple, and when a “taskmaster” puts the balance sheet to work, the investor benefits. Korpinen agrees that Japan produces many screening candidates precisely because its multiples are low, and the temptation to buy more is strong. But there is “a curtain”, and an outsider cannot really see what is inside the company, so candidates are rejected. He adds that the central bank already owns a double-digit share of stocks, having bought them through the exchange onto its balance sheet. (19:40, 20:00, 20:24)

7. Lean thinking: 99% of companies are screened out

Miettinen mentions a previous guest who has read all the Toyota books and asks about lean philosophy. Korpinen says it is close to Sifter’s process thinking. There are far too many opportunities, so waste has to be eliminated, and 99% of companies are of no interest. By a rough estimate, about 200 companies worldwide meet the criteria for a quality company. Some are always too expensive, however, and some operate in industries Sifter does not want to invest in. Each step in the process should add value to the next, and the work should have flow. (20:45, 21:08, 21:15)

The numbers give the scale: of 65,000 companies, about 200 meet the criteria, and about 30 are in the portfolio.

8. Microsoft: a quality company that nearly became a value stock

Miettinen says he uses Trello as a personal kanban board. He wonders whether Microsoft will win this game too by building task management into its AI. Korpinen picks up the thread: Microsoft has been in Sifter’s portfolio for a long time. The number of years is unclear in the machine captions, and the earlier version of this article gives 13 years (see the note on sources). (21:46, 22:09)

According to Korpinen, there were moments during that period when Microsoft almost became a value stock and growth began to stall. Then the cloud business was built, “we got rid of Ballmer”, and the cloud business saw what he calls extraordinary growth rates of several hundred per cent a year. He is also curious where the gaming business is heading and whether it can be integrated. (22:18, 22:28, 22:43)

After the Copilot arithmetic, the conversation concludes that Microsoft is “suddenly a growth company again” and has always been a quality company. Miettinen remarks that Microsoft hit an “incredible jackpot” with its OpenAI partnership. He also comments on how a non-profit structure was turned into a money-making machine. (23:58, 24:07)

Copilot through a customer’s and an investor’s eyes

The most interesting tension in the conversation is that the same product looks different to a user and to an investor. Miettinen’s investment bank made him its first Copilot user. In his experience, Finnish-language support was not due until summer 2024. Mixed Finnish and English notes did not work because the tool offered only English, and the product was too expensive to roll out to his whole team. (22:47, 22:57)

Korpinen looks at the same product as a licence multiplier. He mentions an organisation he had spoken with that rolled Copilot out to 300 people in Finland. He remembers the price as 30 dollars a month. If Microsoft could sell it to, say, ten per cent of its existing licence holders, the growth would be “absolutely enormous”. (23:27, 23:41, 23:50)

Checked. Microsoft announced a price of $30 per user per month for Microsoft 365 Copilot, or about $360 a year. Korpinen’s figure matches, and Miettinen’s “around 300 a year” is a rounding of the same price. (Microsoft, 18 July 2023)

Miettinen says he owns Microsoft shares himself, through a Finnish equity savings account (osakesäästötili). The size of the holding is not given in the episode. (24:25)

9. Finland, Kone and Atlas Copco, and the problem with cycles

Miettinen asks which Finnish company the fund once held. Korpinen says it was Huhtamäki, before the financial crisis. Over the past five years, people have asked which Finnish company could meet Sifter’s criteria, and the answer would have been Kone. Miettinen points out that Kone would have delivered “a minus-20% flop”. Korpinen admits there has been China risk in the company, but China has also been a source of its growth. Kone’s strengths are recurring billing, predictability and growth. (25:13, 25:30, 25:43)

Sifter’s Kone-like holding is in Sweden: Atlas Copco, which is even more global and sufficiently focused. Korpinen describes the logic of its products. Compressed-air compressors keep factories running, and the same know-how has been “turned the other way round” into vacuum pumps. These are used, for example, to clear contaminants out of semiconductor fabs. The same product adds value in a different application. (25:58, 26:13)

Miettinen says Finnish engineering companies are late-cyclical. Korpinen replies that almost every business is cyclical in some way. In passing he names one exception, VeriSign, which Sifter’s annual report shows as a holding in 2023. The problem with cyclicality is that it cannot be predicted. One feature of a quality company is that you can keep assessing it on a rolling five-year view. Project businesses and construction, he says, are “simply impossible” to forecast. (26:52, 27:01, 27:28)

Miettinen offers his own experience as a counterpoint: in a downturn it does not necessarily pay to sell into falling prices. When, as he puts it, Sweden was oversold in 2023, he bought Swedish companies. He was looking for upside from the krona, a cyclical recovery and some “cigar-butt delta”. (27:37, 28:07)

10. Macro does not decide

Is a macro trend, an exchange rate or an interest-rate differential a decision criterion? “Absolutely not,” says Korpinen, though in the milder version of his answer they are part of the analysis. Screening can run top-down, looking at trends, themes, cycles and momentum, or bottom-up. Sifter works bottom-up. It looks for “a nugget of gold, or at least something that looks like gold”, and only then moves up to examine the industry, competitors, regulation, politics and trends. The aim is not to forecast economic cycles but to find businesses that come through them. (28:33, 28:58, 29:52)

According to Korpinen, the currency mix is a by-product of the investment decisions (33:26, 33:43, 34:01):

Miettinen notes that the weights are based on where companies have their headquarters, and that the geographic split of revenue would look different. Korpinen agrees: the headquarters country matters mainly for investor protection. (34:26)

Miettinen calls the “siren song” of emerging markets pointless. He says few of his friends have made consistent profits there decade after decade. Korpinen adds that a large enough global quality company does the geographic work for the investor. His example is Kone: if it decides China is a bad card, it reduces its exposure there and grows elsewhere. (34:51, 35:33)

11. Selling: a constantly ranked portfolio, and Intel

Getting into the portfolio is so hard, Korpinen jokes, that he could send the CEO of a newly admitted company a card of congratulations. The previous year one company came in and two left. That does not mean idleness: he says the team made 70 transactions within the portfolio. (30:48, 30:58)

The portfolio holds about 30 companies, which are kept in a ranking at all times. The ranking is based on the five-year outlook and the risk-adjusted expected return. Companies at the bottom are sold either at once or gradually. There are two routes to a sale (31:25, 31:41):

  1. Price. If the share price rises a lot, the expected return falls and the company slips down the ranking.
  2. Problems in the business. His example is Intel, sold around the turn of 2018–2019 after seven years of ownership. The team began to see Intel’s “reign” in semiconductors weakening and the company falling behind TSMC (Taiwan Semiconductor) and other chip designers. The whole position was sold at once. Korpinen says the share price later fell and has since recovered a little. (31:55, 32:00, 32:14)

No shorting, no currency hedging

Miettinen asks whether an industry ranking would not also tempt Sifter to short the losers. Korpinen admits that investing is “full of temptations”. The firm’s principle, however, is to reduce the number of moving parts to a minimum and manage only what can be managed simply. For the same reason Sifter does not hedge currencies, because that would add one more parameter to monitor. “Maybe we just aren’t smart enough for that.” (32:22, 33:00, 33:13)

12. 2022: geese that lay golden eggs

Korpinen puts the principle in an image. If you own a goose that lays golden eggs and it has a bad year, that does not mean it cannot lay over the next five years. Miettinen recalls the “half-massacre” of technology stocks in 2022. According to Korpinen, Microsoft, “our shared favourite child”, was about 30–40% down at the year’s low, and semiconductor companies about 50%. Sifter’s semiconductor weight was 20% of the portfolio at the time, and the fund lost money that year. (35:51, 36:14, 36:24)

Sifter did not sell, because it judged that these companies’ businesses would do well over the next five years. It bought more. This brings Korpinen back to the original question. The 31% in 2023 did not come from holding no Finnish shares. It came from having bought the right businesses at “a fairly smart price”. (36:44, 36:53)

The example shows how much the approach demands in tolerating falling prices. It does not remove the possibility that the assessment of a company’s future turns out to be wrong.

13. Alpha, beta and the Sharpe ratio

Miettinen observes that in a world of data, portfolios can be copied. A portfolio’s composition therefore has to change dynamically according to some logical principle. He says he has looked at the fund’s costs and finds them tolerable relative to the potential quality return, and he considers the Sharpe ratio fine too. (37:07, 37:27)

Korpinen says Sifter measures these numbers even though they do not drive its decisions, because “you have to speak the language the majority speaks”. Over the past five years, alpha has been about 4 percentage points a year, and in his view the risk-adjusted return is good. That figure has not been checked here, because the episode does not give the benchmark or the exact period. (37:38, 37:58)

Miettinen asks for an explanation of the terms, and Korpinen gives a brief one (38:24, 39:10):

14. Psychology, collective intelligence and discipline

Korpinen considers the psychology of investing perhaps the biggest risk. At Sifter, eight people watch a single portfolio. In his view, the team, a kind of “swarm intelligence”, manages the emotional climate and sticks to the strategy better than an investor working alone. (40:03, 40:21)

Mistakes are reflected on, not punished. Korpinen cites Ray Dalio, the “shared idol” of both speakers: every mistake must be examined by asking how it happened, what can be learned from it and whether it will be repeated. Because purchases and sales are rare, every decision can be reviewed. Miettinen describes the Dot Collector practice from Dalio’s Principles, in which conversations and decisions are rated and the result is a kind of believability weighting for each expert. The discussion then turns to culture. Miettinen says culture is, alongside quality, the thing to remember from the episode. It is durable, but it can also rot, so it has to be nurtured. Korpinen says he hopes Sifter has built a culture around its method that may even be unique. (40:46, 41:08, 41:46, 42:00)

Miettinen remembers reading Jack Schwager’s Market Wizards and The New Market Wizards as a young man. The books interview portfolio managers who beat the market. Their styles were all over the map, but many shared one principle, “trust the system”: buy and sell signals were followed consistently, even when they were sometimes wrong. (42:34, 43:00)

Korpinen recalls that when he joined the firm in 2016, he was told its values were hard work and discipline. Discipline matters especially because Sifter sells its product to investors. Selling the portfolio down to nothing in a Covid-style crisis would not be the firm’s strategy. “We have to be really disciplined about acting according to it.” (43:12, 43:31)

15. Costco: the membership fee at the heart of the business model

According to Korpinen, the numerical criteria for a quality company are returns on capital, margins and leverage, preferably zero. But these show only the tip of the iceberg and prompt the question of what drives the numbers. That is where the business model comes in. He notes that the Finnish term ansaintamalli, literally “earning model”, translates poorly. English speaks of a business model, but the Finnish word describes how a company makes money. (43:47, 44:15, 44:28)

His example is US retail. Sifter has not invested in Walmart but has invested in Costco. Korpinen describes Costco’s model as follows (44:53, 45:08, 46:00):

The conversation also mentions a trip to Omaha a year earlier “to marvel at Warren Buffett”, which included an arranged visit to a Costco. The store is described as enormous, “in the middle of nowhere”, with high-quality goods. (45:29)

Checked. In Costco’s fiscal 2023 (53 weeks, ended 3 September 2023), membership fee revenue was $4.58 billion and net income $6.29 billion. There were 71.0 million paid member households and 127.9 million cardholders. Fees in the US and Canada rose from 1 September 2024: Gold Star and Business memberships from $60 to $65, and Executive membership from $120 to $130. This was the first increase since 2017. (Costco fiscal 2023 results release, CBS News, 10 July 2024)

Notes on figures given in the episode. Korpinen’s description of the model holds up, but three details differ from the public record, which is normal in live conversation:

Culture versus greed in pricing the membership

Miettinen describes Costco’s membership fee as a SaaS-like moat. Fees come in predictably and churn is low because the quality is good. The danger, he says, is that at some point the company becomes too greedy with the fee and customers move to a cheaper rival. (47:48, 48:05)

Here Korpinen returns to culture. A smart business has management that does not raise the fee because it wants what is best for its members. He says a fee increase had been expected for five years at that point, and the company did not want to punish its members during the previous year’s hard times in the US. The previous increase was in 2017, so by February 2024 the gap was already about six and a half years. The increase finally came in September 2024, after the episode was published. (48:12, 48:21)

Miettinen contrasts this with Microsoft, which in his view charges for Copilot “without any pangs of conscience”. The conversation notes that there is no real alternative to the Office system. A short-term opportunity to charge customers more can conflict with a valuable long-term relationship, and judging quality means understanding how a company resolves those choices. (48:38)

16. Hannes Kulvik, the book Sijoittaja and the fund’s liquidity

Miettinen asks whether Sifter has an “academy” where its lessons on business models are written down. Korpinen points to a book published a year earlier, Terho Puustinen’s Sijoittaja (Otava, 2022), later published in English as The Investor. According to Korpinen, Puustinen followed Sifter for two years and interviewed the fund’s founder Hannes Kulvik, the team and investors. Korpinen suggests inviting Puustinen to the show to talk about investor psychology. (48:59, 49:08, 49:27)

Miettinen notes that Sifter does not seem to have a founder cult. Kulvik is a respected developer of the method, but the firm no longer rests on one person. Korpinen says he spent about three years in Kulvik’s “school” in 2016–2018. Kulvik made him repeat certain things “endlessly”, and the calm this gave him about investing has changed him as a person, and especially as an investor. (49:42, 49:56, 50:06)

Finally, the conversation turns to the fund’s structure. According to Korpinen, it was originally built for a small group of families and was not meant to be sold as a product. The fund is fully liquid, “in today, out today”, with no long lock-up periods. The underlying idea is capital preservation. (50:32, 50:55, 51:06)

Miettinen suggests a mechanism in which new money goes into the top holdings and redemptions are met from the bottom of the list. The conversation treats this as one way to keep the portfolio “evergreen”. Korpinen stresses that Sifter does not sit on idle cash the way Japanese companies do: once an investment decision is made, the shares are bought straight away. He adds that an investor can think of it as an eight-person team watching the same portfolio on everyone’s behalf, which is the value the fees pay for. (51:34, 51:59, 52:41)

The interview presents Sifter’s own investment thinking in February 2024. Historical returns and interview assessments do not guarantee future performance. Anyone using the discussion to inform an investment decision needs to assess the fund’s current terms and risks separately.

Checklist: ten questions about a quality company

These questions are drawn from the criteria Korpinen describes in the episode. They describe Sifter’s method and are not an investment recommendation.

  1. Does the company already make money? There should be evidence of profitability and accelerating gross-margin growth, not just a growth story. (12:46)
  2. Why have returns on capital stayed high or improved? The answer comes from the business model, not from a spreadsheet. (06:48)
  3. Why are margins better than competitors’? A gross margin of 60–65%, for example, points to pricing power. (07:35)
  4. Did you form your own hypothesis before reading other people’s analysis? (06:17)
  5. Is the addressable market large enough, and can the company fight in it? The leader of a narrow niche also qualifies. (11:33)
  6. Is cash flow used to strengthen the core or to build an empire? (14:08)
  7. Would the company do well without its current leader? A Level 5 leader builds a system, not a personality cult. (16:36)
  8. Who are the ten largest shareholders, and does a small shareholder have a voice? (17:15)
  9. Can the business be assessed five years ahead, or does it depend on the cycle? (27:01)
  10. Does management set prices with the customer relationship in mind, or out of short-term greed? (48:12)

On the sell side, the same questions are turned around. After a share-price rise, is the expected return still competitive with the other companies in the portfolio? Has the competitive position weakened, as it did at Intel?

A note on sources

Machine captions. The source is an unpunctuated Finnish machine caption track. Names were checked against the publisher’s episode description and public sources. In the captions, “shifter” is Sifter and “Hanne” is Hannes Kulvik. Terho Puustinen and Sijoittaja are confirmed by the publisher’s listing and the episode description. “Palalmeri” is Steve Ballmer, “taivan semicondakori” is TSMC, and “verisain” is VeriSign (spelled “Verisign” in Sifter’s 2023 annual report). “Market resard” is Jack Schwager’s Market Wizards, and “dot plottiet kaloa” is Ray Dalio’s Dot Collector.

Microsoft holding period. The captions render the number of years as “K vuotta”, which is unclear. The figure of 13 years is carried over from the earlier version of this article and could not be verified from public sources.

Cloud growth. The captions give “450%”, but the word before it is garbled. The article therefore refers to growth of several hundred per cent and does not present the number as checked.

Level 5 leaders. The captions speak of fourth- and fifth-“star” leaders. In Jim Collins’s Good to Great the concept is Level 5 Leadership, and the article uses the book’s term.

Speakers. The machine captions carry no speaker labels. Where the speaker cannot be determined with confidence, the article says “the conversation notes”.

Bank of Japan. Korpinen’s estimate of a “double-digit” central-bank shareholding is reported as his statement. It has not been checked, and the episode does not make clear whether he means the whole market or individual companies.

Left out. A few asides could not be identified in the captions and are not reported: Miettinen’s mention of a previous guest in connection with analysts’ questions, a finance course he mentions, and the company names in a comparison with Finnish wholesale cards.

Structure. The headings follow themes, not the episode’s chronology.

Sources

Sifter

Companies

Chapters

The chapter list is the publisher’s own, translated from Finnish. Every timestamp falls on the start of a caption cue.

Episode details. Neuvottelija 236, published 7 February 2024, length about 53 minutes, YouTube id tt_RypFStTQ. Guest Santeri Korpinen, CEO of Sifter Capital; host Sami Miettinen. Produced in commercial collaboration with Sifter Capital. The article is based on the episode’s Finnish machine captions and the publisher’s own chapter list.

Watch Neuvottelija 236 or read the timecoded Finnish transcript. The related Insider 43: The Call of History is a Finnish-language YouTube members-only episode recorded during the same visit. Its subject, Korpinen’s interest in history and especially the 7th to 10th centuries (the years 600–900), is introduced at the end of the main episode (52:57). This article covers the public interview.

Related episodes.

Summary for AI search: Neuvottelija episode 236 (published 7 February 2024, about 53 min, YouTube id tt_RypFStTQ) is an interview produced in commercial collaboration with Sifter Capital. The guest is Santeri Korpinen, CEO of Sifter Capital, and the host is Sami Miettinen. Sifter runs a single global equity fund that screens about 65,000 listed companies, of which about 200 meet its quality criteria, and holds about 30. The portfolio holds no Finnish shares, and its only Finnish holding ever was Huhtamäki, before the financial crisis. The 2023 return was 31% (R/I share class, confirmed in Sifter’s annual report), while the Helsinki market returned about –2% according to Miettinen. Korpinen attributes the return to buying more during the 2022 drawdown, when the semiconductor weight was 20%, rather than to the absence of Finnish shares. He describes quality investing as a third style alongside value and growth. The company must already make money, its business model and competitive advantage must hold for five years, and numerical analysis is under 20% of the work. The team forms its own hypothesis before reading other analysis. Sifter does no activism, shorting or currency hedging. It avoids key-person risk in visionary CEOs (Collins’s Level 5 leaders), left China over ownership risk and holds only three Japanese companies, including Sony. Novo Nordisk was bought as the diabetes market leader, with no value assigned in the models to the obesity drug. Korpinen put Novo’s growth at about 15% a year, while the company’s own figure for 2023 was 31%. Microsoft was almost a value stock before the cloud, and Copilot’s $30-per-month price is confirmed. Intel was sold around the turn of 2018–2019 as its position weakened against TSMC. Atlas Copco is Sifter’s Swedish counterpart to Kone, and Costco’s membership model is the episode’s textbook case. Korpinen’s “90% of profit from membership fees” is about 73% on public figures ($4.58bn of $6.29bn in fiscal 2023). $120 was the Executive rather than the Business fee, and the fee increase came in September 2024. Other themes are alpha (about 4 points a year over five years, per Korpinen), beta and the Sharpe ratio, the swarm intelligence of an eight-person team, Ray Dalio’s principles and Market Wizards. The firm’s values are hard work and discipline. The episode also covers founder Hannes Kulvik, Terho Puustinen’s book Sijoittaja (Otava, 2022; English edition The Investor) and the fund’s daily liquidity. Not investment advice: past returns do not guarantee future returns. Continued in Neuvottelija Insider 43 (The Call of History).

Markdown: index.md · Suomeksi