This article has been updated to reflect the fact that the SEC voted to adopt Rule 18f-4, incorporating changes to the original proposed rules. 

The U.S. Securities and Exchange Commission has passed a set of rules to limit leverage and use of derivatives by US funds. Dubbed 18f-4, the new SEC fund leverage rules will pose a range data management challenges for the investment management community. Any fund holding derivatives in their portfolio (of a value greater than 10% of NAV) will be obliged to comply with certain process requirements (specifically around derivatives risk management) and leverage caps (restricting use of those derivatives).  Those caps will be based on the daily computation of risk metrics and validation tests to ensure funds remain within their limits.

What are the Test Metrics?

The leverage caps will be based on either a ‘relative’ or ‘absolute’ test, which will need to be performed on a daily basis. The test metric is Value at Risk (VaR) calculated over a rolling 20-day period using a 99% confidence interval. The relative test will compare the fund’s VaR against a “designated reference index” (DRI) and would require that the ratio does not exceed 200% (although some closed-end funds will have a limit of 250%). 

Should the fund not have an appropriate index to compare itself against, then the absolute test would require that the VaR measure does not exceed 20% of the fund’s value (in other words, that the fund is 99% likely not to fall by more than 20% over a 20-day period). These limits have been raised from the 150% relative test and 15% absolute tests that were originally proposed.   

How Will VaR be Calculated?

The rules do not specify a methodology for calculating VaR, suggesting that derivatives risk managers should be able to choose between historical, parametric or Monte Carlo models depending on which is most appropriate for the fund and its situation. One specification is that funds should base their VaR calculation on at least three years of historical market data. There is also guidance offered on which risk factors should be included in a firm’s VaR models. The list includes delta (linear relationship with underlying equity prices, interest rates, credit spreads, FX rates and commodity prices); gamma (nonlinear relationships with underlying), and vega (volatility). In response to comments, the Commission has also recognized that some funds may prefer to calculate VaR at a 95% confidence interval and then rescale the value to achieve a 99% confidence interval, given that this approach allows for the calculation to take into account additional market data observations. 

Who is in Scope of the Rule?

The proposed rule would apply to any “fund,” including “open-end” mutual funds and exchange traded funds (ETFs) as well as “closed-end” investment funds or “business development companies” (BDCs). There will be exemptions to the rule. One fund category that would fall outside of the rule are money market funds, which are regulated under rule 2a-7 of the Investment Company Act. Also, if a fund’s total derivatives exposure is less than 10% of its net asset value (excluding derivatives specifically designed to hedge currency or interest rate risk), then it should also be exempt. Finally, funds already in issue that are specifically designed to provide leveraged or inverse exposure to a particular index will be exempt (although subject to additional suitability checks when sold to retail investors), but future leveraged/inverse funds will continue to be in scope (thus effectively capping their leverage at 200%).

Are firms required to back-test VaR models?

The back-testing requirements are relatively stringent and require VaR models to be back-tested on a weekly basis (based on daily data). A fund would need to compare a one-day VaR measure, calculated at the close of the previous business day, against its actual gain or loss for that day. Using a 99% confidence interval would result in approximately 2 ½ expected breaches per year.

It is interesting to note that other equivalent regulations (such as UCITS in Europe) require less frequent back-testing (monthly, although also using daily data). The SEC had originally proposed daily back-testing but has limited the frequency to weekly in response to comments from market participants. 

What needs to be reported?

Under the new rules, most of the reporting obligations are designed to help regulators monitor the effectiveness of derivatives risk management practices and alerting to potential issues. Original proposals for some VaR data to be made public have now been dropped. 

Reporting Limit Breaches – Form N-RN

With regards to the relative and absolute test results, it is important to note that a single breach will not immediately trigger a reporting requirement. Instead, firms are afforded a few business days to come back into compliance. It is only if test limits (200% for the relative test, and 20% for absolute) have been breached for five consecutive days (raised from the originally proposed three days) that a firm would have to file a report to the regulator (no later than the day after the fifth consecutive breach). Should this requirement be triggered, the information to be submitted includes:

• Dates on which the test breaches occurred (fund portfolio’s VaR exceeded 200% of the VaR of its designated reference index for relative test; or portfolio VaR exceeded 20% of net asset value for absolute test)
• The VaR of its portfolio for each of these days
• The value of the fund’s net assets for each of these days (if subject to absolute test)
• The VaR of its designated reference index for each of these days (if subject to relative test)
• The name of the designated reference index and index identifier (if subject to relative test)

Regular Monthly Reports

Under the current proposals some data will also be reported on a more regular basis – monthly – as part of the SEC’s new N-PORT filings. 

Data to be reported includes:

• Median daily VaR (and VaR ratio if applicable) for the monthly reporting period (this information will no longer be made public as had been originally proposed)
• Name of the fund’s designated reference index and index identifier (if applicable)
• Number of exceptions the fund identified from back-testing its VaR model (this information will no longer be made public as had been originally proposed)

When Will This Happen?

SEC fund leverage rules 18f-4 was originally proposed in 2015, but received significant pushback from market participants and did not make it into law. The regulator has taken on-board market feedback and made changes to the re-proposed rules. The re-proposed rules were voted through on the 28th of October, and will become effective 60 days after their publication in the Federal register. Market participants will then have 18 months before the effective compliance deadline.

How Should Firms Operationalise Compliance with the New Requirements?

Give us a call!