Investing ideas, with the method shown
Thirty live screens across 7,000+ companies and 18 exchanges. Every one tells you the exact criteria it used, and, unlike every other list on the internet, the circumstances in which it is wrong.
Not the AI Picks. The dashboard picks six specific names and argues each case with its risk attached. These are whole screens: live lists of every company that passes a stated test, with the method printed. How the six are chosen
See the six picksEvery screen re-runs against the latest data load. Data as of Aug 13, 2026.
Undervalued on cash flow
Companies trading at least 25% below the fair value our published earnings-power model puts on their profits.
When it fails: A valuation model is an argument, not a measurement. Ours is an earnings-power calculation: it multiplies normalized earnings by a capped growth assumption, so it says nothing about the shape of the cash flows behind those earnings and it understates any company whose value is mostly future growth.
Quality at a fair price
Good businesses, not cheap ones. A wide moat and a strong rating, at a multiple that has not run away.
When it fails: A moat score is our estimate of durability, not a fact. Moats erode, and they erode quietly, usually showing up first in a falling gross margin rather than in any rating.
Wide moat compounders
Businesses with a structural reason competitors cannot take their profits away. The rarest and most valuable thing in investing.
When it fails: Great products get copied. Only structural advantages survive, and telling the two apart is the whole difficulty.
Deep value, with a pulse
A very low earnings multiple is usually a warning. This screen demands a low multiple AND evidence the business still works.
When it fails: This is where value traps live. A low multiple can persist for a decade, and being right about the value while nothing ever closes the gap is indistinguishable from being wrong.
Dividend powerhouses
Yields above 3%, filtered for the one thing yield screens never check: whether the company can actually afford it.
When it fails: A high yield is the market's opinion that the dividend will be cut. Sometimes the market is wrong. Usually it is not.
Growing dividend payers
A modest yield attached to a growing business beats a large yield attached to a shrinking one, every time, over any horizon that matters.
When it fails: Growth can stop. A dividend growth thesis is a growth thesis first, and if the growth fails you are left with a mediocre yield.
High return on equity, low debt
A high ROE is easy to fake with borrowing. This screen demands the return without the leverage, which is a much shorter list.
When it fails: A high ROE with a small equity base can also mean the company has bought back a lot of stock, which is fine, or written off a lot of assets, which is not.
Fortress balance sheets
Companies that owe almost nothing. They will survive things that kill their competitors, and they can buy the wreckage afterwards.
When it fails: A pristine balance sheet can also mean management is too timid to invest, and an under-levered company can be a lazy one.
High margin machines
Companies that keep an unusually large share of every pound they sell. That is pricing power made visible.
When it fails: Compare margins only within an industry. Software runs at 80% and groceries at 3%, and neither fact is interesting on its own.
Growth at a reasonable price
Real growth, without the multiple that usually comes attached to it. The middle ground between value and growth, and the hardest place to find anything.
When it fails: A modest multiple on a fast grower is usually the market disbelieving the growth. Sometimes it is right.
Hypergrowth
Revenue compounding above 30% a year. Extraordinary, unsustainable by definition, and occasionally the beginning of something enormous.
When it fails: Growth is priced. These companies are almost never cheap, and the multiple embeds an assumption you should make explicit before buying.
Low volatility defensives
Companies that move less than the market and pay you to wait. The holdings that let you sleep, and let you hold everything else.
When it fails: Low beta is not low risk. Beta measures co-movement with the market, and a stock can have a very low beta while quietly going bankrupt on its own schedule.
Beaten down, but not broken
Good businesses whose share price has been punished. Sometimes the market is wrong. Sometimes it is early. The screen cannot tell you which.
When it fails: 'It is down 40% from its high' is not an investment case. The previous price is an anchor, not a valuation.
Highest conviction
The best overall scores in our universe, restricted to companies where we actually hold enough data to have an opinion worth reading.
When it fails: A single score compresses several judgements into one number, and compression loses information. Read what is underneath it.
AI and semiconductors
The companies building the AI boom. Some are near-monopolies with the best moats in industry. Some are commodities in a lab coat.
When it fails: In every technology build-out the equipment sellers get paid whether or not the buyers ever earn a return. Know which side you are on.
Data centres and power
AI's binding constraint turned out not to be chips. It is electricity, and the grid was not built for this.
When it fails: Regulated utilities are usually capped on what they may earn. Booming demand does not become booming profit if a regulator decides your margin.
Healthcare and pharma
Drug companies are engines that convert one expiring monopoly into the next. The patent cliff is the whole investment case.
When it fails: The moat has an expiry date printed on it. Look up the date before you buy.
Energy and oil
Violently cyclical, generously cash-generative, and structurally hated. All three facts are related.
When it fails: The P/E is lowest at the peak of the cycle, moments before earnings collapse. This is the most counter-intuitive rule in equity analysis and it catches everyone once.
Banks and financials
Almost everything you know about valuation breaks on a bank. Debt is the raw material, and the loan book is where they die.
When it fails: A bank trading far below book is not automatically cheap. The market may simply be correct that the loan book is worth less than stated.
Consumer staples
The companies that sell what people buy regardless of the economy. Dull, and dull is the entire point.
When it fails: Predictable does not mean growing. Many of these are ex-growth, and you are buying an income stream, not a compounder.
Aerospace and defence
Governments are rearming and the order books run for a decade. The customer cannot stop buying, and it also sets your margin.
When it fails: There is a ceiling on how good this ever gets. Do not pay a growth multiple for a margin-capped business.
Mining and materials
Price takers in a violent cycle, with balance sheets that decide whether they survive the bottom of it.
When it fails: The P/E inverts, as with every cyclical. Cheapest at the top.
Recent IPOs
Companies that have listed recently. The most exciting day to buy one is almost always the worst one.
When it fails: The first-day pop went to whoever was allocated stock at the offer price. That was not you.
UK stocks worth a look
The FTSE is a basket of multinationals that happen to be listed in London. That distinction changes everything about what you are buying.
When it fails: In a high-yield market, the highest yields are still the sickest companies. Check the payout ratio every time.
South Africa and the JSE
A commodity and currency market wearing an equity market's clothes. The index can rise on a good day in Beijing.
When it fails: Currency is half your return. A JSE share up 20% in rand can be flat in dollars.
Japan: the governance unlock
Companies that traded below the cash on their own balance sheets for thirty years, now being told to do something about it.
When it fails: A weak yen flatters exporters and inflates the index in yen terms. In dollars the same rally can look far less impressive.
India: growth, already priced
The growth story everyone agrees on, which is exactly the problem. Consensus is expensive.
When it fails: Domestic monthly savings flows provide a persistent, price-insensitive bid. That means the market can stay expensive far longer than a foreign investor expects.
China and Hong Kong
Enormous growth, and a structure in which you may not legally own the company you think you bought. Both are true at once.
When it fails: In many Chinese ADRs you own a Cayman shell with contractual claims on the profits of the operating company, not the company. That structure has never been meaningfully tested in a Chinese court.
Dividend aristocrats
Companies with long, unbroken records of raising their dividend. The record is the point: it is evidence, not a promise.
When it fails: A long streak creates enormous pressure to maintain it, which can push a company to keep paying when it should be reinvesting or repairing its balance sheet.
Mega-cap leaders
The largest companies on earth. Enormously profitable, exhaustively analysed, and mathematically constrained in how much bigger they can get.
When it fails: If you own an index fund you already own a great deal of these, whether you meant to or not. Check the top ten weights before adding more.
None of these quite right?
Build your own screen over the same 7,000+ companies, on any combination of moat, valuation, growth, debt and geography. Free, no account needed to look.
Open the screener