Operational Risk in Managed Accounts Can Cause Problems

A Cerulli report contends that while managed accounts are stable from an investment-risk and a regulatory standpoint, operational risk remains a concern.  

Operational risk in managed accounts threatens to undermine the profitability and reputation of firms that do not mitigate these risks, Cerulli explains in its report, published in the third quarter issue of “The Cerulli Edge-Managed Accounts Edition.”

Cerulli analysts examined specific areas where managers have operational risk concerns. The analysis hones in on unified managed accounts (UMAs) and model-only portfolio submission, as well as with general participation in separate account platforms.

“We asked asset managers about four facets of being on managed account platforms that contributed to operational risk. Three of the areas were identified by more than 75% of respondents. The one most often identified was trade order management, followed by delivering model portfolios (or paper portfolios), and fee processing,” writes Patrick Newcomb, senior analyst in Cerulli’s managed accounts practice.

When it comes to asset managers’ primary concerns with submitting model portfolios to overlay managers/UMA programs, receiving accurate compensation and information on sales and flows rank as the greatest concerns.

“For the most part, the responsibility for these two issues falls on the sponsor, leaving the asset manager with little control. Also, while model portfolio submission can be an avenue for building better relationships with sponsors, it can also build business risks to other products, should the model relationship hit rocky times,” says Sean Daly, analyst in Cerulli’s managed accounts practice.

Part of asset managers’ anxiety around models stems from the fact that there is little uniformity across how sponsors and overlay managers implement models, says Cerulli, and there is little in terms of standardization from an industry perspective. Asset managers remain in a difficult place when it comes to tracking how their models are actually being executed upon once they leave the hands of the manager.