CAN SLIM Investing Model: How Growth and Momentum Work Together
The CAN SLIM investing model combines two forms of evidence that investors often separate: improving company results and confirmation from the market.
That combination is the framework’s defining feature. Rapid earnings growth can look impressive while a stock loses relative strength. A rising share price can attract attention while the underlying company produces little earnings progress. CAN SLIM asks whether several independent signals agree.
William O’Neil developed the method around seven ideas represented by C, A, N, S, L, I, and M. Applying the acronym responsibly requires clear definitions, synchronized data, and an understanding of what the method does not measure.
StockGeniuses translates six company-level components into a deterministic 0-10 diagnostic score. Market direction remains contextual rather than becoming another company score. Required inputs can also make the model Not Meaningful. Incomplete evidence does not earn a neutral assumption merely because the available numbers look strong.
This guide explains the philosophy, the seven letters, the exact scoring logic, and a dated Nvidia walkthrough. The example is not an investment recommendation. It demonstrates why exceptional reported growth can coexist with an unavailable final score.
CAN SLIM is an agreement test, not a growth checklist
The central CAN SLIM question is broader than “Is this company growing?”
It asks whether current earnings, annual earnings, business momentum, supply and demand, relative price leadership, institutional participation, and the wider market environment tell a coherent story.
These signals come from different places:
- income statements describe earnings and revenue
- share data describes dilution or contraction in supply
- trading volume shows participation in price movement
- price history measures leadership relative to other securities
- ownership data describes institutional sponsorship
- broad market evidence supplies context for individual-stock strength
Agreement does not prove future performance. It reduces dependence on any single attractive number. More importantly, disagreement becomes useful: strong earnings without leadership, or leadership without earnings, tells the investor exactly which part of the story remains unconfirmed.
That distinction separates CAN SLIM from approaches that treat investing styles as interchangeable. The guide to value, growth, and quality lenses explains that each lens answers a different question. CAN SLIM is primarily a growth-and-confirmation framework. It does not estimate intrinsic value, and it does not complete a qualitative business-quality assessment.
William O’Neil’s philosophy behind the method
William O’Neil’s public description of his firm’s methodology emphasizes a blend of fundamental, quantitative, and technical research. The philosophy rejects the idea that investors must choose between understanding a company and observing how the market treats its shares.
Three principles carry through the framework.
First, growth should be visible now, not only promised for a distant future. Current quarterly earnings matter because recent acceleration can reveal a change in the company’s economics before it appears in a long annual history.
Second, current improvement needs a longer foundation. One strong quarter can result from an easy comparison, temporary pricing, a tax event, or a one-time gain. Annual earnings add historical context.
Third, the market supplies information that accounting statements cannot. Relative price strength, volume, share supply, sponsorship, and broad market direction can reveal whether other participants are confirming or rejecting the fundamental story.
This does not mean price action is automatically correct. It means that ignoring it removes a distinct source of evidence. The relationship between price action and business quality is complementary: price behavior describes market response, while company evidence describes operating reality.
O’Neil’s philosophy is therefore not “buy whatever is rising.” It is closer to this: require growth, require evidence that the growth is meaningful, and look for market confirmation without confusing confirmation with certainty.
What the seven CAN SLIM letters mean
The original framework uses seven memorable categories:
| Letter | Core idea | Question it asks |
|---|---|---|
| C | Current quarterly earnings | Are recent earnings growing strongly, and is growth accelerating? |
| A | Annual earnings growth | Does the company have a meaningful multi-year earnings record? |
| N | New products, management, conditions, or highs | Is there a new force that can change the company’s trajectory? |
| S | Supply and demand | Does trading participation support the move, and is share supply expanding? |
| L | Leader or laggard | Is the stock outperforming a defined comparison universe? |
| I | Institutional sponsorship | Is professional ownership supporting, weakening, or remaining stable? |
| M | Market direction | Is the broader market environment supportive or hostile? |
Several categories are not directly machine-readable. “New” may refer to a product, executive, market, technology, regulation, or price high. “Leader” requires a defined universe, return period, and percentile method. “Institutional sponsorship” depends on delayed ownership reports and a reproducible trend rule.
A systematic implementation must therefore distinguish the original investing idea from the measurable proxy used to represent it. A proxy is not the original idea in numerical clothing; it is a narrower test chosen for consistency. Without that distinction, two analysts can apply the same acronym and quietly evaluate different companies.
Two evidence layers work together
The seven letters become easier to understand when divided into two layers.
Fundamental growth evidence
C, A, and part of N describe the company:
- current EPS growth asks whether recent profitability is expanding
- EPS acceleration asks whether that expansion is becoming faster or slower
- annual EPS growth tests whether the recent quarter sits on a longer record
- revenue acceleration helps determine whether business activity is also strengthening
These inputs protect the framework from relying on price alone. They also need interpretation. EPS can rise because of revenue growth, margin expansion, buybacks, tax effects, asset sales, or investment gains. The number is evidence, but the source of the number determines its quality.
Market-confirmation evidence
S, L, I, and M describe how securities and participants behave:
- volume indicates whether recent trading activity is unusually strong
- share-count change reveals whether equity supply is expanding
- relative strength compares the stock’s return with a defined universe
- institutional trend describes changes in professional ownership
- market direction places company-specific strength inside the broader regime
Market evidence can confirm a fundamental story, but it cannot repair a weak one. Strong fundamentals also do not guarantee leadership. CAN SLIM becomes most informative at the boundary between the layers: it shows whether operating progress and market behavior agree, not which side must eventually prove correct.
Investors who know how to read a stock analysis model separate those statements. A component score reports what a rule found in a defined dataset. It is not an instruction to buy, sell, or hold.
How StockGeniuses implements CAN SLIM
StockGeniuses converts C, A, N, S, L, and I into component scores from 0 to 10. Every weighted component contributes positively. M remains a contextual market signal with a 0% company-score weight.
| Component | Weight | StockGeniuses measurement |
|---|---|---|
| C – Current earnings | 25% | Quarterly EPS growth plus acceleration |
| A – Annual earnings | 20% | Valid 5-year EPS CAGR, with 3-year fallback and a consistency check |
| L – Leadership | 20% | 12-month total-return percentile in the Model-6 universe |
| S – Supply and demand | 15% | 50-day/200-day average-volume ratio, adjusted for share dilution |
| N – New or innovation | 10% | Quarterly revenue-growth acceleration as a reproducible proxy |
| I – Institutional | 10% | Institutional ownership trend |
| M – Market direction | 0% | Context shown separately from the company score |
The weighted result is rounded half up and clamped to the 0-10 range:
CAN SLIM Score = 0.25C + 0.20A + 0.20L + 0.15S + 0.10N + 0.10I
This is a StockGeniuses diagnostic implementation, not a claim that O’Neil’s original judgment process can be reduced perfectly to one formula. The proxies make the result consistent and auditable. The weights show how much each completed component contributes; they do not express confidence in the underlying data. A precisely weighted unreliable input remains unreliable.
C: current earnings and acceleration
Current quarterly EPS growth compares the latest quarter with the same quarter one year earlier:
Quarterly EPS growth = (latest EPS - year-ago EPS) / |year-ago EPS|
The prior quarter’s year-over-year growth is calculated the same way. Acceleration is the difference between the two growth rates.
- quarterly EPS growth of at least 25% creates a base C score of 8
- positive acceleration adds 2
- negative acceleration subtracts 2
- the component is clamped between 0 and 10
Comparing the same fiscal quarter reduces ordinary seasonality. The acceleration adjustment prevents a large but rapidly slowing growth rate from looking identical to one that is strengthening.
A: annual growth and consistency
The annual component uses a valid five-year EPS CAGR when possible, with a three-year fallback:
Annual EPS CAGR = (ending EPS / beginning EPS)^(1 / years) - 1
The score is 10 at 20% growth or more, 7 at 15% or more, 5 at 10% or more, and 2 below 10%. If the standard deviation of annual EPS growth exceeds 1.5 times its mean, the model subtracts 3 points.
The fixed test improves consistency, but it cannot diagnose every unusual earnings path. Investors still need to examine the years between the CAGR endpoints because a smooth compound rate can conceal contraction and rebound. The consistency penalty is a rule, not a declaration that an unpenalized series is economically smooth.
N: a measurable proxy for change
The original N is qualitative. StockGeniuses uses revenue acceleration to make the concept reproducible:
Revenue acceleration = latest YoY revenue growth - prior-quarter YoY revenue growth
Positive acceleration scores 10; zero or negative acceleration scores 5.
This proxy does not identify the actual new product, market, or strategic change. It asks whether a potentially important change is visible in reported sales momentum.
S: volume participation and share supply
The supply-and-demand component starts with:
Volume surge = 50-day average volume / 200-day average volume
- above 1.2 scores 10
- above 1.0 through 1.2 scores 7
- 1.0 or below scores 4
If shares outstanding increased by more than 2% year over year, 3 points are subtracted. This combines trading participation with a simple supply-side dilution check.
L: leadership requires a universe
Leadership uses a 12-month total-return percentile:
- 80th percentile or higher scores 10
- 70th through below 80th scores 8
- 50th through below 70th scores 5
- below the 50th percentile scores 2
The comparison universe must contain at least 50 securities and follow the deterministic Model-6 membership rule. A return without a universe is not a percentile. Changing the universe can change the component even when the company’s own price history remains identical.
I: institutional trend
Increasing institutional ownership scores 10, stable ownership scores 7, and decreasing ownership scores 3. Missing institutional data receives a neutral 7.
That neutral treatment is intentional. It avoids turning unavailable optional ownership data into a negative company claim. It does not make other required inputs optional.
Why market direction is contextual
The original M recognizes that broad market conditions affect the probability that individual trends persist. Even a company with improving results can struggle when market liquidity, risk appetite, or index structure turns hostile.
StockGeniuses keeps M outside the weighted company score. This prevents the same company from appearing fundamentally stronger or weaker merely because the market regime changed. The user can inspect company evidence and market context separately.
That design also reduces a common interpretation error. A supportive market does not improve earnings quality. A weak market does not erase operating progress. The two facts can coexist.
When the model must return Not Meaningful
CAN SLIM needs more than a few available metrics. The StockGeniuses implementation returns Not Meaningful when required evidence cannot support a reproducible calculation, including when:
- current or year-ago quarterly EPS is non-positive
- prior-quarter EPS acceleration cannot be calculated
- annual EPS history cannot support at least three growth observations and a valid 3-year or 5-year CAGR
- quarterly revenue acceleration is unavailable
- 50-day or 200-day average volume is unavailable
- current or prior-year shares outstanding is missing or invalid
- 12-month total return cannot be calculated
- the relative-strength comparison universe contains fewer than 50 securities
Not Meaningful is not a low score. It means the model’s question cannot be answered from the supplied evidence. Eligibility comes before weighting: a component that cannot be reproduced cannot be compensated for by excellent results elsewhere. Forcing a number would weaken the analysis, not complete it.
Nvidia example: strong growth, incomplete CAN SLIM evidence
Nvidia provides a useful real-world application because its recent financial acceleration is easy to source while its complete model result still depends on market datasets outside company filings. The example separates two questions that are often blurred: what the company reported, and whether the full market-confirmation layer can be reproduced.
The snapshot uses Nvidia’s Form 10-Q for the quarter ended April 26, 2026, filed May 21, 2026, annual results through fiscal 2026, and an analysis preparation date of July 14, 2026. Nvidia’s fiscal 2027 first quarter covers part of calendar 2026; the fiscal label should not be mistaken for a future period.
All EPS figures used below are GAAP diluted EPS adjusted to the same post-split share basis where necessary.
C is calculable and scores 10
Nvidia reported Q1 fiscal 2027 diluted EPS of $2.39, compared with $0.76 in Q1 fiscal 2026:
Current quarterly EPS growth = (2.39 / 0.76) - 1 = 214.47%
The preceding Q4 fiscal 2026 comparison was $1.76 versus $0.89:
Prior quarterly EPS growth = (1.76 / 0.89) - 1 = 97.75%
EPS acceleration = 214.47% - 97.75% = 116.72 percentage points
Growth exceeds 25%, creating the base score of 8. Positive acceleration adds 2. Nvidia’s C component is therefore 10 under the implemented rules.
A is calculable and scores 10, with an important warning
The split-adjusted annual EPS series used for the five-year review is approximately:
| Fiscal year | GAAP diluted EPS |
|---|---|
| 2021 | $0.173 |
| 2022 | $0.385 |
| 2023 | $0.174 |
| 2024 | $1.19 |
| 2025 | $2.94 |
| 2026 | $4.90 |
The five-year CAGR is:
(4.90 / 0.173)^(1/5) - 1 = 95.18%
That rate produces an A base score of 10. The intervening year-over-year growth rates are approximately 122.54%, -54.81%, 583.91%, 147.06%, and 66.67%. Their mean is 173.07%, their population standard deviation is 216.92 percentage points, and the model’s penalty threshold is 259.61 points.
Because 216.92 does not exceed 259.61, no consistency penalty applies. The A component remains 10.
Mechanically, that result is correct. Economically, the sequence is plainly uneven. The example reveals a limitation of any fixed volatility test: an extremely high mean can also create an extremely high penalty threshold. The investor should preserve the score while still recognizing the contraction-and-rebound path.
Historical figures provide context rather than a forecast. The broader guide to historical performance in stock analysis explains why a strong past rate cannot be extended automatically into future years.
N is calculable and scores 10
Q1 fiscal 2027 revenue was $81.615 billion, compared with $44.062 billion one year earlier:
Current revenue growth = (81.615 / 44.062) - 1 = 85.23%
Q4 fiscal 2026 revenue was $68.127 billion, compared with $39.331 billion in Q4 fiscal 2025:
Prior revenue growth = (68.127 / 39.331) - 1 = 73.21%
Revenue acceleration = 12.01 percentage points using the unrounded growth rates
Positive acceleration produces an N score of 10. The result says reported sales momentum increased. It does not identify the durability of demand or prove that every source of growth will persist.
Part of S is visible, but S cannot be completed
Nvidia reported 24.221 billion shares outstanding at April 26, 2026, compared with 24.388 billion at April 27, 2025:
Share change = (24.221 / 24.388) - 1 = -0.68%
The share count declined, so no dilution penalty applies.
However, the S base score requires synchronized 50-day and 200-day average trading volume. Those inputs do not come from the filing and were not available in the repository’s source packet for this snapshot. A share-count calculation alone cannot produce S.
L cannot be reproduced without the Model-6 universe
Nvidia’s own price history would not be enough to calculate L. The model requires its 12-month total return ranked against the deterministic Model-6 universe, with at least 50 valid securities.
That universe’s membership is not available in the source set used for this walkthrough. Substituting the S&P 500, semiconductor peers, or a handpicked technology group would create a different model. The L component therefore remains unresolved.
Missing I would be neutral, but it cannot rescue the run
If institutional trend data is unavailable, the model assigns I a neutral score of 7. This protects an optional field from becoming a hidden penalty.
The missing average-volume and relative-strength inputs are different. They are required. As a result, this public-data subset returns Not Meaningful rather than a final CAN SLIM score.
| Component | Status | Reproducible result |
|---|---|---|
| C | Complete | 10 |
| A | Complete | 10 |
| N | Complete | 10 |
| S | Incomplete – volume averages unavailable | No component score |
| L | Incomplete – Model-6 universe unavailable | No component score |
| I | Missing optional data | Neutral 7 if the full run were otherwise eligible |
| Final score | Required inputs incomplete | Not Meaningful |
This is not a negative conclusion about Nvidia. It is a boundary around the claim. The evidence supports three strong component results, but it does not support the sentence “Nvidia has a high CAN SLIM score.” Analytical discipline means stopping between those statements.
The Nvidia earnings-quality caveat
The completed C component also needs interpretation. Nvidia’s Q1 fiscal 2027 other income was $15.929 billion, compared with an expense of $180 million one year earlier. The filing attributes most of the change to unrealized gains on publicly held and non-marketable equity securities.
Those gains are part of GAAP net income and therefore part of diluted EPS. Using $2.39 follows the model’s stated input. But the resulting 214.47% EPS growth is not purely an operating-growth measure. A C score of 10 therefore means reported EPS accelerated under the rule; it does not mean operating EPS accelerated by the same amount.
The same filing reported 85% revenue growth and 147% operating-income growth, so the quarter’s improvement was not created solely by investment gains. The correct interpretation is more precise: operating progress was substantial, while non-operating gains amplified reported EPS growth.
At that point, investors must still evaluate business quality. CAN SLIM does not determine customer concentration, competitive durability, capital intensity, supplier dependence, margin sustainability, or the repeatability of investment gains.
CAN SLIM versus Peter Lynch GARP
CAN SLIM and the Peter Lynch GARP model both use earnings growth, but they organize the decision differently.
| Question | CAN SLIM | GARP |
|---|---|---|
| Main concern | Is growth receiving market confirmation? | Is earnings growth reasonably priced? |
| Core fundamental evidence | Quarterly acceleration and annual growth | Historical EPS growth |
| Price role | Relative strength and market behavior | P/E relative to growth through PEG |
| Volume role | Explicit supply-and-demand input | Not part of the core calculation |
| Valuation output | None | Relative growth-pricing relationship, not intrinsic value |
| Market context | Explicit through M | Not a core model input |
A company can score well on GARP while lacking CAN SLIM leadership. Another can show strong CAN SLIM confirmation while trading at a PEG that GARP treats less favorably. GARP asks whether growth appears reasonably priced; CAN SLIM asks whether growth is strong and confirmed. Their disagreement reveals which type of evidence is present rather than identifying which model is wrong.
Neither framework replaces the intrinsic-value and defensive approaches described in stock valuation models. CAN SLIM does not say what a business is worth. GARP does not convert PEG into fair value.
Where CAN SLIM helps and what it misses
CAN SLIM is most interpretable for companies with:
- positive and comparable quarterly EPS
- a usable multi-year annual EPS record
- revenue that can be compared across consistent periods
- sufficient trading history and liquidity
- a stable, defined relative-strength universe
- share and institutional data that can be dated consistently
It becomes harder to interpret when:
- earnings cross from losses to profits
- acquisitions or accounting changes break comparability
- one-time gains or charges dominate EPS
- extreme cyclicality distorts annual growth
- thin trading makes volume ratios unstable
- the comparison universe changes
- institutional reports are delayed
- the broader market regime changes faster than financial statements
The framework also leaves major questions outside its score. It does not directly test liquidity, solvency, leverage, refinancing exposure, or drawdown risk. Those belong to financial strength and risk signals, not to a hidden reinterpretation of CAN SLIM.
The underlying stock analysis metrics need similar care. EPS, revenue growth, volume, shares outstanding, and total return are only comparable when their definitions and dates align.
How CAN SLIM fits inside the StockGeniuses Growth category
StockGeniuses includes three Growth Investing models:
- Peter Lynch GARP
- William O’Neil CAN SLIM
- O’Shaughnessy Cornerstone Growth
The Growth Overall Score is an equal-weighted average of valid model scores. At least two of the three models must be meaningful. Missing models are excluded rather than scored as failures.
That architecture treats agreement and disagreement as information. GARP may identify favorable growth relative to price. CAN SLIM may identify fundamental acceleration with market confirmation. Cornerstone Growth contributes another systematic perspective. An overall score is useful only after those distinct meanings remain visible; aggregation should summarize evidence, not erase why the models differ. The broader nine-model comparison extends that same discipline across Value, Momentum, Dividend, Risk, and Sentiment outputs.
This walkthrough does not calculate Nvidia’s other Growth models or infer an overall score. The public-data CAN SLIM run is incomplete, so it cannot contribute a valid company score to that aggregation.
A disciplined CAN SLIM review sequence
Use the framework in this order:
- Confirm eligibility before interpreting any component.
- Align fiscal quarters, annual periods, price dates, volume windows, and ownership dates.
- Recalculate current and prior quarterly EPS growth.
- Inspect the annual EPS path, not only the CAGR endpoints.
- Identify what changed in the business before accepting the N proxy at face value.
- Separate volume participation from changes in share supply.
- Define and preserve the relative-strength universe.
- Treat missing institutional data neutrally, but do not neutralize required missing inputs.
- Read M as market context rather than company quality.
- Audit what created earnings and revenue growth.
- Compare CAN SLIM with a genuinely different model instead of seeking automatic confirmation.
- Return the evidence to a full thesis before making a decision.
This sequence belongs within the broader process used to analyze a stock systematically. A useful model narrows a question, makes its evidence traceable, and states when it cannot answer.
The discipline behind the acronym
CAN SLIM endures because it refuses the false choice between fundamentals and market behavior. Current and annual earnings establish the growth case. Revenue acceleration, volume, share supply, relative strength, institutional participation, and market direction test whether other evidence agrees.
StockGeniuses makes that logic reproducible through fixed proxies, weights, and hard-failure rules. The score can summarize growth and confirmation when the required data is present. It cannot estimate intrinsic value, prove business durability, or predict returns.
The Nvidia walkthrough shows both sides of the method. Reported earnings, annual earnings, and revenue acceleration produce strong component results. The earnings still require a quality audit, and the final score remains unavailable without volume history and a defined relative-strength universe.
That outcome captures the framework’s most useful habit. Do not reward a compelling story by lowering the evidence standard. A strong partial result is still partial. Let independent signals agree, and let missing evidence remain missing.