FINANCEOngoing practice

The Holy Grail of Diversification

Find 15 good uncorrelated bets to reduce risk by 80% without reducing returns

Problem it solves

risk by 80% without reducing returns

Best for

Investors and entrepreneurs making multiple bets who want to dramatically improve their return-to-risk ratio

Not ideal for

Single-bet situations where diversification is not possible

Overview

Why this framework exists

The most important investment insight is understanding the power of uncorrelated bets. If you have 15 good bets that are 60% correlated, adding more only reduces risk by about 15%. But if you have 15 good UNCORRELATED bets with the same expected return, you reduce risk by over 80% without reducing average returns. This improves your return-to-risk ratio by a factor of 5. The key is finding bets that are intrinsically different—driven by fundamentally different factors—not just statistically uncorrelated in historical data.

Core principles

5 total
  1. 15 good uncorrelated bets reduce risk by over 80% while maintaining average returns
  2. Correlated bets give only marginal risk reduction no matter how many you add
  3. Intrinsic difference matters more than historical statistical correlation
  4. The return-to-risk ratio is more important than the absolute return
  5. Distinguish between the risk of ruin (being knocked out of the game) and the risk of painful learning

Steps

5 steps
  1. Distinguish ruin risk from learning risk
    Separate the risk of being knocked out of the game entirely (ruin) from the risk of painful mistakes that teach you (learning). Never take the first kind. Actively seek the second.
    Pro tipLike skiing: if you are not falling a lot, you are probably not learning—but do not kill yourself
  2. Find intrinsically different bets
    Rather than just checking historical correlation numbers, ask: are these bets driven by fundamentally different factors? A Silicon Valley startup and Colorado timber are intrinsically different regardless of what past data shows.
    Pro tipUnderstand what drives each bet's success or failure—if the drivers are different, the bets are likely uncorrelated
    WarningHistorical correlation can be misleading because correlations change across environments
  3. Aim for 10-15 uncorrelated positions
    The magic number is 15 uncorrelated bets for maximum risk reduction. Even 5 good uncorrelated bets provide substantial improvement. Beyond 15, the marginal benefit diminishes.
  4. Test for timeless and universal patterns
    Any relationship you rely on should be tested across different time periods (timeless) and different geographies or contexts (universal). If stocks and bonds behave similarly in Spain, Brazil, and the US, the pattern is more reliable.
  5. Triangulate with quality experts
    For any major decision, find three excellent experts who will disagree with each other because they are committed to the right answer. Listen to both their conclusions and their reasoning. This quality triangulation significantly raises your probability of being right.
    Pro tipThe fear of being wrong combined with curiosity about why someone disagrees is the best motivator for seeking triangulation

Checklist

Saved in your browser

Examples

2 cases
Bridgewater's portfolio construction

Bridgewater manages $160 billion by applying the Holy Grail principle: finding 15+ uncorrelated investment streams that each have good expected returns. They analyze the intrinsic drivers of each investment's price movement to ensure genuine uncorrelation.

OutcomeBy improving the return-to-risk ratio by a factor of 5, Bridgewater built one of the most successful investment firms in history over 42 years.
Dalio's buyer-seller analysis framework

Rather than predicting whether an asset is 'good' or 'bad,' Dalio maps out who the buyers and sellers are, what motivates each group, how big they are, and what they are likely to do. This includes institutional rebalancers, distressed sellers, lockup expiration sellers, and speculative buyers.

OutcomeThis granular understanding of market participants provides a much richer picture than simple value investing approaches.

Common mistakes

3 traps
Diversifying with correlated bets
Adding more bets that are 60% correlated only reduces risk by about 15% no matter how many you add. The marginal benefit is minimal because they all tend to move together.
Relying only on historical correlation data
Correlations change across environments and time periods. Understanding the intrinsic drivers of each bet is more reliable than just looking at past statistical relationships.
Confusing confidence with correctness
The greatest tragedy is holding wrong opinions that could easily be stress-tested. Humility about your own views and eagerness to find disagreement dramatically improves decision quality.

Origin story

How this framework came to be

Ray Dalio discovered this principle through decades of managing Bridgewater's portfolio. He realized that the power of diversification with uncorrelated bets is 'so much more valuable than trying to make any one bet much greater.' This insight—which he calls the Holy Grail of investing—applies far beyond financial markets to business strategy, career decisions, and any domain where you are making multiple bets on uncertain outcomes.

Source

Traced to primary
Source · PODCAST
Ray Dalio
Ray Dalio · 2017
Open source →

Related frameworks

Browse all Finance →