Reimagining Diversification: The Yale Model in Practice

Transcript

SPEAKER_1: Alright, so last lecture we landed on something that I keep coming back to — that tax efficiency isn't just a bonus layer, it's actually baked into the return profile itself. That reframing stuck with me. Let's delve into the theoretical underpinnings of diversification, focusing on correlation as emphasized by Fraser, which is often misunderstood. SPEAKER_2: Right, and this is where the book gets genuinely provocative. Most investors hear 'diversification' and think stocks plus bonds. But Fraser's argument — and it's grounded in decades of academic work going back to Harry Markowitz — is that true diversification is about correlation, not just variety. Owning twenty different things that all fall together in a crash isn't diversification. It's an illusion. SPEAKER_1: So what our listener might be wondering is — why doesn't the classic sixty-forty portfolio protect them during a crash? Bonds are supposed to be the safe side. SPEAKER_2: And historically they were. But in 2022, both stocks and bonds fell simultaneously — the S&P dropped roughly nineteen percent, and long-duration bonds dropped even more. The correlation between them spiked toward one precisely when investors needed them to diverge. That's the structural flaw. When fear drives markets, correlations converge. The hedge disappears exactly when you need it most. SPEAKER_1: So the sixty-forty portfolio fails at the moment of maximum stress. That's a pretty damning indictment. How does the Yale Model actually solve for that? SPEAKER_2: It solves it through what Markowitz formalized as mean-variance portfolio analysis — the idea that you can construct portfolios that maximize expected return for a given level of risk, but only if you account for covariances between assets, not just their individual returns. Yale's David Swensen applied this framework to emphasize the importance of correlation in diversification, selecting assets with genuinely low correlation to public markets. SPEAKER_1: Walk me through the correlation piece more concretely. What does a correlation coefficient actually tell an investor? SPEAKER_2: A correlation coefficient runs from negative one to positive one. Two assets at positive one move in perfect lockstep — owning both gives you zero diversification benefit. At zero, they're completely independent. At negative one, they move in opposite directions — that's a true hedge. Infrastructure, for example, tends to have a correlation to equities somewhere between zero and point two. Private credit is similarly low. That's what makes them genuinely diversifying — not just different labels, but different behavior under stress. SPEAKER_1: And this is what Markowitz called the Efficient Portfolio Frontier — the idea that there's an optimal set of portfolios for any given risk level? SPEAKER_2: Exactly. The Efficient Frontier maps every combination of assets that maximizes return for a given level of risk. The key point on that frontier is the Tangency Portfolio — the one that maximizes the Sharpe ratio, meaning the most reward per unit of risk taken. What Swensen recognized is that adding uncorrelated alternatives shifts that frontier upward. You get more return for the same risk, or the same return for less risk. That's not theory — that's what Yale's numbers showed over thirty-five years. SPEAKER_1: What does Yale's actual allocation look like? Because I think most people would be surprised by how far it deviates from anything conventional. SPEAKER_2: Yale's allocation strategy focuses on assets with low correlation to public markets, which is key to their diversification approach. Domestic equities — the core of most retail portfolios — are often under five percent. The majority of the portfolio is in assets that are either uncorrelated or negatively correlated to public markets. SPEAKER_1: Under five percent in domestic stocks. That's almost the inverse of what most people hold. So what's the historical performance gap between that approach and a standard sixty-forty? SPEAKER_2: Over the long run, the Yale Endowment has averaged returns in the range of twelve to thirteen percent annually, compared to roughly eight to nine percent for a traditional sixty-forty portfolio. That gap compounds dramatically over decades. And critically, Yale achieved that with lower volatility — because the uncorrelated assets smoothed the ride. The Mutual Fund Theorem from CAPM theory actually supports this: all efficient portfolios are combinations of a risk-free asset and the tangency portfolio. Yale just built a better tangency portfolio. SPEAKER_1: So how does Fraser translate this for someone who isn't managing a thirty-billion-dollar endowment? What does the scaling actually look like? SPEAKER_2: Fraser emphasizes starting with accessible alternatives that offer low correlation to public markets, scaling as the portfolio grows. As the portfolio scales toward a million, you add private credit and direct real estate. At the ten-million level, you're building something that genuinely mirrors the Yale model — private equity funds, hedge fund allocations, infrastructure. The architecture scales; the principles don't change. SPEAKER_1: That's a useful ladder. But I want to push on one thing — how does someone actually evaluate whether an alternative asset is genuinely uncorrelated, or whether it just looks that way because it's not marked to market daily? SPEAKER_2: That's the right tension to hold. Private assets use Level 3 fair-value accounting — valuations based on manager models, not real-time prices. So some of the apparent smoothness is a measurement artifact. The honest answer is: you look at behavior through actual stress periods, not just reported volatility. Assets that held value or generated cash flow through 2008 and 2020 — infrastructure, private credit with senior secured positions — those demonstrated real uncorrelation, not just accounting smoothness. SPEAKER_1: So the stress test is the real proof. Not the correlation number on a pitch deck. SPEAKER_2: Exactly. And this connects back to what Fraser emphasizes throughout the book — information asymmetry. In private markets, the manager who has done the operational due diligence, who understands the asset's cash flow mechanics through a downturn, has a genuine edge. That's not available in public markets where millions of analysts are pricing the same stock in real time. SPEAKER_1: So for Sergey and everyone following along — what's the one thing they should carry out of this lecture? SPEAKER_2: That institutional-level returns require a portfolio built around uncorrelated assets — hedge funds, infrastructure, private credit — not just a wider spread of stocks and bonds. Markowitz proved mathematically that correlation is the variable that matters. Swensen proved it empirically over thirty-five years. Fraser's contribution is showing that the same framework is accessible at any portfolio size. The architecture scales. The discipline of seeking genuine non-correlation — that's what separates a weatherproof portfolio from one that just looks diversified until the moment it isn't.