Methodology
How Olbera's Conviction Score works
How strong is the research case? Olbera's Conviction Score answers on a 0–100 scale, bringing together five areas of market evidence before the research is explained by AI.
The short version
A 0–100 research score. It shows how strong the strongest available research case is.
Five areas of market evidence. Technical, news and sentiment, macro, fundamental, and market flow and positioning.
Technical establishes the base reading. The other four areas can strengthen or weaken the case.
Conviction has to be supported by the overall evidence. Evidence pointing the other way weakens it, and no single signal can carry the score on its own.
AI explains the research rather than deciding it. The direction, the score and the selection are established first.
Not every area applies to every asset. Where one doesn't, it's left out — never guessed.
Why one kind of evidence is never enough
Markets are multi-dimensional, and the evidence rarely lines up neatly. A strong technical picture can sit alongside weak fundamentals. Strong fundamentals can sit inside deteriorating market conditions. A macro environment can support or challenge an otherwise strong research case. And positioning can add context that price alone cannot provide.
The point isn't that every signal must agree. It's that the evidence is evaluated together. Combining technical and fundamental analysis is a start; news, the economic backdrop and positioning belong in the same picture. That is what a multi-factor approach is for.
A screener asks one question at a time. A research process asks whether the whole picture holds up.
From evidence to conviction
Several kinds of evidence converge on one research conclusion. Technical analysis sets the base reading, the other areas test it, and AI only explains the result afterwards.
Evidence
Market data, company data, news, and economic and positioning information.
Base reading
TechnicalTechnical analysis establishes which way the evidence leans, and how strongly.
Additional evidence
News, macro, fundamental and market-flow evidence, where it applies, strengthens or weakens the case.
strengthen or weaken
- News & Sentiment
- Macro
- Fundamental
- Market Flow / Positioning
where applicable
Caution
A separate check that can reduce conviction when conditions look stretched or unusual.
Conviction Score
A 0–100 reading of how strong the research case is.
Research conclusion
The strongest research case in each asset class, with its direction.
Research State
Context over time: how the picture has developed across recent readings. A label, not part of the score.
AI explanation
Written afterwards, in plain language. It can't change the result.
Five areas of market evidence
Each area looks at a different kind of evidence, from its own data. Technical analysis establishes the base reading; the other four test it.
Technical
Base reading- Looks at
- How price has behaved: direction and strength of trend, momentum and market structure, plus the breadth of that evidence.
- Contributes
- The base reading and a check on whether the trend looks stretched.
- Applies to
- Every asset Olbera covers.
- On its own, can't
- Explain what is driving price or what comes next.
News & Sentiment
Additional evidence- Looks at
- Recent, relevant developments and how they are being received.
- Contributes
- Additional evidence that can support or weaken the case.
- Applies to
- Assets with relevant coverage.
- On its own, can't
- Create a view when the base reading is neutral.
Macro
Additional evidence- Looks at
- The broader economic backdrop including monetary policy, real yields, growth, liquidity, risk appetite and currency conditions.
- Contributes
- Whether the environment supports or works against the research case.
- Applies to
- All asset classes, with different relevance depending on the asset.
- On its own, can't
- Choose individual assets.
Fundamental
Additional evidence- Looks at
- Company quality including growth, profitability, cash generation, financial strength and whether valuation already reflects it.
- Contributes
- Whether the business behind a stock supports the research case.
- Applies to
- Stocks only.
- On its own, can't
- Say much about the short term.
Market Flow / Positioning
Additional evidence- Looks at
- How large participants are positioned, including positioning and ownership information where reliable data exists.
- Contributes
- Context that price alone cannot show.
- Applies to
- Where reliable positioning data exists.
- On its own, can't
- Form a complete research view.
Agreement strengthens a case. Evidence pointing the other way weakens it. A missing area is never treated as support.
Why conviction has to be earned
Olbera is built to reduce noise, not amplify every possible signal. A strong score has to be supported by the evidence as a whole.
Agreement strengthens the case
When different areas of evidence point the same way, conviction rises.
Evidence pointing the other way weakens it
Evidence that cuts against the research case lowers conviction. It does not raise it.
Stretched conditions cost conviction
When a trend looks mature or statistically unusual, conviction can be reduced, whichever way that trend points.
No single signal carries the score
One striking reading can't produce a strong score on its own. The case needs breadth and consistency across the evidence.
The score doesn't measure how exciting something looks. It shows how strong the strongest available research case is.
Where AI fits — and where it doesn't
Established first — without AI
- The underlying evidence
- The direction
- The Conviction Score
- The selection
Written afterwards — with AI
- A plain-language explanation
- A narrative that synthesises the established findings
- An explanation of what the research shows
AI explains the research. It doesn't decide the research.
The order matters. The evidence is read first. The direction, the Conviction Score and the selection are established. Only then does AI turn those findings into the explanation you read. It cannot change the score, the direction or the selection — it isn't given the ability to. It's there to make the research easier to read, not to replace it.
Research State
The Conviction Score reads today. Research State adds context over time: how the research picture has developed across recent readings.
Strengthening
The research picture has become more supportive across recent readings.
Stable
The research picture has remained broadly consistent across recent readings.
Weakening
The research direction remains intact, but recent readings have become less supportive.
Reversed
The research direction has changed materially from its recent position.
Building History
Not enough recent history is available to assess how the research picture is developing.
It's a label, not another score, and it doesn't change the Conviction Score.
What the Conviction Score is not
Not a probability. It doesn't say how likely anything is.
Not a guarantee. Olbera does not guarantee any particular outcome, return, or level of investment performance.
Not a promise of future performance. Past research or scoring is not a reliable indicator of future results.
Not a prediction of certainty. It describes the research case today.
Not personal financial advice. It does not take into account your individual financial situation, objectives, or needs.
Not an instruction to trade. You are solely responsible for your own investment decisions and their consequences.
Olbera provides general research and analysis for informational and educational purposes only. It is not personal or personalised financial advice, and does not take into account your individual financial situation, objectives, or needs. Olbera is not a financial adviser, broker, or licensed investment professional.
See the research
Every research case in Olbera is laid out the same way. Here is a fictional one, so you can see what each part means.
Example Company
EXMPL
Conviction
71
/ 100
Price trend and company growth are pointing the same way.
Key drivers
- Price has trended higher with steady momentum.
- Company growth has been running ahead of expectations.
What would change the picture
Would strengthen
- Broader conditions turning more supportive.
- Another area of evidence lining up behind the case.
Would weaken
- A reversal in the underlying trend.
- Developments that cut against the case.
Strengthening
The research picture has become more supportive across recent readings.
- 1
Conviction Score
A 0–100 reading of how strong the research case is.
- 2
Direction
Which way the evidence leans.
- 3
Research horizon
The time frame the research looks across.
- 4
Key drivers
The evidence behind the case, in plain language.
- 5
What would strengthen or weaken the case
The developments that would add to, or take away from, conviction.
- 6
Research State
How the picture has developed across recent readings.
Questions
Is the Conviction Score a prediction or a probability?
No. The Conviction Score describes how strong the research case is today, based on the evidence available. It isn't a forecast, a probability or a guarantee, and it says nothing certain about what happens next.
Does every asset get all five areas of evidence?
No. Fundamental applies to stocks only, and market flow applies only where reliable positioning data exists. News and macro apply with different relevance depending on the asset. When an area doesn't apply, it's left out — it's never guessed, and never treated as support.
Which parts of Olbera use AI?
The written explanation. The evidence, the direction, the Conviction Score and the selection are established first, without AI. AI then turns those findings into a plain-language explanation. It doesn't change any of them.
How does Olbera narrow the market?
Each day Olbera researches stocks, ETFs, forex and crypto, bringing the evidence for each asset together into a Conviction Score. Within each asset class, the strongest research case is selected, so you see one clear case per class rather than a list to sift through.
Can I see how a research case was reached?
Yes. Each research case shows the supporting evidence, the conflicting evidence, the key risks, what would strengthen or weaken the case, and its Research State. These details are available during your free trial and to subscribers.
See what Olbera is researching.
Your first 7 days are free. No card required.
No card required · $14.99 AUD/month after · Cancel anytime
AI-generated market analysis for educational purposes only. Not financial advice. Terms of Service · Privacy Policy
Methodology last reviewed: October 2026