Form PF: How do you address Question 42 (Risk Metrics)?
We are currently assisting one of our clients with developing and implementing an enterprise wide process for the Form PF filing (i.e. overall project management, data mapping of developing and implementing a process, 3rd party solution selection). We initially reviewed Question 42 (see below) in a working session with the client’s (say XYZ company) Form PF team, and although it is possible for the risk management team to answer the question as outlined in Form PF, calculating the P/L impact of each market factor scenario independently would not be relevant to XYZ’s investment strategies. Some factors such as “Option Implied Volatilities” are not material to the overall portfolio risk and others seem to be redundant. However other XYZ’s proprietary risk measures, which account for multiple factor changes that would occur in a true market environment, give a more accurate representation of XYZ’s risk management. Based on the discussions, we identified two reply alternatives (see below) and each could affect both regulators’ and clients’ perception of the firm. We would be interested to learn what other group’s members think.
Form PF Question 42:
For each of the market factors identified below, determine the effect of the specified changes on the reporting fund’s portfolio and provide the results. The market factors are the following:
• Equity prices
• Risk free interest rates (changes represent a parallel shift in the yield curve)
• Credit spreads
• Currency rates
• Commodity prices
• Option implied volatilities
• Default rates (ABS)
• Default rates (corporate bonds and CDS)
Note: Assume that changes in a market factor occur instantaneously and that all other factors are held constant. If the specified change in any market factor would make that factor less than zero, use zero instead.
Reply Alternatives – Pros and Cons:
1. Answer Question as Designed & Provide Assumption Write-up – This would require XYZ to generate results for the various shock scenario factors monthly, solely for the purposes of Form PF. Risk Management would be required to validate each calculated factor and sign-off. Then also provide a write-up of XYZ’s true risk process, and record that into the Form PF assumptions section (Question 4).
Impact: Answer would satisfy regulator; but if asked in audit how is each calculation used in risk process, the response would be that it was solely done to respond to the Form PF and other risk measures are used for day-to-day risk measurement. If clients obtain copies of XYZ Form PF filing, they would see risk measures that are not used to manage XYZ every day process. Finally, if these were going to be generated monthly, Risk Management would have to perform some validation on the calculations in Form PF even if not used by firm.
2. Answer “Not Relevant” for each Question & Provide Assumption Write-up – XYZ has the option to answer “Not Relevant” for all of Question 42, and then provide a write-up of XYZ’s true risk process, and put into the Form PF assumptions section (Question 4).
Impact: Not providing a response for each market factor would potentially call attention to XYZ and bring greater scrutiny when audited by the SEC. In addition, if clients get copies of XYZ Form PF filing, they would see figures that are inconsistent with XYZ’s stated Risk Management practices used just to populate the Form PF, but could see the write-up in the assumptions. Finally, no generation and review process would need to be created by Risk Management on a monthly basis.
It is pretty clear from the SEC’s release that your client does not need to build quantitative models that are not currently used by its risk management systems (I think the Release even more strongly suggests that advisors will base reports on their existing internal methodologies). Therefore, if (i) all factors listed are truly irrelevant individually (Note: “relevant” is not defined – and it is not clear that “relevancy” is used here interchangeably with “materiality”) box 1 would be checked; (ii) if any factor is “relevant”, but is not considered formally and individually by the advisor’s risk models then box 2 would be checked; (iii) if any factor is relevant and is formally tested individually or is part of a multi-factor model which can be (and is) considered in isolation from the multi-factor model then the effects on the portfolio should be included.
In any event, based on your description of an alternative multi-factor model process, you would include in your response to Question 4 the advisor’s true risk management framework; that response, in the unlikely event that Form PF data is disclosed to clients (the Rule’s confidentiality provisions should preclude any disclosure by the SEC / FSOC), should be consistent with previous disclosures to clients.
Analyzing this from the SEC audit risk perspective, seems like you need to include your true framework in all events and the SEC’s focus in an audit would be the sufficiency of that framework (i.e. are you using the best factors and in the right combinations – also, irrespective of whether you included responses to the SEC factors, should those factors in fact be part of your regular model). In sum, I don’t see that testing to suit the Form PF will mask deficiencies, if any, in the true and used framework.
NOTE I now post my TRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!