Additional Content Provided by John Lohse CFA, Senior Analyst.

Uncovering the Full Picture: Analyzing Top-Down and Bottom-Up Exposures in Multi-Asset Model Portfolios 

When constructing and evaluating multi-asset model portfolios, we believe it’s beneficial for investors to consider both the top-down asset allocation and the bottom-up exposures of the underlying holdings. By understanding these two approaches, investors can maintain simplicity of portfolio construction and easy comprehension of a model’s investment strategy and risk profile, while also gaining a more comprehensive perspective of a model’s true exposures. 

The Top-Down Approach: Simplicity with Potential Blind Spots 

The top-down approach, analyzing a model’s allocations across different asset classes, sectors, geographic regions, etc., can provide valuable oversight into the overall investment strategy. This approach takes the portfolio’s component fund or ETF’s assigned asset classes at face value, assuming that all of that fund or ETF’s component investments are of the assigned asset class. This high-level perspective offers the benefit of simplicity, allowing investors to quickly understand the model’s broad diversification, risk-return profile, and sensitivity to various asset classes. 

A top-down perspective helps provide a simple approach to asset allocation. It can be a great starting point for portfolio construction, from which a further bottom-up analysis can be completed. In certain models, especially ETF models due to their more pure-play exposure, there may not be significant adjustments needed from that initial starting point.  

However, the top-down perspective can sometimes lead to unintended exposures, as the top-level asset allocation categorization may not fully capture the underlying holdings within each component mutual fund or ETF. This is particularly true for actively managed, balanced, multi-asset, or global investment vehicles, where the stated allocation often doesn’t accurately reflect the actual exposures of underlying holdings. 

The Bottom-Up Approach: Granular Insights with Added Complexity 

In contrast, the bottom-up, or “look-through,” approach focuses on the characteristics of the specific holdings within the component mutual funds and ETFs used to construct the multi-asset model portfolio. This in-depth analysis can help identify exposure drift or style creep as it’s sometimes known, where the underlying holdings do not align with the stated top-level asset allocation. 

By being aware of these nuances in exposure, investors can ensure that portfolio allocations and factor exposures align with an investor’s investment thesis, increasing the likelihood that a model’s performance matches expectations in any given market environment. The downside of this bottom-up approach is that it requires more complex systems and advanced data analytics to implement effectively, as investors need to delve into the specifics of each underlying holding and then reaggregate these to gain a comprehensive picture of the portfolio’s exposures. At first glance, it can be less intuitive to understand, especially if there is exposure drift evident in the component holdings, and interpreting the output, especially of active funds, can be more nuanced. 

Understanding the Nuance of Bottom-Up Analysis for Active Managers 

When utilizing active managers in models, it is common for these funds to invest a portion of assets outside of their core mandate. For example, a large-cap fund may own some mid-cap stocks in its portfolio and conversely, a mid-cap fund may own some large-cap stocks. Through a bottom-up analysis, we are aware of these exposures, and we may want to take corrective action. In some cases, the exposures in one fund may be offsetting or complementary to the exposures in another fund, such that no action is needed. In other cases, we may reduce or eliminate a fund if we believe it has led the model to overconcentration in a certain type of investment. There may also be cases where our Model Portfolio Committee (MPC) and Investment Manager Research (IMR) teams become concerned that a fund has deviated too far from its mandate or purpose, exhibiting exposure drift (that could be a creep into other styles, cap-sizes, geographic regions, etc.). There are many criteria to help us make this determination based on our intimate knowledge of the funds. We seek to understand whether the fund’s investment team has expertise in the areas in which it is investing and if those allocations are excessive. During direct communication with the active manager, we want to hear a solid rationale as to why these investments are appropriate for the portfolio. Additionally, when a fund veers towards securities that are outside of its mandate, it may have a change in its risk profile or performance patterns, which may make it less likely to meet investor expectations. Using a combination of the bottom-up data and our knowledge of the funds held, we determine if we maintain a fund as its holdings still fit the model well, reduce the fund’s weighting to prevent overconcentration, or eliminate the fund from the model if it is no longer an appropriate holding and is demonstrating harmful style drift. 

How Bottom-Up Exposes Size and Style Creep: An Example 

Below is an example of the difference between a top-down and bottom-up analysis on the domestic portion of a hypothetical equity mutual fund portfolio exhibiting size and style creep. There is a significant size creep with the large cap managers going down-cap into mid-caps (and mid-cap managers going down-cap to small caps). On a bottom-up basis, this portfolio has almost double the exposure to mid-caps than shown on the surface by a top-down approach. The bottom-up data also shows that the portfolio exhibits style creep with the portfolio’s spread of growth stocks to value stocks higher than what’s indicated by the top-down perspective. Note, too, that this example is focused on domestic size and style, but a more comprehensive bottom-up analysis would pick up other data points such as geographic tilts, sector weights, and asset class dispersion.

Mutual Fund Portfolio Exposure from a Top-Down Versus Bottom-Up Analysis

Source: LPL Research, Morningstar 04/16/24

Balancing Top-Down and Bottom-Up Perspectives 

In general, models composed of passive investment vehicles and with more component parts tend to have top-down and bottom-up exposures that are closer to each other than a model populated with active managers or with few component pieces. For example, a large cap value equity index ETF is likely to have more pure-play exposure to underlying large cap value stocks than a large cap value active mutual fund, which may have significant style and size creep. A portfolio composed of large cap, mid-cap, and small cap component equity index ETFs will likely be a lot closer on a top-down and bottom-up level than a portfolio composed of a single all-market index ETF (where the top-down would all be categorized as large cap).  

Below is an example of a portfolio made up of a single total-market ETF. On a top-down analysis, the total market ETF would appear to be 100% large blend, but from a bottom-up perspective the portfolio, unsurprisingly, shows market-like exposures across market cap and style.  

All-Market ETF Portfolio Exposure from a Top-Down Versus Bottom-Up Analysis

Source: LPL Research, Morningstar 04/16/24

In the above example, the bottom-up versus top-down approach yielded markedly different results as it was a concentrated portfolio that used a blend/all-market ETF. In the following example of a portfolio made up of market-weighted allocations to large cap, mid-cap, and small cap growth, blend and value ETFs, the bottom-up and top-down analyses yield much more balanced results, although still exhibit some size and style drift. The key factors to this similarity are the granular construction of the portfolio and the index ETFs used to implement exposures.  

Granular ETF Portfolio Exposure from a Top-Down Versus Bottom-Up Analysis

Source: LPL Research, Morningstar 04/16/24 

Summary

Ultimately, we believe the most comprehensive analysis of multi-asset model portfolios, and the most successful portfolio construction process, requires a blended approach that considers both the top-down and bottom-up perspectives. By combining the simplicity of the top-down asset allocations with the granular understanding of the bottom-up exposures (and especially how these may change over time for active managers), we believe we can create resilient portfolios that are equipped to react as expected to changing market conditions. Blending both analytical lenses allows us to gain a deeper understanding of a model’s true exposures and expected behavior, while maintaining the ability for model users to easily comprehend the portfolio’s overall strategy, enabling informed decisions to be made that align with financial objectives.