3 Lessons From Transamerica's SEC Settlement
The implications extend past the company itself.
Costly Mistakes Last week, Transamerica settled SEC charges against its funds by agreeing to pay $97.6 million in refunds, interest, and penalties. That is among the pricier SEC sanctions in recent years.
(As is customary with such actions, the company neither admitted nor denied the SEC's findings. The evidence, though, appears to be substantial. Also, although the case involves four Transamerica entities, for simplification's sake this column refers to all parties as "Transamerica.")
In 2010, Transamerica assigned a freshly minted MBA to construct the investment strategy for several forthcoming asset-allocation funds. Unhappily, the analyst was untrained in financial modeling. Perhaps as a result, he “did not follow any formal process to confirm the accuracy of his work." Moreover, “[management] failed to provide him meaningful guidance, training, or oversight ... to confirm that the models worked as intended.”
The models did not, in fact, operate according to plan. By the time all was said and done, the SEC identified seven legal violations of the Securities Act of 1933, the Investment Advisers Act of 1940, and the Investment Company Act of 1940.
One wonders: Are such problems limited to a single company? Or does this announcement indicate deeper, industrywide problems? For me, the answer lies somewhere in the middle. Transamerica’s example is extreme. I would be surprised to encounter a similar case. That said, the issues are broadly relevant. They affect much of the fund industry, albeit in milder fashion.
Model Concerns Transamerica boasted that the asset-allocation funds were run solely by models. The firm believed, correctly, that today's fund investors--and the financial advisors who serve them--prefer "emotionless" asset-allocation processes to approaches that use "portfolio manager discretion." Active management is not fashionable.
The belief in the machine’s superiority is erroneous. Quantitative models are merely automated forms of portfolio-manager discretion. They incorporate the research, views, and predictions of those who build them. Models insert judgments into the start of the investment act, rather than at the conclusion, as with traditional management. Ultimately, they are no better (or worse) than the abilities of those who create them.
In addition, the Transamerica situation demonstrates that, although models will execute more consistently than human intuition, it doesn’t mean that they will be more accurate. Just as there are no guarantees that a model is well-conceived, neither is there a guarantee it has been built correctly. The possibility for error is therefore twofold: one when specifying the investment strategy and the other when programming it.
This would seem to apply particularly to the many strategic-beta funds that have been launched in recent years, which use sometimes-complex routines based on often-complex research. The possibility for spills exists--and if those were to happen, who on the outside could know? Asset-allocation funds, such as Transamerica’s, are another candidate.
Who's in Charge? The junior analyst who created the funds' asset-allocation strategy was never identified as the funds' sole portfolio manager. Early on, Transamerica did not disclose his name at all. The company then described him as one of three associate managers reporting to a senior manager. The next year, it removed the senior manager's name (although retained the employee), thereby giving the three former associates equal billing. Five months later, it fired that junior analyst.
Thus, the funds’ shareholders never knew who managed their assets. It certainly was not the senior manager, who rejected the legal department’s request that he review the accuracy of the investment-process description because of unfamiliarity. (He also informed the marketing department that he was “not knowledgeable on the products.”) As for the three associates, it turned out that one had real control and two did not. That was impossible to know from the outside.
Such is the state of fund-manager disclosure. The buck reliably stops with corporate CEOs, who are expected to deeply know their businesses and give the requisite speeches, but it stops consistently with portfolio managers. Most funds these days cite multiple managers. Sometimes, they all bear significant responsibility. On other occasions, they do not. This leaves shareholders--and sometimes even Morningstar research analysts--ill-prepared to evaluate portfolio managers’ comings and goings.
Fund Boards The SEC identified two problems with the affected funds' prospectuses.
First, it was several years before the documents mentioned that the asset-allocation funds used quantitative models because Transamerica “drafted the prospectuses using a ‘library’ of approved disclosures”; the library “did not contain any disclosures relating to the use of models”; and nobody suggested that such an item should be required. A stark reminder that lawyers typically draft prospectuses, not investment professionals!
Second, as Transamerica learned through internal audits of the funds’ problems and took steps toward addressing them, including 1) identifying and fixing mistakes in the programming code; 2) manually adjustment the models’ recommendations (oh, the irony); and 3) finally, hiring a new subadvisor, it was slow to tell shareholders. On several occasions, it altered the funds’ prospectuses to reflect the new realities without explaining why those changes were made.
You might think that the funds’ boards of directors would have noticed these omissions. Think again. The good news for the boards was that the SEC determined that Transamerica had withheld critical information, so they would not be held liable for the transgressions. The bad news is that they were so easily duped. Per the SEC’s account, the boards not only were unaware of the models’ inaccuracies but also did not realize that the stated senior manager--who professed being “not knowledgeable” about the funds--was not in fact the decision-maker.
Which leads to the long-standing query of fund-industry boards of directors: Where’s the beef? To be sure, fund boards cause no harm. What is less certain is when they are a benefit. This was an opportunity for fund boards to demonstrate their value. Once again, they failed to answer the call.
(Disclosure: Morningstar Investment Management LLC, a Registered Investment Advisor and subsidiary of Morningstar Inc., is a subadvisor on four of Transamerica Asset Management’s allocation funds. Morningstar Investment Management has material business relationships with affiliates of Transamerica Asset Management.)
John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.