Sell in May and Go Away in the Equity Index Futures Markets, 1993-2019
Total Page:16
File Type:pdf, Size:1020Kb
Sell in May and Go Away in the Equity Index Futures Markets, 1993-2019 Constantine Dzhabarov William T. Ziemba SRC Special Paper No 17 January 2020 ISSN 2055-0375 Abstract The period May 1 to the turn of the month of November (-6 trading day of October) has historically produced negligible returns. The rest of the year (late October to the end of April) has essentially all the year's gains. We show that there is a statistically significant difference and conclude that the strategy go to cash in the weak period and go long in the strong period has about 1 ½ times the returns of buy and hold for large cap S&P500 index and 2 times for the small cap Russell 2000 index during the period 1993-2019 in the index futures markets. Graphs and tables of returns for the entire 1993-2019 period are shown. The 2016-2019 time period had very high returns. For the S&P500, buy and hold and sell in May had essentially identical returns. The Russell 2000 sell in May is higher than buy and hold but with only three years data it is not statistically significantly higher than buy and hold. JEL classification: C19, C41, G11. Keywords: calendar anomalies, sell-in-May effect, monthly small cap and large cap returns, equity futures markets This paper is published as part of the Systemic Risk Centre’s Special Paper Series. The support of the Economic and Social Research Council (ESRC) in funding the SRC is gratefully acknowledged [grant number ES/R009724/1]. Constantine Dzhabarov, Consultant, Surrey, British Columbia, Canada William T. Ziemba, University of British Columbia, Vancouver, British Columbia and Systemic Risk Centre, London School of Economics and Political Science Published by Systemic Risk Centre The London School of Economics and Political Science Houghton Street London WC2A 2AE All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means without the prior permission in writing of the publisher nor be issued to the public or circulated in any form other than that in which it is published. Requests for permission to reproduce any article or part of the Working Paper should be sent to the editor at the above address. © Constantine Dzhabarov and William T. Ziemba, submitted 2020 Sell in May and Go Away in the Equity Index Futures Markets, 1993-2019 ∗ Constantine Dzhabarovy and William T. Ziemba** yConsultant, Surrey, BC, Canada, [email protected] **Alumni Professor of Financial Modeling and Stochastic Optimization (Emeritus), University of British Columbia, Vancouver, BC and Distinguished Visiting Research Associate, Systemic Risk Centre, London School of Economics, [email protected] November 29, 2019 Abstract The period May 1 to the turn of the month of November (-6 trading day of October) has historically produced negligible returns. The rest of the year (late October to the end of April) has essentially all the year's gains. We show that there is a statistically significant difference and conclude that the strategy go to cash in the weak period and 1 go long in the strong period has about 1 2 times the returns of buy and hold for large cap S&P500 index and 2 times for the small cap Russell 2000 index during the period 1993-2019 in the index futures markets. Graphs and tables of returns for the entire 1993-2019 period are shown. The 2016-2019 time period had very high returns. For the S&P500, buy and hold and sell in May had essentially identical returns. The Russell 2000 sell in May is higher than buy and hold but with only three years data it is not statistically significantly higher than buy and hold. JEL: C19, C41, G11 keywords: calendar anomalies, sell-in-May effect, monthly small cap and large cap returns, equity futures markets 1 Sell-in-May-and-go-away August, September and October have historically had low stock market returns with many serious declines or crashes occurring in these months. The months of November to February have historically had higher than average returns; see, for example, Gultekin and Gultekin ∗Thanks go to Alexandre Ziegler for helpful comments on an earlier version of this paper. 1 S&P500 Futures Sell in May (SIM) and B&H Cumulative Returns. 1993-2019. (Entry at Close on 6th Day before End of October. Exit 1st Day of May.) 10 SIM Daily NAV B&H Daily NAV 8 6 4 2 0 (1983) and various papers in Keim and Ziemba (2000), Ziemba (2012) and the tables in this 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 paper. This suggests the strategy to avoid the bad months and be in cash then and only be long the stock market in the good months. Sell-in-May-and-go-away, which is sometimes called the press. Halloween Effect , is one such strategy that is often discussed in the financial Figure 1: S&P500 Futures Sell in May (SIM) and B&H Cumulative Returns Comparison. 1993-2019. (Entry at Close on 6th Day before End of October. Exit 1st Day of May.) For the S&P500 a buy and hold strategy turns $1 on February 4, 1993 into $6.71 on October 25, 2019; whereas, sell in May and move into cash had a final wealth of $9.28, some 38.2% higher. For sell in May this counts interest (Fed funds effective monthly rate) and for the buy and hold this counts dividends (which for futures are capital gains). For the small cap Russell2000, the final wealths were $6.67 and $12.61, respectively, some 89.0% higher. Figures 1 and 2 show this strategy using the rule sell on the first trading day in May and buy on the 6th trading day before the end of October, for the S&P500 and Russell 2000 index futures for the years 1993-2019, respectively. This rule did indeed beat a buy and hold strategy by about 1 risk, respectively. Tables 1 and 2 show the monthly returns, respectively, for those 26 years to the end of October 2019. The strategy worked in most but not all years and in strategy design can be combined with other effects depending upon market conditions. This advantage is greater in absolute total returns, but the percentage advantage was not 1 2 times and 2 times in final wealth with lower standard deviation 2 equty index futures, 1993-2019 Table 1: Data by month for sell in May and go away versus buy and hold for the S&P500 0.014 0.1270.0150.021 3.6210.029 1.9670.026 0.50% 1.4460.028 3.552 20.64% 2.5220.048 15.23% 14.80% 0.2610.029 37.39% 0.018 26.89% 1.2430.017 1.8290.008 3.09% 0.3150.022 3.5790.050 12.72% 0.5530.060 12.14% 0.0900.027 1.69% 0.1190.013 10.28% 1.5090.018 3.99% 1.8470.018 0.22% 2.0420.020 0.47% 2.5650.022 14.62% 0.9850.027 8.88% 0.1580.013 13.06% 0.710 17.00% 2.5570.026 6.86% 0.94% 2.196 6.48% 11.85% 19.19% 0.020 0.5940.0160.032 5.5280.045 1.7420.065 3.78% 1.8440.040 1.185 34.05% 1.352 20.38% 31.51% 0.033 27.21% 0.021 19.16% 2.1060.023 1.2610.017 0.4630.028 2.240 26.20% 0.3900.064 9.42% 0.055 3.39% 1.0570.047 13.83% 0.7250.030 3.42% 0.0710.025 1.2730.024 23.40% 3.0860.041 12.81% 1.3290.030 -0.03% 0.0060.012 13.37% 0.970 29.65% 4.4670.042 11.48% -0.83% 1.536 9.87% 19.67% 21.18% 0.74% 0.05% 1.59% 1.21% 1.19% 2.71% 2.04% 0.36% 1.04% 0.97% 0.15% 0.82% 0.35% 0.13% 0.21% 1.18% 0.72% 1.04% 1.33% 0.57% 0.10% 0.56% 0.95% 1.80% 0.68% 0.36% 2.48% 1.60% 2.40% 2.23% 1.54% 2.01% 0.77% 0.30% 1.10% 0.32% 1.96% 1.15% 0.10% 1.09% 2.22% 0.94% 0.01% 0.83% 1.51% 2.02% 0.10% 1.47% 1.44% 1.01% -0.66% 3 1.56% 1.70% 0.63% 0.22% 0.23% 0.23% 0.22% 1.34% 1.44% 1.01% 1.56% 1.70% 0.35% 0.26% 0.70% -0.09% -1.28% 1.89% -3.08% 1.05% 0.28% 0.26% 0.27% 0.26% 1.18% -1.32% 1.10% -0.09% -1.28% 1.89% -3.08% 2.61% 0.20% -0.67% 3.40% -0.84% 1.89% -1.32% 1.10% Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Mthly Avg StDev t Geom r Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Mthly Avg StDev t Geom r 0.0430.480 0.042 -0.107 0.039 2.104 0.035 2.534 0.007 4.420 0.002 6.060 0.002 6.094 0.002 6.143 0.002 6.129 0.029 2.423 0.041 1.810 0.035 1.487 0.0430.480 0.042 -0.107 0.039 2.104 0.035 2.534 0.036 0.492 0.037 0.359 0.039 0.936 0.044 -0.774 0.050 0.100 0.055 1.383 0.041 1.810 0.035 1.487 3.16% -3.23%2.35%3.15% 2.75% -4.16%5.78% 0.05% 3.96% 0.81%0.89% 0.37% 2.04%2.89% 2.45% 0.88% 6.35% -4.10% -3.59% 0.55% 0.22% 0.35% 5.71% 5.91%3.20% 4.68% 0.68% -9.80% 0.50% 0.78% 0.34% 0.35% 3.34% 0.41% -5.56% 0.51% 1.36% 0.41% 0.48%1.87% 2.03% 0.50% 7.25% 0.50% 0.49% 0.39% 1.30% 0.48% 0.44%2.30% 0.44% 1.79% -1.72% 0.47% 3.41% 0.48% 0.45% -0.09%1.02% 0.42% 0.46% 0.32% -1.67% -4.71% -0.52% -2.36% 0.48% 0.49% 1.63% 0.45% 1.19% 0.34% 0.97% 4.01% 1.63% 1.58% 0.46% -1.76% 0.97% 0.44% 6.85% 0.32% 0.09% 1.84% 4.00% 3.34% 3.23%2.35% -0.18% -0.06% -1.81% 5.61%4.44% 0.25% 0.10% 3.41% 0.73% 5.18% 2.53% 0.42% 0.0265.15% 1.11% 4.30% -2.06% -0.38% 0.13% 0.42% 7.14% -0.079 0.45% 1.34% 2.84% 6.66% 7.48% 2.93% 0.13% 0.47% 0.45% 3.26% -0.68% -0.14% 0.81% 3.25% 0.43%1.71% 1.89% 0.42% 0.50% -1.10% 0.01% 3.88%5.60% 3.88% 0.39% -0.93% 0.25% 0.01% -3.94%7.96% 0.01% -0.15% 3.37% 2.52% 0.01% 1.42% 2.97% -2.63% 0.01% 0.01% 0.90% -4.58% 0.01% 1.82% 1.91% 0.15% 0.01% 0.01% -0.44% 0.34% 0.01% 3.90% 0.34% 0.01% 0.18% 0.09% 0.01% -0.66% 0.15% -0.26% 0.12% 0.10% 0.15% 0.19% 0.17% 0.54% 0.53% 0.10% 3.03% 0.22% 0.16% 0.40% 0.09% 2.05% 0.18% 0.15% 0.29% 0.18% 1.92% 2.92% 1.18% 1.74% 1.06% -9.18% -0.45% 0.036 -0.431 -5.94% 0.40% 3.16% -3.23%2.35%3.15% 2.75% -4.16%5.78% 0.05% 3.96% 0.81%0.89% 0.37% 2.04%2.89% 2.45% 1.30% 6.35% -4.10% -3.59% 0.55% -2.44% 3.21% 5.71% 5.91%3.20% 4.68% 1.85% 3.11% -9.80% 2.59% 0.78% 5.95% 3.34% 1.48% -5.56% -2.54% 3.31% 2.92% 4.66%1.87% -2.94% -5.08% 7.25% -2.13% -0.05% 4.79% 1.30%