Seasonal Patterns in U.S. Equity Volatility

Transcription

1 Seasonal Patterns in U.S. Equity Volatility Gerald A. Hanweck, Jr., PhD Chief Executive Officer Hanweck Associates, LLC 61 Broadway, Suite 1608 New York, NY Tel:

2 Agenda Conventional Wisdom Seasonality of U.S. Equity Volatility Realized and implied volatility patterns Drivers Other Calendar Patterns in Volatility Days of the week Holidays Weekends vs. trading days Interactive Data s Options Volatility Service Database content Queries 2

3 Some Conventional Wisdom Earnings season is more volatile than other periods Summer tends to be quiet October is the most volatile month Weekends don t deliver much volatility Days before holidays are quieter than normal days Options expiration is typically very volatile How wise is the conventional wisdom? 3

4 Realized Volatility Patterns - SPX relative realized variance (rebased, Jan = 1) Chart shows daily close-toclose realized variance of the SPX for each month over the period from present. The values are relative, and have been rebased so that January equals one. So far, so good... roughly speaking: Earnings seasons appear at least as volatile as other periods Summer tends to be quiet October is the most volatile month...but how robust are these observations? 4

5 Realized Volatility Patterns - SPX relative realized variance (rebased, Jan = 1) , ex Sep 2001 Chart shows daily close-toclose realized variance of the SPX for each month over the period from , excluding September The values are relative, and have been rebased so that January equals one. Excluding the financial crisis and September 2001 paints a more muted picture: Some months (Feb, May, Jun, Nov, Dec) appear less volatile than other months; is this related to earnings? Summer seasonality goes away October is the most volatile month, but barely 5

6 Realized Volatility Patterns - SPX Realized volatility; annualized 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% box describes middle two quartiles; line indicates min and max Chart shows distribution of daily close-to-close realized volatility (annualized) of the SPX for each month over the period from , excluding September The box describes the middle two quartiles; the whisker indicates min and max. The distribution of volatility shows things are even less clear: Some months look consistently quiet, particularly Feb, Jun, Nov, Dec Spring and summer look quiet when July 2002 (July max) is excluded October still looks most volatile, perhaps tied with March? Regressions and other statistical tests cannot distinguish a monthly pattern 6

7 Realized Volatility Patterns - SPX relative realized OHLC variance (rebased, Jan = 1) 2.0 Chart shows daily Garman- Klass (OHLC) realized variance of the SPX for each month over the period from 1998-present. The values are relative, and have been rebased so that January equals one. However, the story is not over intraday volatility shows yet another perspective: Feb, Jun and Dec still appear less volatile than other months Summer tends to be relatively volatile October is clearly the most volatile month 7

8 Realized Volatility Patterns Other Indices relative realized variance (rebased, Jan = 1) , ex Sep 2001 SPX Charts show daily close-to-close realized variance of the SPX, DJX and NDX for each month over the period from , excluding September The values are relative, and have been rebased so that January equals one. relative realized variance (rebased, Jan = 1) 2.0 DJX relative realized variance (rebased, Jan = 1) 2.0 NDX 8

9 Implied Volatility Patterns SPX, DJX, NDX relative implied variance (rebased, Jan = 1) SPX Charts show one-month ATM implied volatility (variance) of the SPX, DJX and NDX for each month over the period from , excluding September The values are relative, and have been rebased so that January equals one. relative implied variance (rebased, Jan = 1) DJX relative implied variance (rebased, Jan = 1) NDX 9

10 Daily Realized Volatility - SPX Variance Ratio SPX 1/1/ /12/2010 Realized Volatility 30% var ratio vol Mon-Tue Tue-Wed Wed-Thu Thu-Fri Fri-Mon 3-Day Wkend Fri before 3-Day Wkend Fri after T'giving Expiry Friday Days of the week demonstrate some consistent patterns: Fridays tend to be quiet, particularly Fridays before long weekends The weekend seems to contribute some amount of volatility Expiry Fridays are not particularly volatile 25% 20% 15% 10% 5% 0% Chart shows daily close-toclose realized variance and volatility of the SPX for each category of day over the period from 1998-present. The variances are relative, and have been rebased so that the average of M-F equals one. 10

11 Options Volatility Service Historical, end-of-day equity options database Includes: Historical pricing for options and underlying securities Corporate action histories Option contract adjustments Option analytics (implied volatilities and greeks, interpolated surfaces) Organized in relational-database format Nightly update files and sample update scripts Hosted or deployed on customer s database servers 11

12 Options Volatility Service Sample Queries as int = (select top 1 eqid from eqmaster where ticker='spx' and getdate() between startdate and enddate) ;with prices as (select date_, close_, row_number() over(order by date_ asc) as 'rownum from eqprice where eqid as (select t1.date_, year(t1.date_) * month(t1.date_) as 'yyyymm', log(t1.close_ / t0.close_) as 'logrtn from prices as t1, prices as t0 where t1.rownum = t0.rownum + 1),vars as (select yyyymm, floor(yyyymm/100) as 'yyyy', yyyymm % 100 as 'mm', var(logrtn) as 'hvar from logrtns group by yyyymm) select mm, avg(hvar)*252 as 'hvar from vars where yyyy <= 2007 and yyyymm!= group by mm order by mm asc ;with ivars as (select date_, year(date_) * month(date_) as 'yyyymm', ivmid * ivmid as 'ivar_ from ivcmpr where eqid and expiry = 1 and strike = 100),dates as (select min(date_) as 'date_' from ivars group by yyyymm),vars as (select iv.date_, iv.yyyymm, floor(iv.yyyymm/100) as 'yyyy', iv.yyyymm % 100 as 'mm', iv.ivar_ as 'ivar from ivars iv, dates dt where iv.date_ = dt.date_) select mm, avg(ivar) as 'ivar from vars where yyyy <= 2007 and yyyymm!= group by mm order by mm asc 12

13 This presentation has been prepared for the exclusive use of the direct recipient. No part of this presentation may be copied or redistributed without the express written consent of the author. Opinions and estimates constitute the author s judgment as of the date of this material and are subject to change without notice. Information has been obtained from sources believed to be reliable, but the author does not warrant its completeness or accuracy. Past performance is not indicative of future results. Securities, financial instruments or strategies mentioned herein may not be suitable for all investors. The recipient of this report must make its own independent decisions regarding any strategies, securities or financial instruments discussed. This material is not intended as an offer or solicitation for the purchase or sale of any financial instrument. Copyright 2008, 2009, 2010 Hanweck Associates, LLC. All rights reserved. Additional information is available upon request. 13