Background

Issuers of structured products that reference indices with limited histories often present in their offering documents hypothetical historical data. This "backtested" data is created by applying the index methodology to past historical prices or levels for the index’s constituents. Many view it as particularly useful in the context of proprietary indices, in order to help explain their potential performance in different market conditions.

Typically, when an issuer or index sponsor presents backtested data, the backtested data is accompanied by additional disclosure that explains how the data was derived, that the data is hypothetical and not actual, and that the backtested data cannot accurately predict future results. Generally, issuers also add risk factor disclosures that explain that backtested data is not necessarily accurate and is not predictive of future performance. If the index has actual historical performance results, the backtested data may be presented alongside the historical data, with an explanation of the two time periods and any differences in the data. Although there has been only limited guidance directly addressing the presentation of index performance, many issuers have analogized to the types of explanations and disclosures that have been used in the past in connection with presenting historical performance of private or publicly traded funds.

Given the prevalence of index-linked products, it is not surprising that regulators have become interested in the types of data that are shared with investors. For example, in the Securities and Exchange Commission’s April 2012 sweep letter sent to structured products issuers (question 12), the SEC asked issuers whether they use backtested data and, if so, whether the information was provided to investors and how it was presented.1

FINRA Guidance

On April 22, 2013, FINRA published guidance on the use of backtested, or pre-inception, index performance ("PIP data") in institutional communications, in the form of a letter to a distributor of certain exchange traded products distributed exclusively to "institutional investors," as that term is defined in FINRA Rule 2210(a)(4). The new guidance can be found at http://www.finra.org/Industry/Regulation/Guidance/InterpretiveLetters/P246651.

FINRA’s approval of the use of PIP data in institutional communications is limited to a narrow set of circumstances, and does not apply to retail communications, as defined in FINRA Rule 2210(a)(5). FINRA noted its historic position that the presentation of this information to retail investors does not comply with its disclosure standards.

In requesting interpretive guidance from FINRA, ALPS Distributors, Inc. ("ALPS") noted that their institutional customers of publicly traded securities structured as exchange traded notes, grantor trusts and registered investment companies (exchange traded products or "ETPs") had requested PIP data relating to the ETPs, and that the institutional communications containing the PIP data would not be forwarded on to retail investors.

ALPS proposed to FINRA that they use PIP data in institutional communications, as that term is defined in FINRA Rule 2210(a)(3), subject to certain limitations, including labeling the communication "for use with institutions only, not for retail use," using PIP data only for indices with freely available algorithmic methodologies, presenting the PIP data over at least a ten year period and covering multiple market conditions, with the PIP data being current to the most recent quarter, clearly labeled and dated, and separated from the actual historical data.

ALPS also represented that they would surround the PIP data with cautionary language, such as that the PIP data is hypothetical, is based on retroactive application of the methodology, with the benefit of hindsight, and that such data cannot account for all financial risk that may affect the actual performance of the ETP and that the actual performance of the ETP may vary significantly from the PIP data.

In its response, FINRA agreed to the use of PIP data in communications to institutional investors in the proposed manner. FINRA noted that, in applying suitability standards for recommendations to institutional customers, the firm "must be careful not to give excess weight to PIP data…." To the extent that PIP data "informs the firm’s understanding of the security and its performance characteristics, the firm must consider the correlation between PIP data and actual performance for similar ETPs managed by the sponsor, investment adviser or index provider."

FINRA stated that, in determining whether to use PIP data with institutional investors, firms should consider:

  • The assumptions, rules and criteria used to create the PIP data, in sufficient detail as to permit the firm to clearly understand how the PIP data could be replicated, using readily-available market data;
  • The reputation of the entity that created the PIP data, and if the issuer of the ETP paid for creation of the model, how any material conflicts of interest have been addressed or mitigated;
  • The conditions under which the PIP data may not be effective in predicting how the ETP may perform (e.g., very low or high interest rate environments);
  • The source of the data used to produce the PIP data;
  • The extent to which the PIP data has been tested under varying market conditions and scenarios, based on both an analysis of historical data and simulations or stress tests; and
  • Any reasons why the PIP data would have differed from actual performance of the ETP during the period shown (e.g., transaction costs, market liquidity).

Conclusion

Issuers may use PIP data in institutional communications, provided that the communications are within the parameters set forth in FINRA’s guidance. Keeping in mind suitability standards, PIP data should not be given undue weight when making a recommendation to an institutional customer and consideration should be given to the actual performance of similar products. In the context of this letter, FINRA reiterated its view that the use of PIP data in any retail communications violates FINRA Rule 2210(d).