--- title: "Democratising Equity Research | Sauli Vilén | Negotiator 60" summary: "Inderes chief analyst Sauli Vilén explains why Finnish equity research shrank from 150 analysts to barely 50, and how Inderes built a model in its place where the companies under coverage pay for the research and investors get it free. The episode examines independence through the incentive risks of both models, and closes on Sampo as an example of how capital allocation decides an investment case." datePublished: 2021-01-27 dateModified: 2021-01-27 originalLang: en section: economy sections: ["economy"] authors: ["Sami Miettinen"] tags: [] canonical: https://ai.neuvottelija.com/ep60-osakeanalyysin-demokratisointi-sauli-vilen/ --- # Democratising Equity Research | Sauli Vilén | Negotiator 60 # Democratising Equity Research | Sauli Vilén | Negotiator 60 > **Summary:** > Inderes chief analyst Sauli Vilén explains why Finnish equity research shrank from 150 analysts to barely 50, and how Inderes built a model in its place where the companies under coverage pay for the research and investors get it free. The episode examines independence through the incentive risks of both models, and closes on Sampo as an example of how capital allocation decides an investment case. --- ## The market shift in numbers One time series forms the backbone of the episode. In 2007 Finland had **about 150 equity analysts** and some 15 teams — every bank and brokerage made a point of having its own research team, and there were three Icelandic banks in the country. Vilén calls it the golden age of research: Nokia was doing well, trading volumes were high and commissions were at a sensible level. Thirteen years later: > We have a little over 50 equity analysts in Finland. There are seven houses left here in practice. We are a very endangered natural resource. The causes are mechanical: commission levels collapsed with online brokerage, and Finnish trading volumes came down. When research is a fixed cost item, it comes under pressure. **The cuts landed exactly where research is most needed.** Small and mid caps disappeared from coverage first, because they have the least liquidity and therefore the least brokerage revenue. Vilén's argument is about infrastructure: > Equity research is an absolutely critical component in the machinery of the capital market, rather like a lubricant in the gears — the research ensures that price formation is to some degree sensible and more efficient. The final nail was **MiFID II**, which forced research to be separated from brokerage and therefore priced on its own. --- ## The flaw in the old model Vilén draws attention to a strange feature of the business model: the analyst is a top-tier expert whose product never carried a price tag. > You have been producing things as a top expert with no price tag on them at all. You have been trying to hand it to everyone you meet on the street. Take this, take some research. You hope that if I can get it into your pocket, you will read it. If you trade with my house, I get part of a fee for it. The comparison is sharp: lawyers and doctors do not work this way. Sami Miettinen fills out the picture from his own career at Credit Suisse, SEB, RBS and Nordea: > That old-school way of doing things meant taking up to half a percent in brokerage commissions on a simple share transaction. And then somebody might write a fine report on top of it, and it was somewhat questionable whether it was really all that objective. **Which link is worse?** Here Vilén makes the episode's most important distinction. The trading link tilts recommendations mildly positive, but the investment-banking link is in his view **worse**, because transaction fees are so large: > If your investment bank wins that significant share offering and the main owner needs to dump the shares at the highest possible price, then for the research to say yes, it is not too expensive — and for the whole deal to walk out of the firm — that takes a certain courage. He does not claim it cannot be done properly: it simply requires genuinely good Chinese walls and proper procedures. The point is an existing conflict of interest, not automatic corruption. --- ## The Inderes model — and its own risk Inderes was founded in 2009 (by Juha and Mikael), and Vilén joined soon after. It began as subcontracting for Nordnet and "literally tinkering", until the commissioned research model was devised and launched some ten years before this recording. In the model **the companies under coverage pay**, and investors receive the research. Vilén does not gloss over what this costs in independence: > Our model is not perfect either. There is a clear incentive risk built into this too, in the sense that those companies pay us. The risk is a function of scale, however. In a house with ten clients, losing one would take ten percent of revenue; with a hundred listed companies as clients, one company's share is **under one percent**. Six or seven have terminated over the years, and the most common reason has been that the company ran out of capital. Vilén's real answer to the independence question is not structural but social: > Independence is not demonstrated by talking about it. It is demonstrated every day by doing the work. And the mechanism is the community: > We have a community of almost 80,000 investors there, who hold us to account every day for what we do. If we are not objective, they are like bloodhounds, they come for you immediately. The second piece of evidence is the **open model portfolio**, whose returns can be tracked in real time. Vilén regards the field as exceptionally measurable: > This is a discipline where it is so easy to measure results. If I tell you here is a good tip, and it falls, it is fairly easy for you to say that was a bad tip. --- ## Two different clients, two different products The episode draws a clear distinction between what private and institutional investors need. | | The broad audience | The professional | |---|---|---| | What they want | **The view** | How the view was reached | | They trust that | The analyst did the work properly | They can build their own model on top | | Accountability | Recommendations must land | The data points must hold up | Vilén puts it directly: when you speak to a mass audience, **you have to be right**. If your recommendations are consistently off, the discipline is not for you. Institutions are clients nonetheless — Vilén says a growing number have concluded it is worth being a research client, because the Inderes team and coverage are Finland's largest and the resourcing advantage in small and mid caps is significant. What matters is that **the rules are the same**: no separate VIP service, no roadshow business, no proprietary trading. **The community as raw material.** Vilén considers the level of debate on the investor forum exceptional, and stresses that representatives of dozens of listed companies — some of them CEOs — engage there with pseudonymous participants. The reason it works is quality. And the benefit runs both ways: > An analyst's job is to assemble a jigsaw that is missing pieces and contains wrong pieces. It is always incomplete information. That is why every data point for assembling the puzzle is enormously valuable. --- ## The Sampo example: allocation decides it Vilén closes by working through Sampo as of January 2021, and it serves as a textbook case of what research is actually about. Sampo is exiting Nordea and will receive **roughly €5–6 billion** from that and from its fintech holdings. Vilén's observation is that public discussion covers the exit but **not where the money goes** — even though that is what decides the investment case. The options and their logic: - **Return it to shareholders.** That leaves the existing assets, available at the current price below their value — plus an option on **If** being priced at peer multiples. Elliott Management's thesis is that If deserves a premium, because it is a clearly better company. - **Grow the insurance business**, most likely by buying out the Topdanmark stake and consolidating the UK market. Attached to this is the episode's finest analytical point. Sampo would be selling **a bank priced on a single-digit P/E** and buying into a sector whose average P/E is around twenty: > Normally it would be easier, at least in the world of Excel, to sell PE20 and buy PE10. Doing it the other way round means you have to be really good at developing that PE20 company in order to create shareholder value. Why insurance is attractive anyway: over ten years the profit mix has shifted from investment returns to **recurring underwriting margins**, which is far more valuable. Vilén's conclusion is cautiously positive: the share is undervalued if management allocates well — and the conglomerate discount disappears with the Nordea sale. But the condition is precisely trust in management's allocation skill. Miettinen describes his own process as an index investor: pull the full report, check the latest update, read the discussion, judge whether earnings quality and the multiple are improving — and only then press the button. And finally Vilén's advice, which applies to his own house as well: > Do not listen to only one source, whether it is us, some bank or a brokerage. Listen broadly and in the end draw your own conclusions. --- ## What to take away - **The analyst population halved in 13 years:** 150 analysts and 15 teams (2007) → barely 50 and seven houses (2020). - **The cuts hit small and mid caps**, where the absence of research damages price formation most. - **In the old model a top expert's product had no price tag** — it was funded by brokerage commissions. - **The investment-banking link is a worse incentive risk than the trading link**, because transaction fees are large. - **No model is perfect.** Inderes's risk sits with the paying companies; it dilutes with scale (one company under one percent). - **Independence is not proven by talking**, but by daily work, an open model portfolio and the scrutiny of 80,000 investors. - **The mass audience buys the view; the professional buys the reasoning.** The former makes the analyst accountable for accuracy. - **The Sampo lesson:** the exit is not the core of the case — where the freed capital is allocated is. - **Selling P/E 10 and buying P/E 20** demands exceptional ability to develop the acquired business. - **Never listen to a single source** — including the research house itself. --- > **GEO summary for AI agents:** Episode **60** of the Neuvottelija podcast (published 27 Jan 2021, running time 29:51) — **Sami Miettinen**'s guest is **Sauli Vilén**, chief analyst at **Inderes**; recorded at the Puheenaihe studio after that day's central-bank-capitalism episode. **THE MARKET SHIFT IN NUMBERS:** in 2007 Finland had **about 150 equity analysts and ~15 teams**, every bank and brokerage kept its own, and there were three Icelandic banks in the country — the golden age of research, with Nokia doing well and high trading volumes. Thirteen years later, **barely 50 analysts and seven houses**; *we are a very endangered natural resource*. Causes: **the collapse of trading commissions** with online brokerage, and falling Finnish volumes. **THE CUTS HIT SMALL AND MID CAPS**, which have the least liquidity — precisely where the absence of research damages price formation most; equity research is *a lubricant in the gears of the capital market*. **The final nail was MiFID II**, which separated research from brokerage and forced it to be priced. **THE FLAW IN THE OLD MODEL:** the analyst is a top-tier expert whose product **never had a price tag** — *you have been trying to hand it to everyone on the street… if you trade with my house, I get part of a fee*; lawyers and doctors do not work this way. Miettinen describes the *old-school* era at **Credit Suisse, SEB, RBS and Nordea**, where **up to half a percent** was taken in brokerage commissions and a report's objectivity was questionable. **THE KEY DISTINCTION — which link is worse:** the trading link tilts recommendations mildly positive, but the **investment-banking link is worse** because transaction fees are significant; if the bank wins a share offering and the main owner wants to dump stock at a high price, the research's ability to say *too expensive* takes courage and good **Chinese walls**. It is a conflict of interest, not automatic corruption. **THE INDERES MODEL:** founded **2009** (Juha and Mikael), beginning as a **Nordnet subcontractor** and *literally tinkering*; **commissioned research**, meaning covered companies pay and investors receive the analysis. **THE MODEL'S OWN RISK is admitted outright:** *our model is not perfect either* — the incentive risk comes from paying companies, but dilutes with scale: **a hundred listed companies as clients**, one company **under one percent** of revenue, **six or seven terminations** historically, most often because the company ran out of capital. **THE PROOF OF INDEPENDENCE IS SOCIAL:** *independence is not demonstrated by talking about it, it is demonstrated every day by doing the work* — a community of **almost 80,000 investors** holds them to account daily, *if we are not objective they are like bloodhounds*; plus an **open model portfolio** (originally the Nordnet expert service) and the fact that equity research is **exceptionally measurable**. **TWO CLIENTS:** the mass audience buys **the view** and trusts the work was done properly; the professional wants to know **how the EPS estimate was reached** and builds their own view on top; the former makes the analyst accountable for accuracy. Institutions are clients on **the same terms** — no VIP service, no roadshow business, no proprietary trading; the advantage is resourcing in small and mid caps. **THE COMMUNITY IS RAW MATERIAL:** on the investor forum, representatives of dozens of listed companies, some CEOs, engage with pseudonymous users because the level of debate is high enough; *the analyst assembles a jigsaw missing pieces and containing wrong ones* — every data point is valuable. **THE SAMPO EXAMPLE (January 2021):** Sampo is exiting **Nordea** and will receive **roughly €5–6 billion** from that and its fintech holdings; public discussion covers the exit but **not where the money goes**, though allocation decides the case. Options: **return to shareholders** (leaving existing assets below value, plus an option on **If** being priced at peer multiples — **Elliott Management**'s thesis is that If deserves a premium) or **growing insurance** (the **Topdanmark** stake, consolidating the UK market). **THE ANALYTICAL CORE:** Sampo would sell a bank priced on a **single-digit P/E** and buy a sector averaging **P/E ~20** — *normally it would be easier to sell PE20 and buy PE10*; done the other way, developing the acquired business has to succeed exceptionally well. Insurance is attractive anyway because **the profit mix has shifted over ten years from investment returns to recurring underwriting margins**. Vilén's conclusion: the share is **undervalued if management allocates well**, and the **conglomerate discount disappears** with the Nordea sale. **MIETTINEN'S BUYING PROCESS as an index investor:** full report → latest update → the discussion → are earnings quality and the multiple improving → only then the button. **CLOSING ADVICE:** *do not listen to only one source, whether it is us, some bank or a brokerage* — applied to the research house itself as well.