Conviction Weighting

Equal weighting assumes all ideas are equal. Most investors know this is false. You understand some businesses better than others. You have deeper conviction in some theses than others. Your highest-conviction ideas should logically receive larger allocations.

This is conviction weighting: sizing positions proportionally to your confidence that they will deliver superior returns.

Conviction weighting means allocating capital based on the strength of your belief in each investment thesis, with higher conviction generating larger position sizes within defined limits.

Key Takeaways

• Conviction weighting aligns your capital with your knowledge. If you understand a business exceptionally well, you should own more of it.
• It requires intellectual honesty about confidence levels. Most investors are overconfident. Translating confidence into sizing forces realism.
• Without guardrails, conviction weighting becomes concentration. Many skilled investors have been wrecked by conviction in one idea that ultimately failed.
• Conviction changes as information changes. Sizing rules must allow for downsizing positions as conviction declines.
• The best conviction weightings blend confidence with mathematical limits. You might allocate 50% more to high-conviction vs. low-conviction positions, but cap all positions at 12%.

Defining Conviction Levels

To practice conviction weighting, you must first define what you mean by conviction.

A useful framework has three levels:

High Conviction: You understand the business exceptionally well. You have deep knowledge of the industry, management, competitive position, and long-term drivers. You have tracked the company for years or invested in similar businesses successfully. You can articulate your thesis clearly and explain why the market is mispricing it. Your confidence that the stock will outperform is 75%+ over a five-year horizon.

Example: An investor who has worked in software might have high conviction in a SaaS company they have studied for two years. They understand the product-market fit, unit economics, and TAM.

Medium Conviction: You understand the business well but not exceptionally. You have read the annual reports and understand the competitive positioning. You cannot articulate hidden edges the market has missed, but you see sustainable competitive advantages. Your confidence that the stock will outperform is 55–75%.

Example: A long-term investor in blue-chip healthcare might have medium conviction in a well-established pharmaceutical company. It is a solid compounder, but it is also fairly valued.

Low Conviction: You like the business but do not have special insight. It seems undervalued by simple metrics or has interesting catalysts. But you know you are not expert in the industry, and the bull case relies on assumptions you cannot fully verify. Your confidence is 50–55%.

Example: An investor noticing a retail company trading below book value might allocate capital, but without deep insight, conviction is low.

In practice, most positions fall into medium conviction. High conviction should be rare. If you claim high conviction about 10 positions, you lack honesty about your actual expertise.

Translating Conviction to Sizing

Once you have defined conviction levels, the next step is translating them into position sizes.

A simple framework:

• Low conviction: 3% per position
• Medium conviction: 6% per position
• High conviction: 9–12% per position

If you own 10 stocks—three high, four medium, three low conviction—your portfolio allocation is:

• High conviction (3 × 9%): 27%
• Medium conviction (4 × 6%): 24%
• Low conviction (3 × 3%): 9%
• Cash or flexibility: 40%

This approach allows your highest-conviction ideas to receive 3x capital allocation compared to lowest-conviction ideas, while still maintaining reasonable diversification.

An alternative is to use percentages. Instead of fixed sizes per conviction tier, allocate proportionally:

• High conviction positions split 60% of allocated capital
• Medium conviction positions split 30% of allocated capital
• Low conviction positions split 10% of allocated capital

This maintains flexibility as you add or remove positions.

The Overconfidence Problem

Conviction weighting only works if your conviction estimates are realistic. Unfortunately, investors are chronically overconfident.

A classic study asked investors what confidence level they were 90% certain about earnings. Actual outcomes fell within their predicted range only 66% of the time. If they were truly 90% confident, accuracy should have been 90%. They were routinely wrong.

This overconfidence cascades into sizing. An investor assigns 85% confidence to a position, sizes it at 10%, and then is shocked when the thesis fails 15% of the time. That 15% becomes a blowup.

To combat overconfidence, apply a "conviction discount." If you assign yourself 85% confidence, assume you are actually 75% confident. Downsize accordingly.

Another check: ask yourself, "What would need to be true for me to be completely wrong?" If you cannot articulate plausible failure scenarios, your conviction is false confidence.

Updating Conviction

Conviction is not static. As new information emerges, conviction should shift.

The disciplined conviction weighting investor reviews positions quarterly, asking:

• Have my assumptions about this business been validated?
• Has competitive positioning changed?
• Has management executed as expected?
• Have I learned information that raises or lowers my confidence?

Based on these questions, conviction either remains stable, increases, or decreases. Position size adjusts accordingly.

Many investors treat sizing as a one-time decision. They allocate 10% to a stock and mentally commit to that size. Years pass. The business deteriorates. Conviction should collapse to 40%, but the position remains at 10%.

Conviction weighting requires quarterly recalibration. If conviction shifts from high to low, the position shrinks to the low-conviction allocation. Capital freed is redeployed.

This dynamic rebalancing is more active than equal weighting but remains less active than trading. Positions shift gradually as conviction changes.

The High-Conviction Trap

Investors who successfully identify a compounder and own it with high conviction often face a dilemma: the position grows.

You own a company at 10% that compounds at 20% annually while the rest of the portfolio compounds at 8% annually. In five years, that 10% position may be 18% of your portfolio. In ten years, it may be 30%.

This creates two choices: let the winner run or trim it back to your conviction-weighted target.

Trimming winners back to target creates tax drag and can cause regret if the stock continues outperforming. Letting them run concentrates risk, violating your sizing discipline.

The best approach depends on conviction and circumstances:

• If conviction has increased based on new evidence, allow the position to run larger than the original target
• If conviction is stable, trim back to target periodically
• If the position has exceeded 20% and conviction is not extraordinary, trim regardless

Some investors use a "collar" approach: positions can drift to 1.5x their target before forced trimming. A 10% position can grow to 15% without action. At 15%, it is trimmed back to 10%. This allows winners to run somewhat while preventing excessive concentration.

Conviction vs. Knowledge

A crucial distinction: conviction should be based on relative knowledge, not absolute confidence.

You do not need to be 90% certain about future returns to have high conviction. You need to be significantly more certain than the market is. If the market is 50% confident and you are 70% confident, you have meaningful high conviction—not because 70% is a high absolute level, but because your assessment differs from consensus.

This prevents false conservatism. A skilled technology investor may have 65% confidence in a software compounder, which is high conviction for them even if 35% is a meaningful error band.

Conversely, low conviction does not mean you think the stock is bad. It means you are closer to neutral than bullish, or you lack expertise to assess the case.

Common Mistakes

Mistake 1: Excessive conviction in a narrow thesis. Investors sometimes become intellectually captured by a single argument—"this company will be worth 5x in five years"—and build conviction entirely around that thesis. If the thesis fails, conviction collapses and the position is sold at a loss. Conviction should be based on multiple reinforcing factors, not one.

Mistake 2: Confusing confidence in the stock market with conviction in a stock. Being bullish on markets generally does not create high conviction in a specific stock. High conviction requires specific knowledge.

Mistake 3: Not adjusting conviction as company improves. If conviction starts low but the company executes flawlessly, conviction should rise, and position size should increase. Many investors arbitrarily cap sizes and never adjust.

Mistake 4: Anchoring conviction to entry price. Conviction should be based on current and forward assumptions, not on whether you got a good entry price. An original high-conviction buy can become low-conviction if circumstances change.

FAQ

Q: How often should I reassess conviction? A: Quarterly or semi-annually works well. Annual reassessment is the minimum. More frequent than quarterly is likely excessive.

Q: What if I cannot honestly rank my positions by conviction? A: You may not be ready for conviction weighting. Equal weighting may suit you better. Or, force ranking them: which single position would you own if you could only own one? That is your highest conviction. Rank the rest similarly.

Q: Should I assign conviction percentages before or after analyzing valuation? A: Conviction should reflect the quality of the business and the quality of your knowledge. Valuation affects whether to buy (is it cheap enough?) and sizing (expensive positions can be sized smaller). Conviction about the business is separate.

Q: Can I use conviction weighting with index funds? A: Not in the traditional sense. But you can use conviction to decide how to allocate between index funds. You might have high conviction in a US stock index and low conviction in cryptocurrency, and size accordingly.

Q: What is the maximum size I should give to a high-conviction position? A: 12–15% is reasonable for most portfolios. Some investors use lower limits (10%) for more conservative approaches. Anything above 20% is concentration, not conviction weighting.

Q: How do I know if my conviction is realistic? A: Track record. If your high-conviction picks outperform and low-conviction picks underperform, your conviction assessment is realistic. If they all perform similarly, you are not discriminating.

Related Concepts

• Investment thesis — The articulated argument for why a stock will outperform
• Edge — A structural advantage that allows you to assess stocks more accurately than the market
• Confidence interval — The range within which you believe outcomes will fall
• Position monitoring — Regular review of positions to update conviction

Summary

Conviction weighting is the middle path between equal weighting's mechanical simplicity and the false precision of complex allocation models. It allows for the reality that you understand some businesses better than others, while building guardrails against overconfidence.

The discipline required—articulating conviction clearly, assigning it honestly, and updating it as information changes—is the true value. Most investors benefit from this exercise alone, regardless of the specific sizing formulas they use.

Conviction weighting works best for investors with deep expertise in specific industries or business models, and who commit to quarterly review and rebalancing based on changed circumstances. For others, it may devolve into hidden overconfidence and poor results.

Next: Introduction to the Kelly Criterion

Beyond simplicity and conviction, there is mathematics. The Kelly criterion offers a formula for sizing based on your historical edge and win rate. We explore how it works and why most investors should use it cautiously.