The global financial markets in 2025 were volatile, characterized by giant swings of performance driven by macro and thematic events. The year got off to a rocky start after DeepSeek, a Chinese AI firm, released a lower-cost yet highly performing AI model. This development called the level and benefits of U.S. tech sector spending on AI into question, leading to a selloff of many technology stocks.

Evaluating Investment Manager Performance

On the heels of DeepSeek were uncertainties around President Trump’s tariff policies, which injected additional volatility into the markets leading into and through “Liberation Day” (April 2), when baseline tariffs were imposed on all U.S. trading partners. As the tariffs became more clear to investors and worst-case scenarios around economic and market fundamentals did not materialize, stocks from April 8 through 3Q25 traded sharply off their lows and took on a risk-on posture. A more speculative sentiment entered the markets in late summer/early fall when issues such as quantum-computing stocks ran, adding additional fuel to the AI trade around which market returns had been concentrated.

Needless to say, 2025 was a lot and presented a particularly challenging backdrop for evaluating investment managers, as rapid shifts in market leadership complicated the assessment of process consistency and performance outcomes. In this environment, understanding a manager’s ability to adapt to changing conditions without compromising core investment discipline became increasingly important. With this in mind, institutional investors can examine portfolio construction and behavior through the following considerations:

Bifurcation in intra-quarter market environments should be considered when evaluating end-of-quarter results. Between pre- and post-Liberation Day, we saw the markets’ favor swing in opposite directions, and the same occurred for many managers based on how they were positioned. For instance, if a manager was defensively positioned (generally speaking) heading into Liberation Day, that manager probably picked up some relative return. Conversely, that same manager most likely experienced less upside after April 8.

We also saw this at the beginning of 3Q25. From April 18 to mid-October, only 11% of active small cap managers outperformed the benchmark against a backdrop of continuing speculation and resulting momentum in the markets. However, after the Oct. 15 trading session (where we saw a decoupling of many market-leading trades, particularly around themes like quantum computing), over 80% of small cap managers beat the benchmark through year-end.

What this shows us is two things:

1. End-of-quarter results are not always indicative of market behavior throughout a quarter, and this can have an impact on stock selection success, particularly for portfolios that are designed to be buy-and-hold in nature.
2. Process consistency is best gauged when evaluating portfolios against the backdrop of varying environments. For instance, if a typical “high quality” portfolio had outperformed after April 8 and on, it is worth exploring drivers, outside of stock selection, of that performance pattern. Similarly, if that same portfolio did not experience some improvement after the Oct. 15 trading session, the same exercise is warranted.

Thematic positioning has been critical, even for the typical bottom-up, fundamental manager. The strength of the AI theme has only continued to broaden since ChatGPT marked the beginning of AI exuberance in 2022. The broadening has been supported by the continued strength of investments supporting AI hardware and infrastructure and has shown up in trades within the Information Technology, Industrials, Energy, and Materials/Utilities sectors. The growing level of returns combined with the broadening of these AI themes has made it imperative for managers to have exposure here for any upside participation. Managers with little to no exposure and/or more exposure to “AI losers”—particularly software businesses—have encountered more performance challenges. Given this dynamic, it is important to consider a portfolio’s thematic positioning when analyzing attribution.

Ownership (and non-ownership) of key stocks remains a powerful consideration when assessing total attribution. Market and performance concentration continues to be driven by key stocks. We’ve seen this with the Magnificent Seven cohort for large cap and how the intra-cohort positioning and relative positioning to the aggregate group have determined the directionality of results (see my post on the subject). We’ve also seen this in recent years with the likes of Supermicro for small cap core and growth managers (here’s my post) and Palantir and Applovin’ for mid cap growth managers. In 2025, Bloom Energy—a power provider for AI data centers—was a ~2% position within the Russell 2000 Growth Index, and with a 400% rise in its stock over the year, a strong contributor to index performance. These examples augment the importance of evaluating benchmark relative ownership when determining the drivers of both positive and negative attribution.

In an environment defined by rapid rotations, concentrated leadership, and episodic reversals, disciplined evaluation of manager process, positioning, and adaptability remains more informative than short-term relative performance alone.