Back-tested performance – often misleading, but not off-limits with precautions

January 29, 2015

(Reporting by Julie DiMauro and Jason Wallace)

In separate actions against a Massachusetts-based exchange-traded funds investment manager, the SEC warned advisers to be careful if they advertise their performance, and to pay particular attention to the distinctions between true actual performance, model performance and back-tested performance.

Both actions were filed on December 22 against F-Squared Investments and its former CEO, with the firm settling for $35 million in disgorgement and penalties and the case against the former CEO just getting underway in U.S. District Court. 

The settlement against the firm included a rare admission of guilt,. Both both cases centered around the veracity of F-Squared’s advertising performance claims of its index product, AlphaSector, the investments of which could be rebalanced periodically based on buy-sell signals from an algorithm.

The SEC said the firm and its former CEO, Howard Present, “defrauded investors through false performance advertising about its flagship product.” The advertising performance claims in question occurred during a five-year period, from September 2008 to September 2013.

“F-Squared falsely advertised a successful seven-year track record for the investment strategy based on the actual performance of real investments for real clients,” the SEC said. “In reality, the algorithm was not even in existence during the seven years of purported performance success.”

Real performance, or back-tested?

The data in those seven years, April 2001 to September 2008 allegedly was advertised as real performance data, but “was actually derived through back-testing,” the SEC said.

“Back-testing” is the application of a quantitative model to historical market data to generate hypothetical performance during a prior period. Since AlphaSector was not actually created until late 2008, it could not have had real performance data from that period, the SEC said. Plus, F-Squared “specifically labeled the investment strategy as ‘not back-tested’” in its advertising materials.

The administrative order instituting the settlement went on to say that the hypothetical data also contained a substantial performance calculation error that inflated the results by approximately 350 percent.

“We allege that not only did F-Squared and Present attract clients to this investment strategy by touting a track record they presented as real when it was merely hypothetical, but the hypothetical calculations also were substantially inflated,” said Julie Riewe, co-chief of the SEC Enforcement Division’s Asset Management Unit.

Hypothetical performance data

The use of hypothetical performance data is not an uncommon practice with investment advisers. Traditionally an advisory firm will use a model portfolio to show hypothetical or simulated performance. The model is often presented as an ideal combination of securities for a client’s portfolio.

Unlike the model portfolios, back-tested performance presents hypothetical results based upon the retroactive application of an adviser’s investment strategy over a select market period. There is no real market risk. The goal is to show performance returns that would have been achieved if the investment approach had been in existence during the period shown.

In some cases, hypothetical data is used when a new fund is being offered and little or no historical performance data is available.

It can also be used to illustrate an advantage provided by an adviser’s investment philosophy or methods, such as the benefits of diversifying among a broad range of asset classes, said Tina Petruzziello, founder and compliance principal of Boston Compliance Associates.

“In all cases, though, the hypothetical performance data would be misleading unless it is accompanied by full and clear disclosure explaining what it is, how it was derived, why it is being provided, the fact that it is not the performance of any actual account and, of course, that it is not a guarantee of future results,” Petruzziello said.

The disclosure must be appropriate to the sophistication level of the audience, which can be challenging to a retail audience. But with sufficient disclosure, reviewed by the legal and compliance departments, hypothetical performance need not spell ‘regulatory trouble,” she said.

In an attempt to bring validity to performance figures, a firm may claim compliance with the Global Investment Performance Standards (GIPS). GIPS are a set of voluntary guidelines for calculating and presenting investment performance. The standards were created and are currently being administered by the Certified Financial Analyst Institute.

Currently, a firm can claim GIPS compliance without having an outside firm verify or audit their results.

Hypothetical and back-tested composite returns do not satisfy the requirements of the GIPS standards. To be GIPS compliant, performance data must only contain actual portfolios managed by the firm. Hypothetical or back-tested results can only be included when clearly labeled as supplemental information.

“Advisors to institutional investors continue to see an advantage in adhering to GIPS standards,” said Petruzziello.

She pointed out that, starting on January 1, firms claiming compliance with GIPS standards are to notify the CFA Institute so as to be included on the GIPS Standards website.

The rationale behind this is to motivate firms not currently claiming compliance to do so, resulting in broader adoption.

“The CCO should be involved in this process to make sure there is a common understanding among investment, operations, marketing and compliance personnel about how performance should be calculated and presented, ensuring the required disclosures are included,” Petruzziello said.

Judy Gross, founder and CEO of JG Advisory Services in New York agrees.

Gross advises her hedge-fund clients to have the compliance department sign off on all new advertising material. “The compliance team should keep a detailed record of when these performance materials were last used, what materials they reviewed, any changes or updates they made and when they were made so there is no confusion over who last verified this data,” Gross said. And these recordkeeping steps must be a part of a firm’s policies and procedures.

Gross noted a March 2013 enforcement against two investment advisers at Oppenheimer & Co. with misleading investors about the valuation policies and performance of a private equity fund they manage.

The SEC said in its enforcement order that the firm failed to implement the policies and procedures necessary to prevent investors from receiving misleading information about the firm’s valuation policies and performance numbers.

Compliance practices

The SEC rules do not address model performance, and very little official guidance exists. Firms are forced to rely on regulation from Rule 206(4)-1 of the Advisers Act, which prohibits an adviser from engaging in advertising that “contains any untrue statement of a material fact, or which is otherwise false or misleading” and multiple SEC no-action letters.

In specific, Clover Capital Management, Inc, is arguably the most important of the SEC no-action letters in which the staff presented guidelines for advertising with actual and model performance. The letter specifically lays out the standards the SEC staff uses to determine whether the advertising is fair and not misleading.

In the case of back-tested results, much of the guidance and disclosure provisions come directly from SEC enforcement cases over the years.

In general, the use of back-tested results would be reserved for sophisticated investors who can understand the limitations and disclosures offered. In an attempt to simplify, a collection of recommended practices for presentation and disclosure has been listed. They have been sourced from the SEC’s no-action letters and enforcement cases.

Disclosure content:

  • Incorporate a description of the model, its limitations, assumptions and quantitative parameters necessary to interpret the results. If compared to an index, include all material facts for comparison;
  • Disclose whether the results were obtained with the benefit of hindsight and if advisory clients had outcomes materially different than model results;
  • Include a description of how advisory fees, trading costs, reinvestment of dividends, interest, capital gains, and withholding taxes are treated;
  • Disclose whether strategies reflected in the model portfolio do not relate, or relate partially, to the services currently offered by the adviser and whether the model has changed materially over the time period presented;
  • Include all material economic and market factors that might have impacted the decision-making when using the model to manage actual client accounts;
  • Have a multiple- stage approval process involving compliance, marketing and legal professionals; and
  • Consider a third-party for verification of results.

Disclosure presentation:

  • Ensure disclosure statements are conspicuous. In specific, the font should be similar to other text and above the fold on any web page;
  • Keep back-tested results separate and distinct from actual performance data. Label back-tested hypothetical results as such and include dates that relate to the results; and
  • Maintain sufficient records to support both the calculations and the results. Copies of presentations should be kept in accordance to the adviser recordkeeping rule.

(Julie DiMauro is a regulatory intelligence and e-learning expert in the GRC division of Thomson Reuters Accelus. Follow Julie on Twitter @Julie_DiMauro. Jason Wallace is a senior editor for Thomson Reuters. Jason began his career at TD Waterhouse Securities Inc., now TD Ameritrade Inc., where he held key positions in the Trading, Risk Management and Compliance departments for both retail and institutional sides of the firm. Jason joins Thomson Reuters after serving as an associate director for National Regulatory Services, in San Diego, California. Follow Jason on Twitter @Wallace_iabrief.)

No comments so far

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/