What Makes One Stock a Better Buy Than Another: The Role of Price

Position size. Not stock selection. Not macroeconomic views. Not market timing. Position sizing is the most powerful driver of long-term investment returns, and the single most underutilized lever available to most investors. Here's how to think about it.

The Data on Position Sizing

A classic 2001 study by Brandes Institute researchers examined the investment returns of value investors with access to the same public information. Buffett, Greenblatt, Lynch, and other top investors had similar stock-picking accuracy rates and made similar numbers of mistakes. What differentiated them was not stock selection accuracy, but position sizing: how much capital they allocated to each idea as a percentage of their portfolio.

The investors with the best long-term records consistently sized their highest-conviction ideas larger than their lower-conviction ideas. A stock you believe will outperform by 5% should be a smaller position than a stock you believe will outperform by 50%. That's obvious in principle. It's astonishingly rare in practice.

Why Investors Get Position Sizing Wrong

The two most common position-sizing errors are: sizing all your positions equally (the "equal weight" trap), and sizing your positions based on your confidence without adjusting for your conviction edge (the "overconfidence" trap).

Trap #1: The Equal-Weight Trap

Many investors, especially those in target-date funds or broad index funds, own positions equal in size to their weighting in the index. The S&P 500 gives Microsoft more weight than Berkshire Hathaway, so an S&P 500 index fund does the same. Passive index investing is sensible for some investors, but it's the opposite of strategic position sizing: you're letting market capitalization, not your conviction, determine your allocation.

Active investors frequently fall into the same trap by default, holding equal positions in their favorite stocks. They tell themselves, "I like all five of these stocks," and allocate 20% to each. But if you genuinely believe one of the five has 2x the upside potential of another, and you're treating them equally, you're leaving enormous returns on the table.

Trap #2: Confidence Without Conviction

The second error is even more subtle. An investor might recognize that different positions warrant different sizes, but size them based on overconfidence rather than on an honest assessment of conviction and edge.

This is the trap Daniel Kahneman warned about in "Thinking, Fast and Slow": people tend to be overconfident in their ability to predict outcomes. A stock that seems obviously cheap might be cheap for a reason you don't understand. A stock that has beaten the market for the last five years might be due for mean reversion. An obvious "sure thing" is usually a warning sign, not a blessing.

A Practical Position-Sizing Framework

Here's a practical framework for sizing positions based on conviction:

1. Quantify your expected edge for each position. This doesn't need to be precise. If stock A has a 50% probability of beating the market by 20%, and stock B has a 30% probability of beating the market by 50%, stock A has an expected edge of 10% (0.5 * 0.2) and stock B has an expected edge of 15% (0.3 * 0.5). Stock B should be a larger position, assuming similar volatility.

2. Adjust for conviction by discounting your expected edge. If you're 90% confident in your analysis, discount your expected edge to 90%. If you're 60% confident, discount by 60%. This brutally honest adjustment prevents overconfidence-driven oversizing.

3. Size positions proportionally to your (discounted) conviction-adjusted expected edge, subject to a maximum position size. If your highest-conviction idea has a discounted expected edge of 8%, and your second-highest has 4%, the first should be roughly 2x the size of the second. But cap each position to something reasonable—say 5-10% of your portfolio per position, or 15-20% for your absolute highest-conviction ideas. These constraints prevent outsized losses if your conviction thesis proves wrong.

How the Best Investors Actually Size Positions

Joel Greenblatt's "The Little Book That Still Beats the Market" emphasizes that returns are driven not just by which stocks you own, but how much of your capital you allocate to each. Buffett has repeatedly said that his early partnerships' outperformance came from concentrated, conviction-weighted positions in deeply undervalued businesses.

Pabrai positions-sizes his core holdings at 20-30% of his portfolio when his conviction is highest. The key is that he's honest about which opportunities warrant that level of conviction, and doesn't get caught in overconfidence.

The Tax and Behavioral Costs of Position Sizing

One caveat: taxes and behavioral risk matter. In a taxable account, concentrated positions in appreciating assets create tax-loss harvesting challenges and capital gains liabilities. And behavioral risk—your tendency to panic-sell during market downturns—gets worse with larger positions.

An account sized into concentrated positions needs two things: (1) a long-term (5+ year) time horizon so you can absorb volatility without forced selling, and (2) psychological comfort with large single-position drawdowns. If a 25% drop in your largest position will cause you to panic-sell, your conviction conviction framework was dishonest. Size smaller.

The Ultimate Edge

Position sizing is the ultimate edge because it's the only lever that scales with your research. If you do better research than the average investor, you should size your positions to reflect that advantage. If you've built a competitive advantage in analyzing tech stocks, you should overweight them relative to your index. If you've identified three businesses trading at half their intrinsic value, and the average investor misses them, position sizing lets you capture the full benefit of that research advantage.

Position sizing is also the edge that doesn't require you to be right about everything—just proportionally right, sized to your honest conviction. The investors with the best long-term records are not those with the highest accuracy rates. They're those who size their highest-conviction, highest-edge ideas largest.