Writing an analysis: thesis, figures, risks and a call

What an analysis is

An analysis explains a company, a market or a geological question and ends in a conclusion. Unlike a news summary, it takes a position: what you think will happen, and why.

The analyses are published on the research pages in three categories: energy equities, shipping and geology. They are written by students in the research teams and are student work, not investment advice, which is also stated on every analysis. Each analysis gets its own page that can be shared and found in search engines.

A good analysis has four parts, and the sections below go through them one by one: a thesis, the figures that support it, the risks, and a call that can be checked. Finally come the publishing itself and the rules that apply when you write about shares.

The thesis: one claim the reader can test

Start with one sentence that says what you think and why. Everything else in the analysis should support, qualify or challenge that sentence. If you cannot write it, the analysis is not ready yet.

• Specific. "The company is well run" is not a thesis. "Cash flow will rise when the two new fields start producing, and that is not priced in" is.
• Testable. It must be possible to say afterwards whether you were right. That requires a direction and a time horizon.
• Reasoned. Explain the mechanism: what will happen, and why it affects the share price, the rate or the outcome.
• More than what everyone knows. If it is already in every news article, it is probably priced in. Write what you think others are missing, or why the common view is wrong.

The figures and the sources

The figures should show that the thesis holds. Use primary sources where you can: companies' annual and quarterly reports, the Norwegian Offshore Directorate for fields and wells, the Baltic Exchange for freight rates and the exchange for share prices. State where each figure comes from and which date it refers to, so the reader can check it.

• Key figures have their own field in the form, one per line with a name and a value separated by a colon, for example "Price target: NOK 450". They are shown at the top of the analysis.
• The valuation should be shown so the reader can redo it: which multiples you compare with, or which assumptions a net present value rests on. See the guides on valuing energy stocks, reading an annual report and valuing a discovery.
• Separate facts from opinions. "Production was so much in the second quarter, according to the quarterly report" is a fact with a source. "Production is likely to rise" is an opinion and should read as one.
• Show the calculations. A short worked example often beats a string of adjectives. If a figure appears in the conclusion, the reader should be able to find out where it comes from.

Risks: what would make you wrong

An analysis that only argues one side is a sales pitch. Write what could make the thesis wrong, and how likely you think that is.

• What would disprove the thesis? Name specific events or figures that would show you are wrong, so the reader knows what to watch.
• Sensitivity. How much does the conclusion change if the oil price, freight rates or interest rates move the other way? A simple table with two or three scenarios says a lot.
• Timing. A correct thesis can still turn out wrong if what you expect happens after the end date. Choose a time horizon that fits what you expect.
• Company-specific risk. Debt, project delays, tax and politics, and dependence on one field, one ship or one customer.

The call: how the analysis is followed up

When an analysis tips a stock or a commodity, or says that something will happen, it is entered as a call in the form when the analysis is published. The call is locked, and the outcome counts towards the team's track record, shown on the home page and on the research pages. The analysis itself shows what it called and how it went.

• Stock or commodity. You give the symbol as Yahoo Finance uses it, for example FRO.OL for Frontline on Oslo Børs or BZ=F for Brent, whether the price will go up or down, and an end date between one week and two years ahead. The symbol is checked against market prices before the analysis can be published. A price target is optional.
• Event. You write what it concerns and what will happen by the end date, for example that a company takes an investment decision. After the end date the board decides whether it happened, once.
• How a hit is measured. For a stock, the change is measured from the closing price on the day the analysis is published to the closing price on the end date, including dividends. It is a hit if the price moved the way you called it. The return is also shown against the Oslo Børs main index (OSEBX) over the same period.
• Locked. A call cannot be changed. A mistake can only be corrected on the same day, by the board, so the misses stay.

If the analysis has no tips or calls, you tick a box saying so. The form suggests the stocks mentioned in the text, so no tip is forgotten.

Publishing step by step

1. Log in on the research page with the account you have been given as a writer. The "Publish analysis" button at the top opens the form.
2. Category and author. Choose equities, shipping or geology. As the author you can give your name or the team.
3. Title and standfirst. The title should be short and specific. The standfirst is one or two sentences that sum up the conclusion.
4. Key figures and the analysis itself. Separate paragraphs with a blank line.
5. An English version is optional. Fill in the title, standfirst and text in English, all three, and the analysis also gets an English page.
6. An image is optional.
7. Calls in the analysis. Enter the tips, or tick the box saying the analysis has none.

The board can give approved members early access to an analysis. Members can then read it before it opens to everyone, up to and including a date the board chooses.

Fair and within the rules

Finanstilsynet, the Norwegian financial supervisory authority, has made clear that the rules on investment recommendations apply to professionals and non-professionals alike, including private individuals who publish recommendations about shares. The rules are in Article 20 of the Market Abuse Regulation (MAR). In short, Finanstilsynet says:

• Separate facts from your own opinions, and give your sources.
• Be objective, and do not mislead about the value or price of the share.
• Disclose your own interests and conflicts of interest, for example if you own the share you are writing about.
• A disclaimer does not exempt you from the rules.

So write "in my view" or "the analysis suggests" where something is an opinion, and do not write "buy" or "sell". Read Finanstilsynet's own pages, linked below, before you publish an analysis about a share.

Sources

• Finanstilsynet: Investment recommendations (in Norwegian) – who the rules apply to, objectivity and conflicts of interest
• Finanstilsynet (2022): The rules on investment recommendations also apply to private individuals (in Norwegian) – facts and opinions, sources, holdings and disclaimers
• Euronext: Oslo Børs Benchmark Index (OSEBX) – the index returns are compared with
• Yahoo Finance – the prices and symbols the calls are followed up with

The guides are written by OESA students to explain concepts and how things fit together. They are not investment advice. Figures that are not definitions carry a source and a date; if you find an error, let us know at philipstave@oesa-global.com.