Saturday, January 30, 2016

2016-01-29 Trading Top5s Securities: AMZN

Scrape for 2016-01-29
2016-01-29
Mkt_Cap:MSFT,AAPL,V,BRK.A,AMZN|GILD,MFG,MTU,ABBV
Price:COHR,CIG.C,BBDO,SNE,CNX|MOG.B,MOG.A,ABAX,ALV,RARE
Volume:BAC,CFG,MSFT,SPY,AAPL,FB,PFE,QQQ,F,GE
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-29 data
Bubbles:

Let's analyze $AMZN, which jumped right out at us, topping market capitalization over two days:
geophf:writing geophf$ analyze AMZN
Wrote analysis files for AMZN
AMZN SMA



AMZN EMA

AMZN Stochastic Oscillators

Friday, January 29, 2016

2016-01-28 Trading Top5s Securities: FB

Scrape for 2016-01-28
2016-01-28
Mkt_Cap:FB,AMZN,GOOG,GOOGL,BABA|QCOM,ABT,LLY,NVS
Price:UA,MSTR,CRUS,ECA,OSIS|NTCT,FCFS,ADS,URI
Volume:NGD,LNG,TXMD,NG,CEF,GSAT,FAX,IMO,BTI
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-28 data
Bubbles:

Let's analyze $FB
geophf:writing geophf$ analyze FB
Wrote analysis files for FB
FB SMA

FB EMA


FB Stochastic Oscillators

Wednesday, January 27, 2016

2016-01-27 Trading Top5s Securities: TEX

Scrape for 2016-01-27
2016-01-27
Mkt_Cap:VZ,BIIB,JNJ,RDS.B,AAPL|NVS,GOOGL,GOOG,AMZN
Price:CVLT,TEX,HA,LCI,SC|DV,TSS,TUP
Volume:AAPL,BAC,FCX,FB,QQQ,T,XIV,SPY,GE,PBR
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-27 data
Bubbles

It's rare to see a security leading in price two days in a row, so let's analyze $TEX.
geophf:writing geophf$ analyze TEX
Wrote analysis files for TEX
TEX SMA



TEX EMA

TEX Stochastic Oscillators

2016-01-26 Trading Top5s Securities: XOM

Scrape for 2016-01-26
2016-01-26
Mkt_Cap:JNJ,XOM,RIO,CVX,LFC|LMCB,TSM,COST,REGN
Price:TEX,SWFT,S,FMER,CIE|LMCB,PII,LDOS,HBAN,ANAC
Volume:AAPL,ERIC,AUY,FCAU,SPY,USO,EEM,QQQ,XIV
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-26 data
Bubbles

Let's analyze $XOM
geophf:writing geophf$ analyze XOM
Wrote analysis files for XOM
XOM SMA



XOM EMA

XOM Stochastic Oscillators

Monday, January 25, 2016

2016-01-25 Trading Top5s Securities: ETE

Scrape for 2016-01-25
2016-01-25
Mkt_Cap:WMT,TYC,VRX,AAPL,XOM|GOOG,GOOGL,BRK.B
Price:TYC,GFI,CRAY,KS,DRII|NVAX,CHK,WRK
Volume:BAC,F,HBAN,CSCO,AAPL,FCX,GE,ETE,XIV,S
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-25 data
Bubbles:

ETE is strong both in volume today and yesterday, and price. Let's analyze $ETE
geophf:writing geophf$ analyze ETE
Wrote analysis files for ETE
ETE SMA



ETE EMA

ETE Stochastic Oscillators

Saturday, January 23, 2016

2016-01-22 Trading Top5s Security: GE

Scrape for 2016-01-22
2016-01-22
Mkt_Cap:AAPL,MSFT,GOOG,GOOGL,PTR|AXP,GE,BABA,LMCB,UNP
Price:GLNG,ETE,NGL,WMB,MITL|AXP,LMCB,CHK-D,FCX,NVCR
Volume:BAC,FCX,GE,AAPL,KMI,CSCO,ETE,SPY,F,AXP
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-22 data
Bubbles:

Let's analyze $GE
geophf:writing geophf$ analyze GE
Wrote analysis files for GE
GE SMA

GE EMA


GE Stochastic Oscillators

Friday, January 22, 2016

2016-01-21 Trading Top5s Securities: FCX

Scrape for 2016-01-21
2016-01-21
Mkt_Cap:VZ,GE,GOOG,GOOGL,ALKS|BAC,AAPL,MSFT,GILD
Price:CRZO,CNX,SWN,RRC,WLL|EGL,ALKS,RDUS,MTG,SCHL
Volume:TEF,USO,RSX,SPY,XIV,EWG,BP,FCX,SAP
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-21 data
Bubbles

Let's analyze FCX
geophf:writing geophf$ analyze FCX
Wrote analysis files for FCX
FCX SMA



FCX EMA

FCX Stochastic Oscillators

Wednesday, January 20, 2016

2016-01-20 Trading Top5s Securities: TWTR

Scrape for 2016-01-20
2016-01-20
Mkt_Cap:ABBV,CELG,AMGN,XOM,PTR|BRK.B,IBM,TM
Price:SYNA,MDCO,SWN,WGP,GLNG|NVCR,ENLC,NRF-A
Volume:BAC,FCX,GE,QQQ,XIV,CSCO,AAPL,TWTR,SPY,AA
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-20 data
Bubbles:

Let's analyze $TWTR, just out of curiosity:
geophf:writing geophf$ analyze TWTR
Wrote analysis files for TWTR
TWTR SMA

TWTR EMA


TWTR Stochastic Oscillators

2016-01-19 Trading Top5s Securities: BAC

Scrape for 2016-01-19
2016-01-19
Mkt_Cap:CVS,TI,ABEV,TAP.A,LMCB|STZ.B,MOG.B,MKC.V
Price:GRP.UN,GRT-H,HPT,MOG.B,LMCB|TAP.A,LTRPB,MKC.V
Volume:SPY,BAC,XIV,NFLX,AAPL,USO,NOK,GDX,QQQ
Updated /Users/geophf/Documents/OneDrive/work/1HaskellADay/Seer/data/top5s.csv with 2016-01-19 data
Bubbles

Let's analyze $BAC
geophf:writing geophf$ analyze BAC
Wrote analysis files for BAC
BAC SMA

BAC EMA


BAC Stochastic Oscillators

Tuesday, January 19, 2016

2016-01-18 Martin Luther King, Jr. Holiday

Today is the Martin Luther King, Jr. Holiday. Markets are closed today. You can view the trade calendar here.

Sunday, January 17, 2016

Case study: AAPL investment strategies, 10-year period

Thesis

There is no doubt AAPL has done well the last ten years. A simple buy-and-hold strategy has a ten-bagger yield. That is success, and you can't argue with success.

And we don't. Given the past ten years, a ten-fold gain is a success in anybody's book.

But can we do even better than that with a different strategy.

It is put forward that AAPL's share price becomes chaotic around their Expos and special events. Is this so? And, if so, can this observation be put to good use?

This is the thrust of this paper.

Contents

In this paper we show

1. the results of following a simple buy-and-hold approach, a.k.a. the baseline
2. the share prices around each of the last ten years' expos, a.k.a. the price volatility at those times.
3. the annual sell-then-buy strategy as applied at each expo, a.k.a. the alternative
4. then we compare results of the strategies

Then we close with a summary and further thoughts/future work.

Antithesis

1. Baseline

So, what do we get when we simply buy-and-hold. Simple, right? Just buy ten years ago, then read out the price today, right?

Nope.

Splits.

You must factor in the splits experienced over the last ten years. You can find this information here: http://investor.apple.com/dividends.cfm

(Cold Fusion. And on an Apple web-site. How ... quaint.)

Okay. Let's do this.

The supposition is that if you invested in AAPL ten years ago, you'd have a million dollars today. Okay, true? And how much would we have to make that happen? Let's run the numbers.



So, as you see above, we've actually asked the dual question: if we had invested $10k into AAPL 11 years ago, how much money would result, and the answer is: $250k+.

Instead of a 10-bagger, because of the two-for-one-split at the beginning of 2005, we actually have a 25-bagger.

Now, that is nothing to scoff at.

2.a. Can we do better?

That is the question, along with: "To be, or not to be." But I'll just go with "to be" for now, so we can focus on investment strategies. Looking at the above chart over the 11 years, it seems to be a smooth progression, and it is, year-over-year, but around the expos themselves we sometimes have more chaotic behavior. Let's look at each year's expo.

2016-01-06

The last ten years of expos occurred on the following dates:

2005-01-10
2006-01-09
2007-01-08
2008-01-14
2009-01-05
2010-01-11
2011-02-25
2012-01-26
2013-01-31
2014-01-27
2015-03-12

Okay, borin' list. Let's see the share-price activity around each expo:

2005-01-10



2006-01-09


2007-01-08


2008-01-14


2009-01-05


2010-01-11


2011-02-25


2012-01-26


2013-01-31


2014-01-27


2015-03-12


As you see from the above close examinations of the price histories around the prior Mac World Expos, there is quite a bit of price volatility. Is it something we can exploit?

Nah.

Okay... jk. Chillax.

2.b. Devising a strategy

geophf stands up on a little soap crate, clears his throat, and proclaims: "2.b. or not 2.b.! That is the ..."

A big hook drags him off the stage.

So, the idea is that we can beat the baseline, so, eh, sell just before the expo and buy just after, it looks like, and we make a killing, isn't that right? Let's test this by selling five days before the expo and buying five days after:

*Main> let (ans, log) =
           runWriter (stratSell (USD 10000) "AAPL" cal (stratForAAPL expos 5))

*Analytics.Trading.Data.Strategy> mapM_ print (dlToList log)
PlaceOrder 2005-01-15 "AAPL" Buy $9990.00 $4.70 $10.00
PlaceOrder 2006-01-04 "AAPL" Sell $21013.11 $9.89 $10.00
PlaceOrder 2006-01-14 "AAPL" Buy $21003.11 $11.27 $10.00
PlaceOrder 2007-01-03 "AAPL" Sell $21228.55 $11.39 $10.00
PlaceOrder 2007-01-13 "AAPL" Buy $21218.55 $12.92 $10.00
PlaceOrder 2008-01-09 "AAPL" Sell $38884.56 $23.68 $10.00
PlaceOrder 2008-01-19 "AAPL" Buy $38874.56 $20.70 $10.00
PlaceOrder 2008-12-31 "AAPL" Sell $22660.75 $12.07 $10.00
PlaceOrder 2009-01-10 "AAPL" Buy $22650.75 $11.79 $10.00
PlaceOrder 2010-01-06 "AAPL" Sell $53775.53 $28.01 $10.00
PlaceOrder 2010-01-16 "AAPL" Buy $53765.53 $28.60 $10.00
PlaceOrder 2011-02-20 "AAPL" Sell $84645.80 $45.04 $10.00
PlaceOrder 2011-03-02 "AAPL" Buy $84635.80 $47.83 $10.00
PlaceOrder 2012-01-21 "AAPL" Sell $100595.15 $56.85 $10.00
PlaceOrder 2012-01-31 "AAPL" Buy $100585.15 $60.68 $10.00
PlaceOrder 2013-01-26 "AAPL" Sell $100053.68 $60.37 $10.00
PlaceOrder 2013-02-05 "AAPL" Buy $100043.68 $61.38 $10.00
PlaceOrder 2014-01-22 "AAPL" Sell $124706.64 $76.51 $10.00
PlaceOrder 2014-02-01 "AAPL" Buy $124696.64 $68.99 $10.00
PlaceOrder 2015-03-07 "AAPL" Sell $226808.92 $125.50 $10.00
PlaceOrder 2015-03-17 "AAPL" Buy $226798.92 $126.82 $10.00

*Main> ans ~> 1924.12%

Huh, the sell-before-and-buy-after strategy does spectacularly well! ... as the chart of this approach shows:


Except for the fact that we made ~$30,000 less than the simple buy-and-hold strategy, and here, also, we cheat, because each sale incurs a tax on the profits! So not only do we do worse just running the numbers, but we do worse-worser-worsest! after paying additional taxes each year.

How much additional tax? I leave that as an exercise to the reader.

2.c. 'just before' and 'just after'

The problem here is the 'just' in 'just before' and 'just after' is a matter of more than just trying five days and hoping for the best. This needs a little machine-learning-love to get the absolute best result, and then it requires kicking the machine-learning algorithms to the curb when the absolute best results only work in backtesting.

So, five days works great! except just not as well as the buy-and-hold strategy.

How about a 10-day sell-before-and-buy-after strategy?

*Analytics.Trading.Data.Strategy> let (ans, log) =
          runWriter (stratSell (USD 10000) "AAPL" cal (stratForAAPL expos 10))

*Analytics.Trading.Data.Strategy> mapM_ print (dlToList log)
PlaceOrder 2005-01-20 "AAPL" Buy $9990.00 $4.69 $10.00
PlaceOrder 2005-12-30 "AAPL" Sell $21165.54 $9.94 $10.00
PlaceOrder 2006-01-19 "AAPL" Buy $21155.54 $10.12 $10.00
PlaceOrder 2006-12-29 "AAPL" Sell $23288.12 $11.15 $10.00
PlaceOrder 2007-01-18 "AAPL" Buy $23278.12 $11.77 $10.00
PlaceOrder 2008-01-04 "AAPL" Sell $46704.63 $23.63 $10.00
PlaceOrder 2008-01-24 "AAPL" Buy $46694.63 $17.29 $10.00
PlaceOrder 2008-12-26 "AAPL" Sell $31101.51 $11.52 $10.00
PlaceOrder 2009-01-15 "AAPL" Buy $31091.51 $10.95 $10.00
PlaceOrder 2010-01-01 "AAPL" Sell $80787.11 $28.47 $10.00
PlaceOrder 2010-01-21 "AAPL" Buy $80777.11 $26.30 $10.00
PlaceOrder 2011-02-15 "AAPL" Sell $148308.86 $48.30 $10.00
PlaceOrder 2011-03-07 "AAPL" Buy $148298.86 $47.32 $10.00
PlaceOrder 2012-01-16 "AAPL" Sell $177023.53 $56.49 $10.00
PlaceOrder 2012-02-05 "AAPL" Buy $177013.53 $61.72 $10.00
PlaceOrder 2013-01-21 "AAPL" Sell $194278.33 $67.74 $10.00
PlaceOrder 2013-02-10 "AAPL" Buy $194268.33 $64.78 $10.00
PlaceOrder 2014-01-17 "AAPL" Sell $226505.76 $75.53 $10.00
PlaceOrder 2014-02-06 "AAPL" Buy $226495.76 $71.92 $10.00
PlaceOrder 2015-03-02 "AAPL" Sell $402125.54 $127.69 $10.00
PlaceOrder 2015-03-22 "AAPL" Buy $402115.54 $125.57 $10.00

*Analytics.Trading.Data.Strategy> ans ~> 3524.33%

Ook! Oh, yeah! Let's chart that puppy!



Hm ... this looks suspiciously similar to the previous chart. Let's leave that aside for the moment.

So, okay, can we do even better? like two million percent? Let's see!

*Analytics.Trading.Data.Strategy> let (ans, log) =
          runWriter (stratSell (USD 10000) "AAPL" cal (stratForAAPL expos 30))

*Analytics.Trading.Data.Strategy> mapM_ print (dlToList log)
PlaceOrder 2004-12-11 "AAPL" Buy $9990.0 $5.21 $10.0
PlaceOrder 2005-12-10 "AAPL" Sell $19081.59 $9.96 $10.0
PlaceOrder 2005-12-10 "AAPL" Buy $19071.59 $8.64 $10.0
PlaceOrder 2006-12-09 "AAPL" Sell $26046.6 $11.81 $10.0
PlaceOrder 2006-12-09 "AAPL" Buy $26036.6 $11.46 $10.0
PlaceOrder 2007-12-15 "AAPL" Sell $55686.63 $24.53 $10.0
PlaceOrder 2007-12-15 "AAPL" Buy $55676.63 $16.95 $10.0
PlaceOrder 2008-12-06 "AAPL" Sell $43552.92 $13.26 $10.0
PlaceOrder 2008-12-06 "AAPL" Buy $43542.92 $12.83 $10.0
PlaceOrder 2009-12-12 "AAPL" Sell $88890.88 $26.20 $10.0
PlaceOrder 2009-12-12 "AAPL" Buy $88880.88 $26.43 $10.0
PlaceOrder 2011-01-26 "AAPL" Sell $153522.89 $45.65 $10.0
PlaceOrder 2011-01-26 "AAPL" Buy $153512.89 $46.61 $10.0
PlaceOrder 2011-12-27 "AAPL" Sell $176367.3 $53.56 $10.0
PlaceOrder 2011-12-27 "AAPL" Buy $176357.3 $69.93 $10.0
PlaceOrder 2013-01-01 "AAPL" Sell $185787.32 $73.68 $10.0
PlaceOrder 2013-01-01 "AAPL" Buy $185777.32 $56.70 $10.0
PlaceOrder 2013-12-28 "AAPL" Sell $249938.47 $76.28 $10.0
PlaceOrder 2013-12-28 "AAPL" Buy $249928.47 $73.3 $10.0
PlaceOrder 2015-02-10 "AAPL" Sell $421876.49 $123.27 $10.0
PlaceOrder 2015-02-10 "AAPL" Buy $421866.49 $125.22 $10.0

*Analytics.Trading.Data.Strategy> ans ~> 3713.14%


Still good. No 'two million percent,' but (slightly) better than the ten-day sell/buy-approach. So, apparently a five-day before-after duration in the strategy was too close to the 'heat' of the expo. Good to know!

Similarities

Looking at the charts in isolation doesn't give us a good picture of how significantly different the impacts of following these three approaches are. So, let's look at the three approaches on one chart and get a clearer picture:


Here we see that even though all of the approaches are of the same form (which is not surprising, after all: they are all of the same strategy: sell and buy around each Mac World Expo), but what is striking here is the impact of distancing these orders from the epicenters of the Expos: the further one is from the events themselves, the more significant the orders become. Particularly, placing orders within the week of the Expo gets lost in the chaos of the Expo, itself; ten days away from the expo and the stock becomes reliably stable enough so that profitability is nearly double that of the simple buy-and-hold strategy.

Synthesis

The last ten years have shown that simply buying and holding AAPL yields incredibly good returns of over 2500% on the initial investment. We could stop there and call it a day and be very happy with our result vis-à-vis most of the rest of the markets.

But, a simple observation yields an insight into this security: there is high volatility of the share-price around the Mac World Expo. What does this insight mean? Simply selling just before each expo and buying right after each Expo yields a (very) good set of returns, but, unfortunately, worse than the simple buy-and-sell strategy.

No. It was the insight then backed by some straightforward data analytics that yielded a payout over 3700%, leading to the formula:

Insight + Analytics = Winning

Apotheosis

We reviewed two strategies in this paper: the buy-and-hold strategy and the sell-then-buy around the Mac World Expos strategy. This is the start of the inquiry of this high-performing security. It can be the end, as well: 3700+% return over ten years is nothing to sneeze at!

Fine, but can we do even better than that? Perhaps. We looked at buy-and-hold, then we examined sell-then-buy around the time of the Mac World Expos. There are other things we can do:

  • The Mac World Expos were not the only major events Apple has sponsors, sometimes even two or three other major events or new product reveals have occurred during the year. What happens when we include those events in the sell-then-buy strategy?
  • Buy-and-hold and sell-then-buy both deal with one initial investment over a 10+-year period. What if we take the DRiP approach of small, regular purchases over time, either adopting the buy-and-hold or the sell-then-buy approach to regular (monthly, quarterly or yearly) investments?
  • Options. One approach is to used options to leverage the buy or sell orders.

These are three possible areas that are good sources of research.

Sources:


  • Since May, 2015 to January, 2016 AAPL has been mentioned on top 5s trading securities list more than double than any other security making that list during that period. Leading to the initial buy-and-hold case-study: http://lpaste.net/5472780506909114368
  • @MacNN_Mike countered with the sell/buy-approach around the Mac World Expos on twitter https://twitter.com/MacNN_Mike/status/674279321612218368
  • This approach with the 5, 10, and 30 day was subject to analysis (eventually) at http://lpaste.net/2656933684896071680 (this is a revision to incorporate the stock splits that occurred)
  • The three approaches were then compared and contrasted as one Big Up! at http://lpaste.net/2722111763528024064
Caveat

These approaches fall into the domain of back-testing. Investing in AAPL has worked, and has worked very well because AAPL has captured that ineffable quality of innovation and cool, giving them preeminence in the marketplace. 

The caution is this: as soon as AAPL stops being innovative or cool, then that fickle thing – consumer loyalty – just goes away. Their past performance has been stellar, but their future performance is not guaranteed, and this article makes no such promise on future earnings.