Michael Drozd

Michael Drozd is chief investment officer, executive committee member and chair of the investment committee at LaSalle St., where he leads market outlook, asset allocation and multi-asset strategy. He also serves as senior portfolio manager for LaSalle St. Capital Management, overseeing the firm’s OCIO platform.

He previously held senior roles at Merrill Lynch, Citi Private Bank, J.P. Morgan and Calamos Investments, advising institutions and high-net-worth clients.

Michael earned a B.S. from Drake University, is a CIMA professional, and holds Series 7, 31 and 66 licenses.

CIO: How are you adjusting strategic and tactical asset allocation for a “higher-for-longer” rate environment, persistent inflation uncertainty, and increased geopolitical fragmentation?

Drozd: The post-zero-interest-rate-policy environment has created meaningful stress for portfolios constructed under prior regime assumptions. Rising sovereign debt levels have reintroduced term premium uncertainty, and equity-bond correlations have, at times, turned positive under inflationary conditions, reducing the diversification benefits that traditional balanced construction relied upon.

Geopolitical conflict, tariff disruption, large-cap technology concentration, artificial intelligence valuation dispersion and private market illiquidity add further complexity. In our view, this suggests the portfolio response may need to be structural, rather than tactical.

The core view is that diversification should function at the factor and regime level, not solely at the asset class level.

The framework rests on five components:

• Factor-diversified equities: using a value/growth balance across domestic and international markets to reduce single-regime concentration;
• A trend-following allocation: that, in certain environments, can exhibit differentiated behavior relative to traditional assets;
• Precious Metals: utilized as a potential diversifier in scenarios involving sovereign or currency related risks;
• Rules-based tactical overlay: a threshold-driven process to adjust risk during periods of macro deterioration without abandoning strategic positioning; and
• Reduced duration: an underweight to long-duration fixed income to help manage sensitivity to fiscal imbalances, sovereign balance sheet deterioration and interest rate risk.

CIO: What is one principle from your career that has proven especially relevant in today’s environment, and how have you applied it recently?

Drozd: The principle that has guided me most is staying intellectually curious and refusing to stop at surface-level answers, especially when clients are counting on you. In today’s environment, that matters more than ever. Markets are moving on narratives as much as fundamentals, and the difference between a good decision and a great one often lives several layers deeper than the consensus view.

Recently, I’ve applied this by going well beyond standard due diligence when building out our investment framework, stress-testing manager track records across full market cycles, pressure-testing our own tactical positioning for internal consistency, and being willing to say definitively that certain exposures simply don’t belong in client portfolios, even when they’re widely held.

This combination—staying genuinely curious and following through to the hard answer—is what I believe moves the needle for clients.

CIO: How is AI changing the way you generate insights, provide advice or otherwise work with clients—and where do you see its greatest limitation today?

Drozd: AI has had a measurable impact on research synthesis and investment infrastructure development. Large language models reduce the time required to process manager commentary, macro reports and due diligence materials from hours to minutes, allowing for more time to be directed toward analysis and judgment, rather than aggregation.

One concrete illustration: building the CMA framework, CAPM extensions and Carhart factor model that underpin our investment process required nearly six weeks of development and calibration. This is work that would typically involve a team of five to 10 analysts in an institutional setting. The same framework parameters, submitted to an LLM tool, produced a functional model in approximately 30 seconds. The output was roughly 80% accurate and structurally sound, but lacked the domain-specific decisions and calibration embedded in the original.

That gap defines both the opportunity and the limitation. AI accelerates the work; it does not replace the judgment required to evaluate it. Without the experience to assess what the model gets right and what it misses, the output carries real risk. The interpretation, validation and active decisionmaking remain human, and that is unlikely to change.