WahLau Backtesting

"Sell In May And Go Away" "5窮". Can we use it in KLSE?

Wah Lau
Publish date: Mon, 26 Oct 2015, 09:36 PM
Wah Lau
0 41
Backtesting study.

This month, I go few seminars. Those trainers talked about "Sell In May And Go Away".=.= I first time listen it, due to I still new here. I google it and found points as below:

 

According to the Stock Trader's Almanac, since 1950, the Dow Jones Industrial Average has had an average return of only 0.3% during the May-October period, compared with an average gain of 7.5% during the November-April period. 

Read more: Sell In May And Go Away Definition | Investopedia http://www.investopedia.com/terms/s/sell-in-may-and-go-away.asp#ixzz3pfzcXG3W

Or

http://www.businessinsider.my/chart-sell-in-may-and-go-away-2014-4/#FjjdHOm4I015Wug3.97

 

 

I taken KLSE data 1994-2014 (20 years). The average KLSE monthly changed as below:

  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Average 3.08 15.04 -8.59 11.22 0.10 -2.69 3.98 -17.88 -11.02 12.74 -5.91 22.74

If we buy in start of May and wait until Sep (5 months). Average change KLSE -27.52. 

But If we buy in Oct until Feb, Average change KLSE +47.68.

According chart as below best buying point is Oct and bad buying point is May.

 

 

I try take DOW 30 (DJI) 994-2014 (20 years), so we can compare with KLSE. And both chart about same with KLSE chart.

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Average -43 7.15 134 253 -11 -91 82 -83.8 -63.1 154 179 154

 

**Just sharing as study case and take your own risk to use it in your investment.

 

If you need more data, check below:

 

KLSE

  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1994 -164.29 10.64 -170.47 99.64 -62.53 16.83 21.15 105.62 -1.72 -25.55 -96.22 -40.87
1995 -76.98 85.97 1.13 -30.58 107.75 -32.28 34.06 -44.95 -14.25 -44.17 -5.07 43.25
1996 64.77 25.20 64.22 43.27 -52.40 -3.97 -68.42 49.82 18.62 32.52 58.72 10.89
1997 -21.24 53.95 -67.57 -122.93 27.38 -30.25 -64.52 -201.63 -0.29 -138.60 -135.15 53.95
1998 55.04 99.51 -27.41 -92.96 -87.64 -82.87 -51.32 -108.65 77.92 33.47 94.84 75.97
1999 14.51 -50.58 -37.98 172.87 66.92 77.00 -50.78 -2.91 -90.03 64.12 -5.26 78.14
2000 115.91 52.48 -3.86 -78.85 15.46 -83.99 -30.54 -4.33 -83.54 41.40 -24.19 -52.49
2001 53.32 -21.76 -61.87 -54.10 -18.73 18.36 68.40 26.29 -69.20 -17.40 38.08 57.37
2002 25.47 -11.93 47.68 38.65 -53.17 -14.17 -6.40 -10.58 -72.84 21.85 -29.99 17.22
2003 17.76 -17.71 -15.59 -0.29 41.76 19.01 29.13 22.11 -9.89 84.64 -38.68 14.05
2004 24.87 60.01 18.50 -56.25 -28.76 8.81 12.22 -3.73 20.85 10.28 56.02 -9.09
2005 10.78 -12.88 -31.49 5.36 -17.74 27.13 50.12 -25.22 13.76 -16.31 -15.08 3.73
2006 16.43 11.17 -0.15 21.17 -20.15 -13.04 20.80 23.63 8.13 21.13 91.95 16.51
2007 99.15 8.76 45.57 78.89 19.23 9.42 9.51 -83.36 53.72 81.73 -18.83 43.13
2008 -50.43 -48.79 -95.59 39.04 -13.36 -86.56 -28.64 -62.27 -80.60 -149.44 -2.55 13.81
2009 7.88 3.53 -13.75 114.78 56.88 27.51 99.77 -0.63 28.14 35.76 26.18 7.96
2010 -11.11 15.32 43.27 24.45 -59.99 28.72 51.26 65.54 38.68 42.24 -28.00 41.84
2011 0.21 -29.86 52.16 -11.05 25.02 19.82 -31.04 -89.13 -78.85 102.32 2.79 36.68
2012 3.01 42.28 25.68 -22.10 1.23 26.17 28.20 21.41 -14.10 40.47 -67.54 76.84
2013 -50.61 1.93 35.41 43.13 56.54 0.43 3.73 -51.94 42.60 37.31 9.07 52.98
2014 -69.84 38.55 11.66 23.40 -1.70 11.44 -13.13 -0.65 -18.46 9.82 -35.30 -64.37
Average 3.08 15.04 -8.59 11.22 0.10 -2.69 3.98 -17.88 -11.02 12.74 -5.91 22.74

 

Dow 30 or DJI

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1994 224.3 -146.4 -197.5 47.2 76.7 -133.4 139.5 148.6 -69.9 64.58 -168.58 95.2
1995 9.5 167.2 146.6 163.6 143.8 91 152.4 -98.3 179.6 -36.1 321.9 40.1
1996 279.2 93.3 100.4 -21 76.3 10.7 -125.7 82.9 266.6 148.7 493.5 -73.1
1997 363.3 63.5 -290.8 425.49 322.01 347 544.8 -591.1 316.8 -502.7 379.32 85.08
1998 -0.21 634.48 257.09 265.05 -165.66 58.04 -75.7 -1337.53 295.57 754.37 520.85 67.46
1999 175.59 -52.21 479.05 1002.6 -228.03 405.04 -311.22 173.61 -492.75 395.09 145.69 625.38
2000 -564.11 -809.63 735.17 -113.86 -217.15 -81.91 73.45 695.73 -560.48 307.15 -549.45 374.16
2001 93.9 -391.57 -616.09 856.89 179.52 -408.62 22.43 -580.4 -1101.01 241.48 761.48 172.78
2002 -98.67 188 291.03 -457.17 -20.96 -684.69 -506.67 -73.31 -1066.23 802.6 507.31 -560.57
2003 -288.64 -163.5 102.59 485.65 372.97 132.21 249.02 183.99 -139.87 525.66 -17.11 667.39
2004 35.04 94.47 -224.73 -130.25 -40.09 246.82 -295.55 31.67 -88.08 -53.31 397.07 357.95
2005 -294.03 279.32 -264.47 -312.57 270.86 -189.27 368.19 -160.34 88.06 -129.43 365.96 -87.73
2006 143.84 131.11 119.75 254.78 -198.75 -19.69 35.46 197.95 296.24 401.26 140.72 238.57
2007 157.66 -348.57 85.89 708.23 565.94 -219.09 -198.51 147.3 537.32 28.45 -555.94 -106.4
2008 -623.65 -373.81 2.28 551.7 -180.67 -1293.03 35.25 165.74 -698.23 -1521.36 -499.15 -54.64
2009 -771.63 -944.14 549.65 561.28 334.12 -54 726.12 318.67 219.28 0.53 631.69 86.87
2010 -361.7 257.11 531.21 152.29 -875.66 -360.67 695.55 -452.81 773.71 330.58 -113.07 570.2
2011 315.07 333.99 94.53 489.14 -240.75 -155.07 -270.12 -530.92 -701.2 1039.43 94.68 174.98
2012 411.57 319.53 259.07 2.8 -822.6 488.15 127.76 84.68 345.51 -338.47 -71.46 76.57
2013 756.28 193.91 524.05 261.26 283.75 -211.95 592.25 -702.3 330.94 425.52 529.11 485.05
2014 -874.48 624.02 136.34 122.21 136.59 111.68 -266.83 535.72 -56.96 350.44 436.37 -4.2
Average -43.42 7.15 134.34 253.11 -10.85 -91.46 81.52 -83.83 -63.10 154.02 178.61 153.86

 

Discussions
2 people like this. Showing 28 of 28 comments

firehawk

yr tables' unit is "point", right? it is better if there is another option in "%", some ppl may feel more meaningful if it is express in % .

2015-10-26 22:02

WahLao

firehawk, I agree with u.I will add it ^^
D7zul, please share it. I still new and I wish can learn more here.

2015-10-27 01:13

YiStock

How about running a statistic to confirm this hypothesis at 95% and 99% confident level. I have used a lot of statistic during higher degree study..very sad to say, have totally forgotten how to input these data. But the result will definetely carry a lot of weight. Regards

2015-10-27 08:40

Ooi Teik Bee

Dear WahLao,
A very good article, I appreciate the hard work done. Hope to read more article from you.
Thank you.
Ooi

2015-10-27 08:43

JT Yeo

1. This is the whole index, which is very different from individual stock movement
2. Past does not represent the future, especially things as superficial as using months to determine your decision
3. Dont be fool by randomness, these are all 'noises', that is one of the problem with back testing. It is just as easy to find the perfect hour (sell at 10am come back at 3pm) or the perfect day (sell on Mon come back at Thurs). Heck you can even use temperature or weather to prove a certain temperature or certain weather has positive return and maybe thunder day has negative return. Which has nothing to do with stocks at ll.

2015-10-27 09:28

JT Yeo

wow D7zul you can make money everyday, 20 days a month since Jan 2015. If 1% a day, this is conservative, because top gainers cant be 1%, you are at least 200% gain right now.

2015-10-27 09:49

kcchongnz

"Sell in May and go away", but do not forget to come back in September was a famous academic research paper by my supervisor, Professor Ben Jacobson, a Dutch national originally, when I did my master in finance.

The paper shows statistical significant in most of the developed and developing countries in the world. He actually have made money doing some fund management with one of his students.

But bear in mind a couple of things:

1) the countries under study were mostly in the Northern Hemisphere, US and Europe in particular where there have their summer holiday starting in May.

2) Any of this anomaly usually will be quickly arbitraged away when something is made popular,and yes, the future will not necessary be like the past.

3) Whether this phenomenon still persists i do not know as the paper was published about 10 years ago.

2015-10-27 10:14

Probability

I think the fundamental reason for above....is generally CEO's of company gives a very pessimistic hope to achieve their forecasts to their shareholders at the beginning of the years (by second quarter of the year) while keeping a few good orders under their sleeves (as done by most sales person to handle the pressure on them for growth)...but as the final 3rd and 4th quarter results are observed....things will prove otherwise.

2015-10-27 10:17

Kevin Wong

Only for mkt timers, otherwise investors better off stay invested at all times over the long term.
Focus on individual stocks instead of the mkt in general then, ""anytime is a good time to invest""...
Anyone believe that it's also anytime is a good time to trade?

My best wishers to all investors/timers/traders...good luck everybody!

2015-10-27 10:44

JT Yeo

I think you are underminig your own potential, charging people RM3.

You see, Genneva guarantee 3% per month or 36% per year, people are willing to refinance their house to buy their gold. Thats the power

Considered you can do more than 200% a year, I dont think Malaysia's top 10 billionaires can match you. Imagine all the individual investors, institutions will come to you.

2015-10-27 11:49

JT Yeo

Well, if they can predict top gainers accurately, why would they want to work in institutions? Same for you, if you can do it consistently over the long time, why do you want to make RM3

2015-10-27 12:59

JT Yeo

if someone can predict the stock, aka economy accurately, they will be worth 100 million, that is a conservative figure. Why would someone worth 100 million wants company car, and climb corporate ladder.

2015-10-27 13:20

fayeTan

the reverse of Sell in May = Buy in October (which is now)...

2015-10-27 15:19

JT Yeo

common sense right, if you can prove you are that good making 200% a year. it is easy for people to give you 10 mil easily. end of the year it will turn into 30 mil, you collect 20% , that is 4 million. You are a millionaire in a year, u dont need capital.

2015-10-27 15:39

Jonathan Keung

nothing is easy. the best trader maximise their profit but cut their losses. nobody says i can predict 100%. just like the fund houses. they try to average their bets

2015-10-27 15:43

CCCL

D7zul, earn your first bucket of gold. The rest will follow. Nothing is impossible.....

2015-10-27 20:09

WahLao

Firehark - I done my % data but cant edit it =.=.
YiStock - Good idea
Ooi Teik Bee - Thanks.
D7zul - Now I still learning fishing, not yet open account. My target is open account at KLSE <1400. I wish learn something from u.
JT Yeo - Yup, I agree with u. I backtesting FKLI and still dont get good formula for it.
kcchongnz - Thanks for sharing.
Probability - Thanks for your sharing, I plan talk about probability VS 1:2:7
Kevin Wong - I still waiting buying point =.=
faye Tan - =.=.... I not yet open a account.
CCCL - I agree.

2015-10-27 21:21

Ooi Teik Bee

Dear WahLao,

I suggest you do a study on "Buy when everyone is fearful" strategy.

In this case we buy on 24/8/2015, what will be the return of investment if we hold them until 31/5/2016 ?
Likewise, what is the return of investment if we bought on 28/8/2013 and hold until 31/5/2014.
What is the return of investment if we bought on 17/12/2014 and hold until 31/5/2015.

I always like to buy when everyone is fearful.
Thank you.

2015-10-28 08:38

speakup

Right now is "Sell when everyone is greedy". Look at SEB & ESCERAM yesterday, semua REBUUUUUTTTTTTTT!

2015-10-28 08:41

BuLLRam

Dzul..u need capital to make reasonable profits in stockmkt..hv to take a resdonable sized position to make effective profits..otherwise u are just makinf few hundreds...only lunch or beer money ...for all your time n effort...scale it up n realise your potential...cheers.

2015-10-28 10:10

WahLao

Dear Ooi Teik Bee,

Thanks for your suggestion.

1st I need convert "Buy when everyone is fearful" to number and logic/rule 1st.
Can advice what signal you c base on KLSE?

For example
"Buy when everyone is fearful"
= buy when got signal "Many People scare" X "throw they stock without thinking"
= "High Volume" X "Low KLSE" than we buy next day


Any1 if you have any ideas please comment. I hope can get the interesting results and I will post here.

2015-10-28 21:50

WahLao

D7zul - If we use daily chat, need wait how many day to confirm it?

2015-10-28 22:57

Ooi Teik Bee

Dear WahLao,

For example
"Buy when everyone is fearful"
= buy when got signal "Many People scare" X "throw they stock without thinking"
= "High Volume" X "Low KLSE" than we buy next day

Ans : Buy on the next day when "the first blue color candle appeared". Volume was high, it is even better. Low volume is also never mind.
Thank you.

2015-10-29 08:17

WahLao

Ooi Teik Bee - I try checked the results 1st

2015-10-30 00:56

D7zul

So, how's the result?

2015-11-21 14:11

mooi

Wake me up when September ends

2015-11-24 12:32

D7zul

mooi, september end already.. now is november.. hehe ;p

2015-11-25 12:41

gogogogo

another bad month is august

2017-04-17 13:49

Post a Comment