Summary will populate from computed model data.
Rather than relying on a single model to estimate expected returns, this framework adopts a multi-lens approach: it decomposes the expected return of an asset or portfolio into interpretable building blocks, each grounded in a different economic mechanism.
The core idea is that long-run expected returns can be understood as the sum of: (1) the current yield the asset provides, (2) the expected growth in that yield or in the underlying cash flows, and (3) any valuation change (multiple expansion or compression) that the market prices in through factor exposures.
The framework assembles expected returns from the bottom up, estimating each component independently and then combining them.
$$r_u \approx y_u + g_u$$
where $y_u$ is the unlevered yield and $g_u$ is the unlevered real growth rate.
$$E[r_e] = y_e + g_e + \Delta\text{PE}$$
Earnings yield plus earnings growth plus expected change in the price-to-earnings multiple.
$$E[r_i] = r_f + \sum_{k=1}^{K} \beta_{ik}\,\lambda_k$$
where $\beta_{ik}$ is asset $i$'s exposure to factor $k$ and $\lambda_k$ is the expected risk premium for factor $k$.
$$E[r] = \underbrace{y}_{\text{yield}} + \underbrace{g}_{\text{growth}} + \underbrace{\sum_k \beta_k \lambda_k}_{\text{factor premia}} + \underbrace{\Delta V}_{\text{valuation}}$$
$$g^* = ROE \times b$$
where $ROE$ is return on equity and $b$ is the retention ratio $(1 - \text{payout ratio})$.
The waterfall chart above shows the additive decomposition of expected returns. Each block represents a distinct economic source of return. This section will be populated with computed data from the MATLAB analysis pipeline.
The same framework applies to bonds (yield + roll + spread change), real estate (cap rate + rent growth + cap rate change), and commodities (carry + spot return + roll yield).