Order Making Fiscal Year 2016 Annual Adjustments to Transaction Fee Rates

I. Background

Section 31 of the Securities Exchange Act of 1934 (“Exchange Act”) requires each national securities exchange and national securities association to pay transaction fees to the Commission. Specifically, Section 31(b) requires each national securities exchange to pay to the Commission fees based on the aggregate dollar amount of sales of certain securities (“covered sales”) transacted on the exchange. Section 31(c) requires each national securities association to pay to the Commission fees based on the aggregate dollar amount of covered sales transacted by or through any member of the association other than on an exchange.

Section 31 of the Exchange Act requires the Commission to annually adjust the fee rates applicable under Sections 31(b) and (c) to a uniform adjusted rate. Specifically, the Commission must adjust the fee rates to a uniform adjusted rate that is reasonably likely to produce aggregate fee collections (including assessments on security futures transactions) equal to the regular appropriation to the Commission for the applicable fiscal year.

The Commission is required to publish notice of the new fee rates under Section 31 not later than 30 days after the date on which an Act making a regular appropriation for the applicable fiscal year is enacted. On December 18, 2015, the President signed the “Consolidated Appropriations Act, 2016”, providing $1,605,000,000 in funds to the SEC for fiscal year 2016.

II. Fiscal Year 2016 Annual Adjustment to the Fee Rate

The new fee rate is determined by (1) subtracting the sum of fees estimated to be collected prior to the effective date of the new fee rate and estimated assessments on security futures transactions to be collected under Section 31(d) of the Exchange Act for all of fiscal year 2016 from an amount equal to the regular appropriation to the Commission for fiscal year 2016, and (2) dividing by the estimated aggregate dollar amount of sales for the remainder of the fiscal year following the effective date of the new fee rate.

The regular appropriation to the Commission for fiscal year 2016 is $1,605,000,000. The Commission estimates that it will collect $502,582,684 in fees for the period prior to the effective date of the new fee rate and $35,649 in assessments on round turn transactions in security futures products during all of fiscal year 2016. Using a new methodology described below, the Commission estimates that the aggregate dollar amount of covered sales for the remainder of fiscal year 2016 to be $50,672,728,301,509.

The uniform adjusted rate is computed by dividing the residual fees to be collected of $1,102,381,667 by the estimate of the aggregate dollar amount of covered sales for the remainder of fiscal year 2016 of $50,672,728,301,509; this results in a uniform adjusted rate for fiscal year 2016 of $21.80 per million.

III. Effective Date of the Uniform Adjusted Rate

Under Section 31(j)(4)(A) of the Exchange Act, the fiscal year 2016 annual adjustments to the fee rates applicable under Sections 31(b) and (c) of the Exchange Act shall take effect on the later of October 1, 2015, or 60 days after the date on which a regular appropriation to the Commission for fiscal year 2016 is enacted. The regular appropriation to the Commission for fiscal year 2016 was enacted on December 18, 2015, and accordingly, the new fee rates applicable under Sections 31(b) and (c) of the Exchange Act will take effect on February 16, 2016.

IV. New Methodology for the Baseline Estimate of the Aggregate Dollar Volume of Covered Sales

The methodology used to generate the baseline estimate of the aggregate dollar amount of covered sales is required to be developed by the Commission in consultation with the Congressional Budget Office (“CBO”) and the Office of Management and Budget (“OMB”). The Commission recently completed a comprehensive review of the methodology and determined that modifications to the methodology would improve the accuracy of the estimates. The Commission consulted with CBO and OMB regarding the modifications to the methodology, as required under Section 31 of the Exchange Act. Consequently, the Commission has adopted the new methodology to generate the baseline estimate of the aggregate dollar volume of covered sales, which is used to determine the new fee rates. The methodology is explained in Appendix A attached to this order.

V. Conclusion

Accordingly, pursuant to Section 31 of the Exchange Act,

It is hereby ordered that the fee rates applicable under Sections 31(b) and (c) of the Exchange Act shall be $21.80 per $1,000,000 effective on February 16, 2016.

Appendix A

This appendix provides the methodology for determining the annual adjustment to the fee rates applicable under Sections 31(b) and (c) of the Exchange Act for fiscal year 2016. Section 31 of the Exchange Act requires the fee rates to be adjusted so that it is reasonably likely that the Commission will collect aggregate fees equal to its regular appropriation for fiscal year 2016.

To make the adjustment, the Commission must project the aggregate dollar amount of covered sales of securities on the securities exchanges and certain over-the-counter markets over the course of the year. The fee rate equals the ratio of the Commission's regular appropriation for fiscal year 2016 (less the sum of fees to be collected during fiscal year 2016 prior to the effective date of the new fee rate and aggregate assessments on security futures transactions during all of fiscal year 2016) to the estimated aggregate dollar amount of covered sales for the remainder of the fiscal year following the effective date of the new fee rate.

For 2016, the Commission has estimated the aggregate dollar amount of covered sales by projecting forward the trend established in the previous decade. More specifically, the dollar amount of covered sales was forecasted for months subsequent to November 2015, the last month for which the Commission has data on the dollar volume of covered sales.

The following sections describe this process in detail.

A. Baseline estimate of the aggregate dollar amount of covered sales for fiscal year 2016.

First, calculate the average daily dollar amount of covered sales (ADS) for each month in the sample (October, 2005-November, 2015). The monthly total dollar amount of covered sales (exchange plus certain over-the-counter markets) is presented in column C of Table A.

Next, model the monthly change in the natural logarithm of ADS as a first order autoregressive process (“AR(1)”), including monthly indicator variables to control for seasonality.

Use the estimated AR(1) model to forecast the monthly change in the log level of ADS. These percent changes can then be applied to obtain forecasts of the total dollar volume of covered sales. The following is a more formal (mathematical) description of the procedure:

1. Begin with the monthly data for total dollar volume of covered sales (column C). The sample spans ten years, from October, 2005-November, 2015. Divide each month's total dollar volume by the number of trading days in that month (column B) to obtain the average daily dollar volume (ADS, column D).

2. For each month t, calculate Δ LN ADS (shown in column E) as the log growth rate of ADS, that is, the difference between the natural logarithm of ADS in month t and its value in the prior month.

3. Estimate the AR(1) model

with D ^m _t representing monthly indicator variables, Yt representing the log growth rate in ADS (Δ LS ADS), and ε_t representing the error term for month t. The model can be estimated using standard commercially available software. The estimated parameter values are β =⁻ 0.2671 and α − α₁₂ as follows:

α₁ (JAN) = 0.0854, α₂ (FEB) = 0.0425, α₃ (MAR) = 0.0124, α₄ (APR) =⁻ 0.0466, α₅ (MAY) = 0.0501, α₆ (JUN) = 0.0031, α₇ (JUL) =⁻ 0.0482, α₈ (AUG) =⁻ 0.0004, α₉ (SEP) = 0.0335, α₁₀ (OCT) = 0.0614, α₁₁ (NOV) =⁻ 0.0296, α₁₂ (DEC) =⁻ 0.0801. The root-mean squared error (RMSE) of the regression is 0.1140.

4. For the first month calculate the forecasted value of the log growth rate of ADS as

For the next month use the forecasted value of the log growth rate of the first month to calculate the forecast of the next month. This process iterates until a forecast is generated for all remaining months in the fiscal year. These data appear in column F.

5. Assuming that the regression error in the AR(1) model is normally distributed, the expected percentage change in average daily dollar volume from month t - 1 to month t is then given by the expression

where σ denotes the root mean squared error of the regression (RMSE).

6. For instance, for December 2015, using the β parameter and the α₁₂ parameter (for December) above, and the change in the log-level ADS from November, 2015, we can estimate the change in the log growth in average daily sales as β γ_Nov + α_Dec = ((−0.2671 ×−0.02892) −0.0801) = −0.0724. This represents the estimated change in log average daily dollar volume for December 2015 relative to November 2015. To estimate the percent change in average daily sales from November, 2015 to December, 2015, use the formula shown in Step 5, above: exp (−0.0724 + 1/2 0.1140²) = −0.0638. Apply this estimated percent change in ADS to the ADS for November, 2015 to estimate the ADS for December, 2015 as $291,167,469,596 × (1−0.0638) = $272,602,991,941. Multiply this by the 22 trading days in December 2015 to obtain a total dollar volume forecast of $5,997,265,822,693.

7. For January 2016, proceed in a similar fashion. Using the estimates for December, 2015 along with the β parameter and the α₁ parameter (for January) to generate a forecast for the one-month change in the log level of average daily sales. Convert the estimated log change in average daily sales to estimated percent change in ADS as in step 6, above to obtain a forecast ADS of $304,668,090,424. Multiply this figure by the 19 trading days in January 2016 to obtain a total dollar volume forecast of $5,788,693,718,050.

8. Repeat this procedure for subsequent months.

B. Using the Forecasts From A To Calculate the New Fee Rate

1. Use Table A to estimate fees collected for the period 10/1/15 through 2/15/16. The projected aggregate dollar amount of covered sales for this period is $27,314,276,282,567. Actual and projected fee collections at the current fee rate of $18.40 per million are $502,582,684.

2. Estimate the amount of assessments on security futures products collected from 10/1/15 through 9/30/16. First, calculate the average and the standard deviation of the change in log average daily sales, in column E. The average is 0.005148 and the standard deviation is 0.12233. These are used to estimate an average growth rate in ADS using the formula exp (0.005148 + 1/2 0.12233²)−1. This results in an average monthly increase of 1.271%. Apply this monthly increase to the last month for which single stock futures' assessments are available, which was $2,828.72, for November, 2015. Estimate all subsequent months in fiscal year 2016 by applying the growth rate to the previously estimated monthly value, and sum the results. This totals $35,649 for the entire fiscal year.

3. Subtract the amounts $502,582,684 and $35,649 from the target offsetting collection amount set by Congress of $1,605,000,000 leaving $1,102,381,667 to be collected on dollar volume for the period 2/16/2016 through 9/30/2016.

4. Use Table A to estimate dollar volume for the period 2/16/2016 through 9/30/2016. The estimate is $50,672,728,301,509. Finally, compute the fee rate required to produce the additional $1,102,381,667 in revenue. This rate is $1,102,381,667 divided by $50,672,728,301,509 or 0.00002175493.

5. Round the result to the seventh decimal point, yielding a rate of 0.0000218 (or $21.80 per million).