What O'Neil actually did
William J. O'Neil (1933–2023) founded *Investor's Business Daily*. He did something most authors of investing books don't: he went back and studied what the biggest winning stocks looked like before they took off.
The sample spanned decades and hundreds of large winners. He wasn't looking for what *should* work. He was looking for what those stocks actually had in common at the point they started moving.
Take the first letter of each of the seven traits he found and you get CAN SLIM.
One thing up front: this is not a formula. Matching all seven guarantees nothing, and a stock matching three can still multiply. It's a checklist of characteristics, useful mainly for filtering out the names that don't look anything like a big winner.
C — Current Quarterly Earnings
The most recent quarter's EPS should be up at least 25% versus the same quarter last year.
Note that's versus *the same quarter last year*, not versus last quarter. Same reason as in the EPS article — many industries have seasonal patterns, so comparing this Q3 to Q2 picks up noise. Comparing to last year's Q3 is clean.
O'Neil's observation: by the time a big winner starts moving, earnings are usually already accelerating. Not about to improve — already improving, just not fully priced in yet.
The 25% threshold came from his US sample. It isn't a law. The point is "clearly growing" rather than "growing" — 3% and 40% are not the same thing.
A — Annual Earnings Growth
Earnings up each of the past three years, compounding at 25% or better.
C looks at right now. A asks whether this has been a genuinely profitable business over time.
A single explosive quarter can be one-off — a property sale, an FX gain, a subsidy recognised. Three consecutive years of growth is harder to fake; it usually means the business model is actually expanding.
O'Neil also looked at return on equity, with 17% as the bar. That number asks how efficiently the company turns shareholders' capital into profit.
N — New
A new product, a new service, new management — or a new price high.
The first three are qualitative and require reading news and filings. The fourth is quantitative, and it's counterintuitive: O'Neil argued for buying stocks making new highs, not cheap stocks.
Most people's instinct runs the other way — it's fallen a lot, it's cheaper, surely that's less risky. But his data showed that the stocks which went on to rise the most were frequently bought just as they broke to a new high.
The reason is supply. A new high means nobody overhead is trapped. Everyone holding the stock is in profit, nobody is desperate to get out at breakeven, and selling pressure is at its lightest. A stock making new lows has a wall of trapped holders at every level above.
The new-highs article covers the same idea, and the "New High" column in this tool is what flags it.
S — Supply and Demand
A smaller float, plus expanding volume on the breakout.
The same buying pressure hitting a stock with fewer shares outstanding produces a much larger price move. Simple supply and demand — limited supply, rising demand, price jumps.
So O'Neil favoured small and mid caps and expected less explosiveness from megacaps. Doubling TSMC requires a completely different order of capital than doubling a mid-cap.
The other half is volume. On the day price breaks the prior high, volume should expand noticeably — a common rule of thumb is 40% or more above the recent average. A breakout without volume means nobody really wanted it, and it often fails.
L — Leader, Not Laggard
Only buy the strongest one or two names in a sector.
This is the most overlooked item in CAN SLIM, and possibly the most valuable.
When a theme takes off, everything in the sector rises — but by very different amounts. The common thought is that the leader has run too far, so the second-tier name is cheaper with more catch-up room.
O'Neil's data says don't. Laggards lag for a reason, and when the theme cools, second-tier names fall harder than the leader. He used a Relative Strength rating and bought only names in the market's top 20%.
In this tool, the closest approach is: screen for strict bullish alignment in the screener, then look at the sector distribution. When several results share an industry, that industry is the current leadership — then pick the one or two rising fastest and making new highs first.
I — Institutional Sponsorship
You want institutions buying — but not already full.
Big moves need big money; retail flow alone doesn't build a real trend. So O'Neil wanted to see institutional ownership increasing, ideally with new funds starting positions over recent quarters.
But there's a ceiling: a stock institutions have already loaded up on is dangerous. The buyers have finished buying; what's left is sellers. Those names fall especially hard on bad news.
In Taiwan you can watch the three major institutional investors' net buying and ownership ratios; most trading platforms show this. It's not something this tool currently covers.
M — Market Direction
Three out of four stocks follow the market.
O'Neil called this the most important of the seven — and the reason most people lose money. However well you pick stocks, going long in a bear market loses.
The first six items pick stocks. M asks whether you should be in the market at all. When the market turns down, his advice was blunt: get out, hold cash, wait. Not rotate — stop.
The Market Barometer in this tool exists to answer M. Glancing at it before every screen is the cheapest risk control available.
Does it work in Taiwan? What needs adjusting
Translates directly: N (new highs), L (leaders and sectors), M (market direction), and the volume half of S. These are price and supply behaviours, and they don't differ much across markets.
Needs adjusting:
Reporting frequency differs. US companies report quarterly; Taiwan's quarterly filings lag by over a month. But Taiwan has monthly revenue disclosure, which the US does not. Monthly revenue YoY works as a leading indicator for C, far faster than waiting for the quarterly report — which is exactly why this tool has a "Monthly Revenue YoY" column.
Discount the float argument. Plenty of Taiwan small caps have a small share count *and* poor liquidity, so you can't get out. Check volume alongside size.
The 25% and 17% thresholds come from a US sample. Taiwan's market is dominated by electronics manufacturing, where the cycle swings harder than US software or consumer brands. Applying fixed thresholds may screen out everything at a cyclical low and pass everything at a high. Treating them as relative standards works better — find what's growing fastest in the market right now, rather than defending an absolute number.
Running it with this tool
Of the seven items, this tool helps with four:
- Start with M — open the Market Barometer and confirm bullish alignment. If it's bearish, don't
- Screen for N and L — screener set to top 300 by market cap plus strict bullish alignment, then read the "New High" column
- Use sectors for L — check the sector distribution of your results; the industry with the most names is the current leadership, and you want its strongest members
- Use monthly revenue for C — set Monthly Revenue YoY to the highest bracket to filter out names whose fundamentals haven't followed
The rest — A (three-year growth), S (float and volume), I (institutional flow) — require separate fundamental and ownership data this tool doesn't cover.
The honest part
A few things worth knowing.
It's built for bull markets. O'Neil said M mattered most precisely because this method has almost nothing to buy in a bear market. It is not an all-weather strategy.
The sample has a survivorship problem. "What did the eventual big winners look like beforehand" and "will stocks matching these criteria go on to rise" are two different questions. The first can be studied; the second has to be verified in live trading.
Applied strictly, very few stocks qualify. All seven at once is a single-digit list at any given time. Most people in practice lean on three or four — usually M, N, L plus earnings growth.
It requires stop-loss discipline. O'Neil's original method paired the criteria with a hard rule: cut any position down 7–8% from your entry, unconditionally. His logic was that when a CAN SLIM stock works it works big, so you can afford many small losses. Without that discipline the expected value of the whole method changes.
Who it's for
Suited to people trading swings to intermediate trends, willing to read financials, and able to take frequent small stops.
Not for buy-and-forget investors — CAN SLIM names are volatile and need watching. Not for value investors either, since it deliberately buys what has already risen and looks expensive.
If you want more of O'Neil's thinking, the quote cards on the home page include passages from him; they come round periodically.
Use this tool in line with your own investing experience.
Nothing here constitutes investment advice.