a small note on the previous companies - the fact that they have not increased as expected is I believe partly due to the current uncertain situation in Europe, what I can say is that the companies below have not decreased to any level related to other companies, and that charts on all of the companies continue to look good.
I have added analysis of the greater Global Indexes and will be making comparisons between these and my alphas for a better understanding of Global positions - another late night coming up.
Thursday 29 September 2011
Wednesday 28 September 2011
MWA NWS
beta5 both
MWA currently 4.5 looking for 5.01
NWS currently 85 looking for 94.7
charts to the right ->
Do Your Own Research and Due Diligence
BOY, SMDR, WSH
the interesting aspect of these three likely lads is that whilst pretty much everything else is down, these three are up between 3.5 and 6.9%
bodes well for beta4
bodes well for beta4
Tuesday 20 September 2011
Monday 19 September 2011
WSH
so where does all this analysis, charts, blogging get us ?
I get a list of companies each day from my betas now that on the face of it look promising for an 11.5% rise over about 3 to 5 days.
with the usual 'DO YOUR OWN RESEARCH' caveat and I take no responsibility on your trading actions, I dont know your trading schemes or limits or anything else, I will now publish the companies coming up in the best betas each day.
this is beta4, a complex analysis of the rsi, vi, psars, macd and adx, and then another analysis on the psar of the rsi.
so for today its WSH, charts to the right, currently at 249, therefore looking for 278.
Saturday 17 September 2011
100%ers and betas
the only problem of finding a 100%er is that the companies occur, in the instance of beta1, every 6 months, only need another 36 of these and I am away :)
I have so many alpha's now that it is difficult keeping track as to which have completed development - alphas are now turning into betas when complete
DPL and GEM, yey
Friday 9 September 2011
balancing an elephant on the edge of a razor blade
alpha16 is a strange one - it seems to highlight specific months and especially Aug 2011 and right now, and what is more important is that the trend post buy signal is pretty much always in the up direction with a clear indication on the psar of the rsi of the close.
the reason that this seems to be period dependent could be that this alpha may be following a greater trend.
alpha16 is approximately:
report all companies that are for volume shares, the stochastic on vi is 100 and the stochastic on the macd is 0, the psar trigger on the rsi on the close must be d to u, the stochastic of the psar of close must be 0, there must have been a stochastic equal to 0 over the last 7 days and the close rsi must be less than 20 over the last 7 days.
the relationship between the days, the indicators and the analysis is vital.
the admin page and an example company CTY is included to the right
Wednesday 7 September 2011
psar on rsi
just for fun ran the psar against the rsi and it looks like a very good indicator - developing the next alpha on this just now - its the last chart in the image to the right, and this is for MXP
Tuesday 6 September 2011
rsi
from a kind LSE forum txter it seems that the rsi is an indicator often used, a little digging and obtaining the logic, and a late night and my system is updated with this as well as incorporated into my newest alpha
still going through the results but all the companies checked until now follow an upward trend post signal for this new alpha
to the right is a random company XCH with the RSI the last chart superimposed onto the close price (black) for reference
there is a slight dip post buy signal at the moment on this company that needs ironing out but it is still in development
this alpha did though signal a buy at the start of an almost 100% increase in the stock price so things are looking up
Monday 5 September 2011
EOG sample
EOG is again quite interesting at the moment and am awaiting to see how things pan out - the psars, macd, adx and vi look good yet the st looks very high
Sunday 4 September 2011
more charts
the stochastic on the psar of the psar on close has proved to be quite interesting and is detailed for MXP on the eighth chart in the diagram to the right - very cyclic and looking to be following rise and falls quite closely - while we are at it lets add the stocastics for the psars on the vi and the macd as well, just for fun you understand
will add to the next alpha to see how this goes - the calcs are running for all companies at the moment
just a note here - its not important to know the derivation of the charts, or what a stochastic means or the other terms, apart from that it goes from 0 to 100, the important thing here is to establish a link between the pattern on each chart and the close price in black
on each chart if we can say something like 'buy when red > 80 and sell when < 10' with some degree of certainty then we stand a chance
analysis 1
where to start ?
I start with a list of all companies that have risen more than 45% over a period of 10 working days (2 weeks) since the 1st Jan 2011
this is the first listing to the right - the list is cropped because it is quite large which is really great news as it means I have a good chance of picking one of these - there are about 7 per day - and many more if I reduce the % increase, 45 sounds like a good number at the moment
I am also looking just at the close prices, the thinking is that these are a little more stable than looking at the close at the start and then highs, but would presume this would give me more results, and there are quite enough to go through at the moment
then I choose a company at random... lets look at HUM - the second chart list to the right and appears in the list under 11 Aug 2011
this is really where the science takes a back seat and the art begins
a cursory scan of the charts and we see that in the sixth chart the vi psar goes from negative (below the vi line) to positive (above the vi line) at exactly the date indicated, the vertical black line
if we only looked for this rule then we would get a lot of companies being returned and with an unacceptable future rise/decrease ratio, we need to add other indicator formulae
another quick glance and we see the stochastic k has a minimum of 0 over the last 10 days prior to the vertical black line, something that I have found happens very often on risers
I create a new alpha and put these analysis in...
and the analysis continues, taking into consideration anything that you can think of that might possibly bring the listing to risers only without over restricting so that no companies are returned, ones that I am looking into at the moment are gradients and differences between different psars, ahh... and then there is the riemann hypothesis , my search for pattern recognition and indicator analysis has on more than one occasion lead back to this, more on this at some point
writing this and looking at the charts makes it look interesting to calculate the stochastic on the psar of the psar on close
Saturday 3 September 2011
KEFI sample
and this is what I have on KEFIthe interesting thing here is that a lot could be made almost by just following the stochastics - it is very cyclic - buy on k < 5 sell on k > 80
chart definitions are on the MXP post
not all a bed of roses
just to prove that it is not all good, this is EOG from alpha11 with buy signal 16th march 2011, not good, not good at all
but it is not all bad, the good news is that out of 9 random companies researched only one was like this
alpha11 is still in development
chart definitions are on the MXP post
MXP - 3rd Sept 2011
and this is for my MXP brethren:
the first chart is close price in black, blue psar on close, red psar on open
the second chart is close black, close psar in red, and then psar on close psar in blue (yes psar on the psar)
the third chart is close black then a mash of psar on low, psar on high, psar on high up low down, psar on open up close down
the forth chart is adx with + in blue and - in red and adx in green, with the close manipulated to fit between 0 and 100
the fifth is the macd with the macd in red and the macd psar in green, zero is blue, close is black and manipulated to fit between min and max
the sixth chart is vi, vi in red, psar on vi in green, zero in blue, close is black and manipulated to fit between min and max
the seventh chart is the stochastic, red is k, blue is smoothed k (called d), green is smoothed smoothed k (called d1), black is close manipulated to fit between 0 and 100
the eighth, ninth and tenth charts on the more recent charts are the stochastics on vi, macd and psar(close),
and the eleventh chart is rsi and psar on rsi, again seen on more recent charts
ZOX sample
this is a randomly selected company from the top of the list - 31 Aug 2011 - this would be a good experiment to see if this share does reach 11.5% over the next 5-10 dayschart definitions are on the MXP post
TIG sample
from the alpha11 sample previously this is what I get from a randomly selected company from the bottom of the list, about march (in this case march 1st indicated by the vertical line):
just to confirm, I know nothing about TIG, all I do know is that with the analysis for alpha11 said to buy on march 1st
chart definitions are on the MXP post
in this case the 11.5% sell would have been met on about 25 may, if we bought at about 16 and sold at about 18, but I would miss the close price going to 23
system sample
this is what the alpha development system looks like, run for alpha 11 on Sept 3 2011:the first 2 input fields are the name and reporting start date
the next section is the sql statement to be run against the database and is sectioned into separate lines for ease of reading and use - the where clause is the one that takes the most development.
it is the where clause that governs the companies that are returned.
this whole statement is linked to the field marked sql: just below the word success, this field would contain the whole statement run against the database as the database would see it, this aids debug - the success word is there to indicate that the report completed successfully
the reset function resets all the fields to default values
and then we get to the output data
the first column is the company with a link to the LSE page for that company (London South East)
the second column is the date of the buy signal with a direct link to the charts for the company on the LSE site
the third column is the date reference
the forth column (A) links directly to the underlying data in the system for the company
the fifth column (G) is the system graphs, examples posted later
the sixth column is the close price for the particular day
the colons are place holders for other information that can be added if needed
day 3 - nothing for free
all of this has now been running for one year.
has this worked ? predominantly yes
getting up signals is not too much of a problem, with the best alpha's this is over 80%, one alpha approaching 100% (alpha7) but only produces a stock pick on average about once per month, which is great if you want 11.5% per month, I need 20 alpha7's.
the sell signal is a little more problematic
currently I am working on a flat 11.5% up from buy price to sell, which I am trying to reduce to between 3 and 5 days giving me the required about 3% per day.
my trading strategy needs some thought now
has this worked ? predominantly yes
getting up signals is not too much of a problem, with the best alpha's this is over 80%, one alpha approaching 100% (alpha7) but only produces a stock pick on average about once per month, which is great if you want 11.5% per month, I need 20 alpha7's.
the sell signal is a little more problematic
currently I am working on a flat 11.5% up from buy price to sell, which I am trying to reduce to between 3 and 5 days giving me the required about 3% per day.
my trading strategy needs some thought now
day 2 - next 6 months
so here we are with a system that downloads eod data each evening, uploads to a database and then calculates some indicators
psars are calculated on close, open, high, low and vol (although psar on vol is a little strange to say the least because the numbers involved are so large), as well as psars of open up and close down and high up and low down (psars trip at the top and bottom of trends, so these switch the up and down basis)
to make sure we have a trend we need the other indicators:
adx, vi and st are calculated as per wikipedia.
psars on vi and macd are also calculated just for fun and have proven very useful, as is the psar on the close psar
I enabled the embedded web server in the database and output reports to a web page, my ip is static and the database allows the creation of http code - I have programmed an http5 canvas reporting system. OEL has apache 2.2, may use this if I split the server into web and dbase in the future, but simple is the name at the moment
the interesting analysis happens when I link the indicators... as these are the start of the stock picking process I have called these alpha's.
alpha's consist of statements such as
show me all companies which have an average volume greater than 10 x 10,000 / close price (indicating that buying and selling of the shares will not be a problem), that over the last 10 days have had a minimum st of 0, +adx is greater than -adx, psar on vi is positive but was negative over the last 3 days, psar on macd is positive but was negative over the 3 previous days and the psar on close is positive with the previous 3 days being negative, and give me all companies that follow this rule from 1st Jan 2010 to the present day
this is approximately alpha 12, an alpha that works very well
the coding of alpha's interface is also via a web page, so the development cycle is I pick which alpha I want on a web page, edit, save, run, check resulting company charts, all via the web.
for live alpha, I merely run and due diligence the companies coming out with the date of today.
I run this statement from a specific date and can gauge effectiveness of alpha's, at the same time have programmed a routine that automatically goes through the alpha's and indicates which are good and which are not, this takes a lot of processing. current analysis is choosing companies at random and checking out movement, if price goes up after indicated date, then choose another company until I find one that does not follow the up rule and start the process of finding out what indicator runs against the other companies. I do this iteratively, each time changing the code and then running and checking again. each time the number of returned companies that follow the rule reduces - the art is to not restrict the companies to an extent that no companies are returned, but return all companies that increase in value over the proceeding days.
the production of a fully tested alpha takes weeks
the result is that I have built up and am building a set of alpha's that under specific circumstances indicate a good probability that the company share price will go up...
psars are calculated on close, open, high, low and vol (although psar on vol is a little strange to say the least because the numbers involved are so large), as well as psars of open up and close down and high up and low down (psars trip at the top and bottom of trends, so these switch the up and down basis)
to make sure we have a trend we need the other indicators:
adx, vi and st are calculated as per wikipedia.
psars on vi and macd are also calculated just for fun and have proven very useful, as is the psar on the close psar
I enabled the embedded web server in the database and output reports to a web page, my ip is static and the database allows the creation of http code - I have programmed an http5 canvas reporting system. OEL has apache 2.2, may use this if I split the server into web and dbase in the future, but simple is the name at the moment
the interesting analysis happens when I link the indicators... as these are the start of the stock picking process I have called these alpha's.
alpha's consist of statements such as
show me all companies which have an average volume greater than 10 x 10,000 / close price (indicating that buying and selling of the shares will not be a problem), that over the last 10 days have had a minimum st of 0, +adx is greater than -adx, psar on vi is positive but was negative over the last 3 days, psar on macd is positive but was negative over the 3 previous days and the psar on close is positive with the previous 3 days being negative, and give me all companies that follow this rule from 1st Jan 2010 to the present day
this is approximately alpha 12, an alpha that works very well
the coding of alpha's interface is also via a web page, so the development cycle is I pick which alpha I want on a web page, edit, save, run, check resulting company charts, all via the web.
for live alpha, I merely run and due diligence the companies coming out with the date of today.
I run this statement from a specific date and can gauge effectiveness of alpha's, at the same time have programmed a routine that automatically goes through the alpha's and indicates which are good and which are not, this takes a lot of processing. current analysis is choosing companies at random and checking out movement, if price goes up after indicated date, then choose another company until I find one that does not follow the up rule and start the process of finding out what indicator runs against the other companies. I do this iteratively, each time changing the code and then running and checking again. each time the number of returned companies that follow the rule reduces - the art is to not restrict the companies to an extent that no companies are returned, but return all companies that increase in value over the proceeding days.
the production of a fully tested alpha takes weeks
the result is that I have built up and am building a set of alpha's that under specific circumstances indicate a good probability that the company share price will go up...
day 1 - first 3 months
with an interest in number manipulation, databases, finance and beating systems, I have set about doing all 4 - this is the story of my journey
trading, that sounds interesting. a little digging and the next thing I know I am developing a system and buying a server.
so...
lets have some fun with the stock market, all indicators point to this being the biggest game in town
it seems that this game is where you pick a company, buy some stock in said company and watch while either the value of your stock increases or doesn't - the first stage of this process is the choosing of the company stock, and where I start
anyone who has tried to stock pick knows it is not easy, random picking works randomly, which means that there has to be a system.
the premise is that money can be made from stocks and that the system is geared in some way making this possible - there is no point in carrying on if this is not true -
if the premise is true then we can get data, manipulate it, report on it to pick stocks with certain indicators/patterns and trade
dont want to get bogged down with the due diligence although some has to be done - this is not the point and as a black art as stock picking itself and highly dependent on market sentiment - I have heard of systems tracking posts on forums, texts and news reports looking for company names and then calculating the overall tone of the post to indicate sentiment - this is not what this is about but I am not discounting it in the future.
its a game, ok with real money but still a game, a serious game.
and so to set up...
what is available data wise ?
seems that easy to get data is stock close, open, high, low and vol at the end of a day
(I am looking at stocks only at the moment, options and derivatives will come later as will shorting - at the moment keeping it simple and going long on (buying) equity only)
there is other information available but I have to start somewhere.
what indicators are out there - indicators that tell us how a stock is doing ?
I have chosen:
psar, vi, adx, st, macd and rsi - Wikipedia, yey - all will be explained - there are others, I have chosen these for the moment
psar is interesting as this is the one indicator that tries to look into the future whilst the others are looking at the past. remember the name... parabolic sar
how about another goal ? 1 million, 1 year, from 10,000 which equates to approx 2% per day compound - surely this is possible ? on 2% this gets interesting at about day 90 when funds reach about 50,000
my system... I like oracle. I like unix and windows. unix for the server (oracle enterprise linux), windows 7 and ubuntu for the laptop for testing
the server is currently a 12 core Xeon with 24Gb ram and ssds, lots of fans and lots of noise, and a ups;
the laptop is an i7, dual boot win7 and ubuntu and gets very hot when in full calc mode
downloaded data manipulated with sed and awk, scheduled with crontab, and uploaded to the database, validated and then the calcs start, all in the evening over night...
buying stock... a First Direct sharedealing account seems a fair place to start, internet based - there is stamp duty and trading costs, must remember to add these to the 2% per day.
being uk based I look at LSE and see where data is out there - first land on google finance.
create script (wget) to download all data via http from all companies each evening... I guess google does not like to be whacked and found data missing and a rather nice error message saying that google thinks that the ip address is up to no good and has decided not to send, reeks havoc with my calcs. the exact same thing then happens with yahoo finance
need a reliable source of end of day (eod) data - a quick search and 25 a month later I have a regular feed of eod data, and ftp access, yey.
so the scene is set...
trading, that sounds interesting. a little digging and the next thing I know I am developing a system and buying a server.
so...
lets have some fun with the stock market, all indicators point to this being the biggest game in town
it seems that this game is where you pick a company, buy some stock in said company and watch while either the value of your stock increases or doesn't - the first stage of this process is the choosing of the company stock, and where I start
anyone who has tried to stock pick knows it is not easy, random picking works randomly, which means that there has to be a system.
the premise is that money can be made from stocks and that the system is geared in some way making this possible - there is no point in carrying on if this is not true -
if the premise is true then we can get data, manipulate it, report on it to pick stocks with certain indicators/patterns and trade
dont want to get bogged down with the due diligence although some has to be done - this is not the point and as a black art as stock picking itself and highly dependent on market sentiment - I have heard of systems tracking posts on forums, texts and news reports looking for company names and then calculating the overall tone of the post to indicate sentiment - this is not what this is about but I am not discounting it in the future.
its a game, ok with real money but still a game, a serious game.
and so to set up...
what is available data wise ?
seems that easy to get data is stock close, open, high, low and vol at the end of a day
(I am looking at stocks only at the moment, options and derivatives will come later as will shorting - at the moment keeping it simple and going long on (buying) equity only)
there is other information available but I have to start somewhere.
what indicators are out there - indicators that tell us how a stock is doing ?
I have chosen:
psar, vi, adx, st, macd and rsi - Wikipedia, yey - all will be explained - there are others, I have chosen these for the moment
psar is interesting as this is the one indicator that tries to look into the future whilst the others are looking at the past. remember the name... parabolic sar
how about another goal ? 1 million, 1 year, from 10,000 which equates to approx 2% per day compound - surely this is possible ? on 2% this gets interesting at about day 90 when funds reach about 50,000
my system... I like oracle. I like unix and windows. unix for the server (oracle enterprise linux), windows 7 and ubuntu for the laptop for testing
the server is currently a 12 core Xeon with 24Gb ram and ssds, lots of fans and lots of noise, and a ups;
the laptop is an i7, dual boot win7 and ubuntu and gets very hot when in full calc mode
downloaded data manipulated with sed and awk, scheduled with crontab, and uploaded to the database, validated and then the calcs start, all in the evening over night...
buying stock... a First Direct sharedealing account seems a fair place to start, internet based - there is stamp duty and trading costs, must remember to add these to the 2% per day.
being uk based I look at LSE and see where data is out there - first land on google finance.
create script (wget) to download all data via http from all companies each evening... I guess google does not like to be whacked and found data missing and a rather nice error message saying that google thinks that the ip address is up to no good and has decided not to send, reeks havoc with my calcs. the exact same thing then happens with yahoo finance
need a reliable source of end of day (eod) data - a quick search and 25 a month later I have a regular feed of eod data, and ftp access, yey.
so the scene is set...
Subscribe to:
Posts (Atom)