Making a 2+2 trading system
 

NOTE: ahnuld: If you think I should just get a blog, then lock this and I’ll move on. Also, the main piece of advice I’ll give is for people not to follow my advice, but just consider it with reference to your own situation and try and test anything of interest yourself…... Then you can only blame yourself when you bust your wad.

I’d like to develop and implement a new trading strategy on this forum. There are regulars here who are much more qualified than me to do this, but I thought the process might be useful for others to watch and help me.. Besides, it’s something that must be done by me eventually anyway. The system will be mechanical in nature and probably non-discretionary, so I won’t be considering fundamental analysis (I suspect 90% of you will stop reading here). You can call it TA, but I won’t be using any classical chart-reading or Japanese candlestick techniques here as I’m still practicing those disciplines. Most of this will be relevant to anyone trading anything; whether by FA, TA, tea-leaves, solar flares, lunar cycles, or the Stu Ungar Method. I will be using some very familiar indicators and hope to look for an edge primarily from my exits and bet-sizing, and using the indicators to help maintain discipline and quantify prices beyond the noise. For the record, I believe using indicators is inferior to price alone since indicators are a derivative from price. Some are a derivative of a derivative (eg- MACD, TMA) and the further you depart from the price, the further you depart from the reality of the market.

I currently trade 2 systems. One is long-term (months/years) and the other is medium term (weeks/months). Both have had mixed but reasonable results and are by no means magic bullets, but they have met my (modest) objectives. I have a little more time on my hands, so I thought I’d try to make a shorter term strategy (days/weeks). Provided I trade each system on different markets I should be able to improve my overall risk/reward as the above 2 systems are quite different in nature, and hopefully this 3rd one will be to. Posting live trades for these 2 systems would be like watching paint dry for most people, but maybe daily positions (or at least market orders) might spark an interest amongst you gambooolers. I have neither the time, ability nor inclination to daytrade. I would probably have a better edge at an Absolute Poker blackjack table than being sat in front of screens full of 100s of blinking red and green arrows (aka-the world’s biggest slot machine).

I suggest the trading plan be written something like this: (and each section can be discussed/flamed in a separate sub-thread)

1. Introduction – finding a concept or an exploitable edge.
2. Objectives – I believe developing a trading system is like making your own clothes. I may feel good and look fantastic in purple and green jeans, but they may not be suitable for you.
3. Timeframe – tick by tick, hour by hour daytrading, EOD, weekly charts?
4. market(s) and trading vehicle - Commodities, forex,, stocks? Also, am I going to trade equities, futures, spread-trading, fruit & vege, CFDs,?
5. Trade set-up – What conditions need to be present before I place an order? This might involve various trade filters.
6. Entry – How, where, and why am I going to enter the market?
7. Exit – How, where, and why am I going to exit the market?
8. Stops – buy limit, stop-loss, GTC, GTS, OCO etc. How am I going to use stops to enter/exit the trade and protect capital/take profits?
9. Money management – Know how much to bet, pyramiding, ‘bankroll’ management. Experiment (play) with the Kelly Criterion. This section will probably overlap #7
10. Understanding my system’s performance – This will include back-testing, forward testing and paper trading and is very time consuming and difficult (for me). I need to know how my system will behave under various conditions, what the expected profits/losses are (EV), risk/reward, std.dev, sharpe, opportunity factor, blah blah. It is very important to be able to psychologically whether drawdowns knowing you still have an edge. (aka, my current 500BB downswing – FU MINBET)
11. Further concepts such as contingency planning, record keeping/reviewing.

All the above things need to be defined and considered before any trade is taken.

I will try and address some of the concepts discussed, but hopefully just keep to the bare details. I’m sure others here could do a better job than me explaining the intricacies and reasoning behind the various aspects of system development.

I have lots of ideas for systems and get more every time I stand and think in the shower. I’ll start by suggesting a basic concept around which a strategy can be developed. I don’t claim to have any specialist knowledge and I’m not planning to re-invent the wheel, but I will be developing something from the ground-up that I will be trading live (if it seems to work). Bear in mind this could all end in abysmal failure (happened many times), or I could find something that works well (happened a few times). Also, many of the above points may have to be re-visited as new information/data/discoveries/realities come to light.

It is my opinion that money management exits, and psychology are the most important parts to any trading system. Psychology however is not really in the scope of this trading plan and there are some excellent texts available on the subject.

Please leave any comments, ideas or suggestions before we get started.

Re: Making a 2+2 trading system
 

I wouldnt be able to add much/anything in terms of market knowledge, but if you want some programming/math guys, Id be interested in the experience.

Re: Making a 2+2 trading system
 

Sounds like a good plan.

I doubt I'll be able to add anything constructive as I'm really just learning at the moment, but will chip in if I think it'd be helpful1!

I'd ideally like to see something developed for Forex, as this is probably the most accessible market - I am currently learning trading 0.01 lots, so can buy/sell as lttle as $1000 at a time. Would it be suitable for any pair or for a certain one?

If there is any testing you'd like me to do, just let me know!

Re: Making a 2+2 trading system
 

Aside: I never understood the idea of "forward testing". It is just the slowest possible way to do backtesting: later.

eastbay

Re: Making a 2+2 trading system
 

Kimichi,

Nice post covering alot aspects of TA...but there is NO magic bullet foolproof sytstem that will conquer the markets.....period.

This is what makes the best traders in the world what they are...they have the talent/discipline/split second ability making qualities needed to be flexible in their decisions,and not being tied into a numbers specific system or program.Their money management disciplines are uncanny as well

All the things you covered are GREAT tools in becoming/enhancing a traders chance of success....but the HUMAN element is what separates the great ones from the one hit wonders

Go ask the quant geeks with thier phd's in math,what happened during this recent market meltdown with their recent exposure to the CDO instruments tied to the sub-prime disaster?

They majority of these 'fool proof' long/short HIGHLY levered hedged vehicle "traders"....lol.... all scratched their heads saying ... what just happened?

"But we WERE UP almost 30% for the year ,and are now we are down 60%" .....as they had NO idea of a liquidation exit plan,with many of these so called quant programs crowding the trade.Many of them will now be looking for a new job to boot,or just leave WS.I cant tell you how many of thse ideas/programs eventually BLOW UP over the years

I am still in LV now...,as I have stated before ,I am going to start a thread about trading/investing when I return home,that covers ALL of the traits IMHO needed to be a successfull trader/investor(esp. the human/physcological importance of the markets)...not just some mathamatical theoried program ,that might work on paper for a short time frame.....when eventually "mr. market" decides to rear its ugly head and teaches this new system of the week a lesson as well

Sorry to say guys...I only wished it was that easy for ALL of us:(

TTYL,
Stephen [img]/images/graemlins/cool.gif[/img]

Re: Making a 2+2 trading system
 

[ QUOTE ]
Kimichi,

Nice post covering alot aspects of TA...but there is NO magic bullet foolproof sytstem that will conquer the markets.....period.

This is what makes the best traders in the world what they are...they have the talent/discipline/split second ability making qualities needed to be flexible in their decisions,and not being tied into a numbers specific system or program.


[/ QUOTE ]

Tell it to Jim Simons and his crew. He is on the record as saying they never override their automated systems.

eastbay

Re: Making a 2+2 trading system
 

Start at a base system first.

3. Daily
4. Equity Index (to smooth for unexpected exogenous events)
5. Trade set up is the money engine obv
Start simple and expand from there
6,7,8,9 are all subsets of 5 imo
10. Rigorous back testing is a good idea
11. If you plan on writing a book, i guess

Re: Making a 2+2 trading system
 

Interesting read on Simons that was on berg the other day. Here's just a few parts of it.

By Richard Teitelbaum
Nov. 27 (Bloomberg) -- On a hot afternoon in September,
Renaissance Technologies LLC founder Jim Simons is too busy to take a phone call. It is, he says, from Cumrun Vafa, a
preeminent Harvard University professor and expert on string
theory, which describes the building blocks of the universe as extended one-dimensional filaments.
``Get another time when I can talk to him,'' Simons tells
his assistant. Then he mentions that the next day, he'll be meeting with Thomas Insel, director of the National Institute of Mental Health, to discuss autism research. And he's slated that Saturday to host a gala honoring Math for America, or MFA, a four-year-old nonprofit he started that provides stipends to New York City math teachers.
``I'm undoubtedly involved in too many things at the same
time,'' Simons says in his 35th-floor office in midtown
Manhattan. ``But you make your life interesting.''
String theory, autism, math education: It's fair to ask how Simons, 69, manages his day job overseeing the world's biggest hedge fund firm. The answer, judging from the numbers, is very well.
Renaissance is on fire: Its Medallion Fund -- which uses computers and trading algorithms to invest in world markets -- returned more than 50 percent in the first three quarters of 2007. It had about $6 billion in assets as of July 1.
Simons registered that performance as subprime and related markets were collapsing, sending two mortgage-related hedge funds run by Bear Stearns Cos. into bankruptcy. The turmoil pummeled the Goldman Sachs Global Alpha Fund, a rival to Renaissance's funds, which fell more than 25 percent during the same time. Morgan Stanley's computer jockeys lost $390 million in a single day in early August.


Doubling Assets

Along with routine personnel and marketing tasks, Simons takes time for the researchers and programmers who stop by his office to discuss mathematical and statistical issues they've encountered as they work on new trading strategies.
More than 200 employees, of whom about a third have Ph.D.s,
work in East Setauket. Another 100 are based in Manhattan, San
Francisco, London and Milan. ``He creates an environment where
it's easy to be creative and works hard to keep the [censored]
level to a minimum,'' says former managing director Robert Frey,
who worked at Renaissance from 1992 to 2004.
Even without the new commodities fund, Renaissance's assets have more than doubled in a year from about $16 billion on Sept. 30, 2006. That growth has catapulted Renaissance past such titans as Daniel Och's Och-Ziff Capital Management Group LLC, Ray Dalio's Bridgewater Associates Inc. and David Shaw's D.E. Shaw & Co. to become the world's largest hedge fund manager, according to data compiled by Hedge Fund Research Inc. and Bloomberg.

Leaving Academia

In 1977, frustrated with a math problem and eager for
change, he abandoned academia to start what would become
Renaissance, hiring professors, code breakers and statistically minded scientists and engineers who'd worked in astrophysics, language recognition theory and computer programming.
``All the quants in the world are trying to follow in Jim's footsteps because what he's built at Renaissance is truly extraordinary,'' says Andrew Lo, director of the Massachusetts Institute of Technology Laboratory for Financial Engineering and chief scientific officer of quant hedge fund firm AlphaSimplexGroup LLC. ``I and many others look up to him as a tremendous
role model.''
The tendency for fund managers to try to emulate Simons may become more curse than blessing in the years ahead. As the selloffs in July and August showed, many quant funds are chasing the same investments. For example, as of June, Renaissance and rival AQR Capital Management LLC had four of the same top 10 stock holdings: Johnson & Johnson, Lockheed Martin Corp., International Business Machines Corp. and Chevron Corp.

Wise-Cracking

Renaissance is under increasing pressure to stay ahead of the pack -- and to keep its secrets under wraps. Save current employees and a few former ones, nobody knows precisely how the firm makes its millions. Medallion stopped taking new money from outside investors in 1993 and returned pretty much the last of their capital 12 years later. Today, the fund is run almost exclusively for the benefit of Renaissance staff.
The wise-cracking Simons himself is mum on virtually all of its details. What can he say about Medallion's trading strategy?
``Not much,'' Simons says with a chortle, and then takes a drag on one of the Merit cigarettes he often smokes.
What kind of instruments does it trade?
``Everything.''
How many different strategies does it use?
``A lot.''
Simons says his Ph.D.s laugh when they read the far-fetched theories about what their fund might be doing. One chat room participant speculated that Renaissance uses audio hookups to futures exchanges and analyzes the noise from the pits with voice-recognition software.

Re: Making a 2+2 trading system
 

I just read that before I saw this post. Since this is 2+2, also:

"At MIT, Simons worked hard and played hard -- mostly late- night poker."

and link: http://www.bloomberg.com/apps/news?p...efer=exclusive

Re: Making a 2+2 trading system
 

[ QUOTE ]
I just read that before I saw this post. Since this is 2+2, also:

"At MIT, Simons worked hard and played hard -- mostly late- night poker."

and link: http://www.bloomberg.com/apps/news?p...efer=exclusive

[/ QUOTE ]

Damn, that was one long article. I guess you could say he's the Chris "Jesus" Ferguson of Hedge Funds.

Re: Making a 2+2 trading system
 

[ QUOTE ]
...but there is NO magic bullet foolproof sytstem that will conquer the markets.....period.

[/ QUOTE ]

I mentioned this is my OP. Everybody starts off thinking thsy have some magic key. I have discovered myself over the years that there's no magic key. The only wizardry (imo) would be mastering your own behaviour as you make discretionary trades.

[ QUOTE ]
This is what makes the best traders in the world what they are...they have the talent/discipline/split second ability making qualities needed to be flexible in their decisions,and not being tied into a numbers specific system or program.Their money management disciplines are uncanny as well


[/ QUOTE ]

I agree too. Ed Seykota (paraphrased):

[ QUOTE ]
There are 3 golden rules for trading
1. Always follow the rules
2. Always follow the rules
3. Know when to break the rules

[/ QUOTE ]

A good trader knows when to follow his rules and when to change or modify the rules. A machanical system eventually degrades, but only an expert knows when this is simply an expected drawdown, or the system has lost its edge.

Richard Dennis and his crew did fantastically in the 1980s trading a simple trend-following method. But markets do change (from a trader's perpective) and this model no longer is practically tradeable - although it still has an edge as I've tested it recently. There just aren't the susutained trends available in today's markets.

Also: (I think) William Eckhardt said:

[ QUOTE ]
You have to know when to really step on the gas

[/ QUOTE ]

In refering to money management. Opportunity factors may limit the number of high-probablility trades you can make, forcing you to take lower EV trades while waiting. When all your ducks are in a row, a good trader knows when to increase their bet-size and thus improve their edge and expectancy.

I developed a system with a very high profitability. It's losers were both smaller in number and value than winners - and I've never seen another system like it. The problem was, it would only be in the market about 0.5% of the time - ie. one quick trade per year per market, I would have had to scan 1000s markets every night looking for that trade. The opportunity factor wasn't there to make the system worthwhile. It would be these trades when a good trader would increase their position sizes. Since I am not a good trader, I was unable to trade this system.

I think many successful systems traders have these subtle discretionary elements in them. After all, if these human inputs were not needed, then the greatest math and computer brains would have developed a system happily chugging on autopilot vacuuming up the world's money while they slept.

Re: Making a 2+2 trading system
 

[ QUOTE ]

4. Equity Index (to smooth for unexpected exogenous events)

[/ QUOTE ]
Stops can protect against nasty surprises and if I trade a volatile stock which regularly gaps overnight, then I have the option to use guaranteed stops.

My bias (illusion) is that any market can be traded in a simlar way, but I'm leaning towards liquid midcap equities with little or no dividends (dividends screw up my simple models). However, I'm planning on trading shorter term than I have previously and a quick eyeball at my broker's FTSE ex-350 quotes (ie somewhere between FTSE AIM and 250) shows spreads approaching 0.5% which may be too much like paddling upstream. I haven't checked other markets' small cap spreads but I imagine they're at best the same.

The upward bias in stocks however (and the limit on the amount a market can fall), has made trading from the short side significantly less EV than the long side during previous backtests.

I think following highly liquid, highly traded markets like some commodities, forex, DJIA puts you up against the high stakes pros:


from the above Simons article:
[ QUOTE ]
As the selloffs in July and August showed, many quant funds are chasing the same investments. For example, as of June, Renaissance and rival AQR Capital Management LLC had four of the same top 10 stock holdings: Johnson & Johnson, Lockheed Martin Corp., International Business Machines Corp. and Chevron Corp.

[/ QUOTE ]

These guys have to trade the same big markets as their position sizing limits their scope. The little guy with his piddly little account (me) can skip around these guys without worrying about slippage, but can probably also do so on relatively illiquid and less followed (less efficient?) markets.

[ QUOTE ]

5. Trade set up is the money engine obv
Start simple and expand from there
6,7,8,9 are all subsets of 5 imo


[/ QUOTE ]

I'm wondering why you think so. I believe money management is the most important part of a system as your edge is worthless if you bust out. Exits are also more important than entries as they define your initial risks and your profits.

It's evident from many threads on these boards that most people concentrate on entries. "When do I sell GOOG" suggests the investor bought without clearly defined exit criteria. He might have made a perfectly timed elegant entry, but appear not to have considered the exit. It's so easy to allow a winner to turn into a loser, or to prevent a small winner turning into a big winner without proper exits.

I think all the above points (1-11) are related, but need to be considered individually to meet whatever objectives are decided upon in #2

Re: Making a 2+2 trading system
 

i think talking about points 2-11 is kind of pointless until you/we figure out point 1.

i'm quite interested in following such a project and possibly contributing and even putting some money into the market if we actually devise a decent system, but discussing exit points and money management seems silly when we havn't even begun figuring out an edge.

Re: Making a 2+2 trading system
 

What security would this hypothetically trade?

Re: Making a 2+2 trading system
 

[ QUOTE ]
I just read that before I saw this post. Since this is 2+2, also:

"At MIT, Simons worked hard and played hard -- mostly late- night poker."

and link: http://www.bloomberg.com/apps/news?p...efer=exclusive

[/ QUOTE ]

I don't think they're quite on the same level.

Re: Making a 2+2 trading system
 

[ QUOTE ]

I'm wondering why you think so. I believe money management is the most important part of a system as your edge is worthless if you bust out. Exits are also more important than entries as they define your initial risks and your profits.

It's evident from many threads on these boards that most people concentrate on entries. "When do I sell GOOG" suggests the investor bought without clearly defined exit criteria. He might have made a perfectly timed elegant entry, but appear not to have considered the exit. It's so easy to allow a winner to turn into a loser, or to prevent a small winner turning into a big winner without proper exits.


[/ QUOTE ]

In short term trading entries and exits points are all you have- so we are agreeing here. The points sit at the center of a trading system. Specific price zones will define the risk of a particular trade- which is equally important.

Probably most important is stress testing a system for situations that might lead to a total blow-up of the account capital. The equivalent of a "major leak" in poker.

I always ask "how much can I lose" or "where am I going to sell and take a loss" before assessing the potential for upside opportunity. A randomly picked backroll number -- ie "$5000" -- is not a target or a stop loss because it runs the risk of being determined by your personal financial situation rather than the specific influences impacting the direction of the trade. One ought to size positions in a fashion where bankroll never faces a crippling loss with one trade.

But risk management is irrelevant in the absence of the logic layer that determines when one enters and exits a trade.

OBJECTIVES
 

I decided to think about my objectives first, since until I make some kind of inventory, I won’t really have any guidelines on what I want from this exercise. What is it I want to achieve? I think this is pretty much mostly a personal issue, but I’ll make a few generalisations about my own situation as it relates to the mechanics of using a trading system.

I’ll have available the equivalent of $25,000 for this particular system. I have stable money tied up elsewhere but while it isn’t money I can really afford to lose, I won’t top myself if I did. I think this is a reasonable amount so that it is meaningful and I’ll be able to practice decent position sizing using most of the instruments I intend to trade.

I have an income and don’t need profits from trading. If I take this system live, then I hope to have more than $22500 after 1 year in this account. I think it is reasonable to assume I might not make any money (at least in the first year) but my initial profit objectives are actually to limit my losses to less than an average 1% per month over the course of a year. People might flame me for this, but it is not easy losing this little while learning to actively trade. Unless you’ve never tried actively trading, or your name is Timothy Sykes, you’ll know what I mean.

Attainable objectives, baby!

I work during the day but have enough free time. I want to trade using EOD data, but may only be able to adjust (guaranteed) stops during trading hours. I currently live in Asia, so the easiest markets for me to trade would trade during European business hours (GMT 0 - +2) so I can enter orders before the open and adjust them before, or close to the close.

I have a little more time to devote to this system than others. One of my systems involves spending 5-10 minutes per day checking some prices. The medium term system takes a little longer, plus some time to prepare a trade if there’s one available or being managed. I expect to spend a few hours each night managing this system and record keeping depending upon how many markets I can trade, or how often signals are issued.

I need to learn more about statistics as related to trading so I can properly analyse results. (if anybody gas a recommended book/website then let me know).

I also need to learn to program some backtesting software. The entries, exits, and position sizing I use is simple yet still difficult to program (for me) into various backtesting software. I've previously done all testing in Excel and while it allows you to become more intimately aware of your systems behaviour, it takes a [censored] long time.

I’m not going into detail about my personal finances or experience except to say I’m not flush and have limited experience and reasonable success (ie-haven't busted yet!) trading. Follow me to stockmarket rags, SHOE.

Re: OBJECTIVES
 

Also, if people are interested and want to contribute/flame - please keep responses in the relevant sub-threads.

ie-this subthread is titled 'objectives'

A TRADABLE CONCEPT?
 

Many traders don’t have a plan. A plan should be a set of rules to guide each trade. Writing down what you intend to do and how you will react to each situation. A plan helps to remove pesky emotions from trades and helps you become somewhat detached from the market on an emotional level. A plan should remove panic and indecision since we have predetermined responses which have been tested, are logical, understandable concepts, and are known to work.

I suggest we try and trade a simple concept. I don’t have the knowledge or expertise to start looking for arbs or deep statistical relationships to exploit. Simple concepts work well and tend to be more robust. I’ve spent years trying to re-invent the wheel doing this in the past I am currently studying some well-known statistically relevant relationships in various markets. There are lots of apparently good relationships, but it’s difficult to formulate a tradable plan around them, and I suspect I’m still busy trying to re-invent the wheel here…..

The first simple concept I think is whether we are going to buy high or buy low (and sell high or sell low). My first system buys high and sells higher (I can’t short this one for tax reasons) – which is opposite to what people intuitively try to do. This is essentially a trend-following system which follows the strongest uptrends in 25 stock markets/sectors, 10 bond markets and 1 moneymarket. It holds a maximum of 5 positions at any one time and makes approximately 2 round-trip trades per year, and holds positions for usually a few months or years.

Another system I use is a mixture between swing and trend trading. I look for value in 7 long-term stock index uptrends (and v-v) and hold the position for as long as I can until the trend reverses. This holds positions for weeks or months. It has worked wonders on Asian markets but bombed in US markets.

Both the above are relatively difficult to trade psychologically because of the poor win-rate (~30% for #1) and heavy drawdowns accompanied by many consecutive losses, (12 in a row stings) and watching and waiting as yet another tidy winner turns around into a loser.

I propose we develop a shorter-term modification of the latter system. Look for value in uptrends and sell rallies in downtrends – nothing new here. The main difference (aside from the timeframe) from system #2 will be profit targets. I’ve never liked profit targets as it is often the profit from one wild runaway trade that makes up for all the little losers and accounts for more than your year’s profit. Limiting profits by using a target may or may not make sense. It will, however increase the systems win-rate at the expense of perhaps reduced expectancy (possibly negative), and despite the reduction in profit expectancy, may make the system easier to follow and thus more likely to be stuck to during nasty drawdowns.

It will involve selling longs in a runaway uptrend, but avoid holding positions during failed breakouts which gap back or holding on to winners as they retrace back to somewhere near our initial stop – you can’t have it all.

I’ll dig out some charts to better illustrate this concept – but it’s really simple. Then we can try and look for an edge together, or at least a way to trade the concept.

It’s difficult to define or quantify an edge without proper testing and much experience. If the concept makes logical sense then we can move onto the next step.

Obviously, I hope others can make suggestions/flameless criticisms along the way and maybe propose superior ideas before I start digging too much into anything.

Re: A TRADABLE CONCEPT?
 

I’ve found a stock (RGU) which looks to have some nice swings. Notice that this would be refered to as a "penny stock" in the US. It's currenctly trading around 80 pence. Penny stocks are sometimes avoided for various reasons in the US, but the price of a stock isn't really so relevant outside the US. Blue chips can be priced below a pound.

http://i180.photobucket.com/albums/x...irain/RGU1.jpg

There was a nice uptrend beginning after the tech meltdown and lasting until Q2 2005 when it entered quite a wide trading range until the recent summer volatility during which is has tanked. This is a weekly chart using a daily scale, but markets are fractal allowing me to illustrate a basic concept.

I’ve drawn a basic trend line with channels either side 2 std devs away each side from the main trend. This is easy to do in the middle of the chart, but I’ll look for effective ways to quantify (or at least qualify) the trend and the channel when thinking about set-ups making it less ambiguous at the right edge of the chart.

I’ve drawn green circles around oversold (value) areas of the uptrend and red circle around overbought areas. The old adage “the trend is your friend” can be illustrated with this market trend. Assuming 2 trades in either direction:

Long:
2 – 3 (35p to 70p) = 35p move
4a – 5 (75p to 105p) = 30p move
TOTAL 65p

Short:
1 – 2 (45p to 35p) = 10p move
3a – 4 (75p to 55p) = 20p move
TOTAL 30p

So trading with the trend gives more than double the potential rewards here. Obviously only a fraction of each swing can be caught and perhaps most trades will lose money, but the basic concept remains valid.

This means I will be filtering out shorting signals during uptrends and long signals during downtrends. – Another basic fundamental but very important concept.

This stock appears to be following an Elliot Wave pattern during this uptrend. 1-2-3a-4-5a followed by the first retrace of the down-leg (a). The final leg (4a - 5a on this chart) being the most tradable move - but I’m not going to concern myself with market cycles at this point.

This next chart zooms in to the upleg 2 – 3 on the above chart.

http://i180.photobucket.com/albums/x...n/RGUdaily.jpg

Again there is the solid blue line representing the underlying trend with blue dashed confidence lines 2 standard deviations away – easy to draw in hindsight. This time I’ve overlaid a black 13-bar exponential moving average (EMA13). This EMA is enveloped by a channel drawn by shifting this EMA 20% in either direction. This isn’t a Bollinger Band as we are not yet concerned with volatility, but this EMA envelope can be modified later so that volatility data is used in its construction. Notice how wide this envelope is. The distance between the top and the bottom of the envelope represents approximately 40% of the value of the market, yet it contains around 90-95% of the market’s prices. An equity index’s envelope by comparison, might only be as wide as 4-5% of the market. A market with a wide trading envelope represent a bigger target from which we can extract profits with less room for error and a lesser impact from trading costs.

We can better see a similar patter being played out on a smaller scale. 3 legs up with 2 retraces in between. I haven’t labeled these #1-#5 as the chart is already a little cluttered. The green circles again show areas of value in an uptrend and the red circles show areas where the market might be temporarily overbought. The blue ovals show brief areas of potential consolidation which might lead to whipsaws and getting stopped out of losing trades. We can look for ways to limit these losses and potentially eliminate them to some degree through the use of filters – can be discussed further in the ‘set-ups’ sub-thread.

I think we need to be looking for entries in the green areas (below the EMA in an uptrend), looking for exits in the red areas (below the upper envelope in an uptrend) and looking for ways to preserve our capital in the blue areas (when the short-term trend become undefined).

Please make any suggestions that come to mind.

Re: A TRADABLE CONCEPT?
 

kimchi,

I am watching this with some interest. You might want to set some parameters that limit your scope, focus your attention and simplify the universe of options. For example, you might decide to go with

1. Long equities no margin;
2. Selecting only stocks with 1B or more marketcap, selling for more than $10 and avg volume of at least 500K shares a day.

This criteria eliminates shorting, options and thinly traded stocks. I do believe that a trader must get experience beating long-stock without margin before getting fancy. I know this from first-hand experience.

You appear to have done considerable homework. You seem to have an good grasp on many important trading concepts. Given my impression of your study in trading, I am surprised to see that you are apparently thinking about zeroing in on one equity instrument.

Using such an instrument exposes you to the temptation of over-optimizing your system parameters, potentially leading to all sorts of trading tragedies.

Truly robust systems work well using a very narrow range of 'optimal' parameters tested against a great many instruments. Such systems allow you to screen for potential longs that exhibit a sought-for chart pattern or price-action state. The instruments that come through the screen often provide the dividend of helping to identify promising sectors that are on the way up. For example using a screen for specific chart states often reveals several issues in the same sector that exhibit that state.

The book New Trading Systems and Methods 4rd edition by Perry Kaufman is an excellent resource for more information on this and many other essential system construction concepts.

Book:Trading Systems and Methods

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
Aside: I never understood the idea of "forward testing". It is just the slowest possible way to do backtesting: later.

eastbay


[/ QUOTE ]


I meant more about splitting data into chunks and doing separate test on each.

Yeah, I wouldn't use forward testing for strategy testing so much. I think forward testing is more useful to see how you get along with actually using your system day to day. For example, I spent some time backtesting some ideas on The US markets only to discover during forward testing that my broker didn't allow the movement of guaranteed stops outside of market hours. And US market hours didn't fit into my normal daily routine.

Or maybe searching, screening, filtering, datamining and entering/adjusting orders is proven to take too much time.

[ QUOTE ]
You appear to have done considerable homework. You seem to have an good grasp on many important trading concepts. Given my impression of your study in trading, I am surprised to see that you are apparently thinking about zeroing in on one equity instrument.

[/ QUOTE ]

I'm not thinking about trading a single market, and certainly not necessarily RGU. It's just that RGU exhibits some of the qualities that I feel might suit this type of trading. I think we can discuss markets and trading vehicles more later in a 'markets' sub-thread.

Hopefully, the forum will suggest a basket of (perhaps lowly/non correlated) markets to try and trade after we've made a list of criteria needed to be condidered before a market is included in the system.

[ QUOTE ]
Truly robust systems work well using a very narrow range of 'optimal' parameters tested against a great many instruments

[/ QUOTE ]

I intend to do some testing using the same parameters across different markets which will include some stress conditions (ie - 10/1987, bond/debt crises, 9-11, etc)

[ QUOTE ]
The book New Trading Systems and Methods 4rd edition by Perry Kaufman is an excellent resource for more information on this and many other essential system construction concepts.

[/ QUOTE ]

Thanks for the suggestion. I have a few of Kaufman's books on my bloated Amazon wishlist but I've only read some articles and such, and never a book. I have tried to incorporate his adaptive moving average into some systems without successs. I suppose I should just read more about his ideas in his books. I think his adaptive MA is one of those fantastically simple ideas which I've been unable to implement. I actually spent a long frustrating, fruitless time trying to develop my own AMA before I realised it had already been done before - Doh!

Re: A TRADABLE CONCEPT?
 

Here’s another chart. This time a forex pair (GBP/CAD) showing a transition from a long term uptrend to downtrend earlier this year.

http://i180.photobucket.com/albums/x...ain/GBPCAD.jpg

This time I’ve overlaid a longer EMA (70) to illustrate the medium-term trend, which might help us to trade in the direction of least resistance. It’s blue during uptrends (when the previous bar’s EMA value is below the most recent value) and red in downtrends. I have the same EMA15 with an envelope, but this time an envelope shift of 1.5% away from the EMA15 is enough to enclose most of the recent prices. Compare this with the 20% shift needed for the volatile mid/small cap stock (RGU) discusses in my previous post.

So again, we have 2 good potential trades. Point 1 to 2 shows a similar entry/exit area as before, but then there was a change of trend which prevented us from looking for longs as the trend was down, according to the EMA70. Area 6 – 7 shows a good potential trade from the short side.. Notice how we would have covered our short just inside the lower envelope (around 2.26 to 2.22 – a decent profit) preventing us from following the downtrend to around 2.08 – a huge move lost. We may have had other opportunities to enter the trade, but this illustrates a psychologically difficult part of trading. Watching as some big fish get away is tough, but if that isn’t part of your objectives or part of your trade plan, then you must be satisfied with just making a correct trade. Save following trends for another system.

Points 3,4, and 5 are the tricky areas. They would probably have made us enter positions destined to be losers as they were in the middle of a transitioning trend as EMA 70 changed from up to down trending. This creates nasty, but often unavoidable, whipsaws. These losses can be limited though careful use of stops, exits, and/or applying some filters. This is a good example of the importance of getting your stops and exits right. Filters remove some opportunities to trade but hopefully leave the cream of the set-ups to trade. These can be discussed later, but a quick look at the chart suggests a few possible filters. For example, if the entire bar is below the short term trend (EMA15) during an uptrend (blueEMA70), then stand aside – and vice-versa. Or another one – The long-term trend needs to be a certain strength before a trade is taken. There are ways to objectively quantify the trend strength.

These ideas can be discussed later but illustrate the 3 market states; up, down, sideways, and hopefully the 3 trade positions you will match to the market state; long, short, stand aside. Trying to incorporate these concepts into a simple system is difficult.

Re: A TRADABLE CONCEPT?
 

You are on the right track about using price and volume as indicators over 1st and 2nd derivative indicators and oscillators.

Here is an idea:

0. Settle on the value of N and determine when the N-day trend of the S&P is 'up'. Use simple MA values or even simpler compares (if current price is higher than N-ago price).
When the trend is up by your measure, proceed to next step.

1. For the set of all stocks with price > 10
and daily average volume > 500K
and market cap >= 1B

a. If close is > (1.3 * close from X days ago ) //price is 30% up in X days
b. and close is <= (Y-day MA)and close is >= Z% below Y-day MA //price is at or slightly below Y-day MA
c. and Y-day MA is >= (1.2 * close from ((X * .80) days ago) //price is 20% up in (x * .80) days

..Then consider for purchase. This is a starting point. It narrows the universe down to a small list of stocks that:

a. Are trending higher in both X days (30% or more higher) and X*.80 days (20% or more higher)

The hi-order filter (Step 0) limits your universe of stocks for consideration.

The lower order criteria selects stocks that are trending higher but for some reason are challenging the trendline treshhold you define.
Stocks that are doing this while the market is trending higher may be mispriced temporarily.
The risk characteristics are good because a tight stop can be placed very close to entry.

This is just the beginning of a decent robust trend-following system concept.
It buys strongly trending stocks that are threatening to violate a major trend line,
at a time when the market overall is trending higher. It allows a tight stop, since you
are buying at a spot where the instrument is likely to bounce.

I notice that this approach is good at selecting candidate longs when the overall market is trending higher.

I also notice that this approach is- somewhat suprisingly-- good at selecting candidate SHORTS when the overall market is trending lower.

Here are the candidate names generated from this screen, as of yesterdays close for the following values:

X = 300 (EMA)
Y = 60 (EMA)
z = 5

PDA PTR PKX SDA BBL LFC BTM

Thus the universe of over 8000 USA names is reduced to 6 or 7 promising names.
Note that when the overall market is in a strong period, the list shrinks and may have zero members.
When the overall market is in a weak period, the list of names tends to grow.

yahoo link:
http://finance.yahoo.com/q/cq?d=v1&s...DA+BBL+LFC+BTM
Date of screen, and this post: 11/30/2007

Keep in mind that the market may not be in an uptrend as of 11/30/2007; that depends on your subjective value for N.
Looking at the names and industries can often help you ID promising sectors as well.

This is a very simple concept and a mere 1st-step starting point.

However, looking at the names and charts and industries of these instruments
may be interesting to some readers.

Re: A TRADABLE CONCEPT?
 

That looks like a good set-up for a longer term system. I'll have a proper look at your screen tomorrow when I have more time.

I've just had a look at the chart for the first stock you mentioned: PDA

http://i180.photobucket.com/albums/x...rain/PDA-1.jpg


If I understand it right, you are buying short-term value in a longer-term uptrend. Maybe a little similar to what I've been working on above - but using a much longer timeframe.

I think the exits on longer term systems pretty much decide whether they are workable or not. They are so difficult to get right and I think giving up some EV is sometimes worth making them easier to trade.

[ QUOTE ]
he risk characteristics are good because a tight stop can be placed very close to entry.

[/ QUOTE ]

I'm wondering about your entry here. I always think it's best to find a long entry above the market with the intention of getting stopped in as/if the market moves in your expected direction. If the market retreats and fails to trigger your buy-stop, then this stop is simply lowered to the new latest high. This allows for tight stops, depending upon the timeframe you're using. I've highlighted a possible placing for a buy stop - just above the last high. A stop loss can be placed just under the last bar's low or under a low of previous bars depending on how long-term you want to trade this market. Obviously longer term stops will mean higher risk trades but potentially higher rewards if you give the stock a suitably wider trailing stop or some other looser exit.

Using stop loss 1 ($45.50) and assuming an entry around $49, then we have $3.50 per share initial risk. Quite high, but not for a volatile stock like this. I like to incorporate average true range (ATR) into calculation of stop placement as a temporary contraction in volatility can otherwise mean placing relatively illogical stops.

I've also highlighted another interesting bar I'm currently trying to work into another system. Pin bars often signal reversals in the opposite direction of 'Pinocchio's nose' and this stock shows an excellent example of that on 16th August.

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
1. Long equities no margin

[/ QUOTE ]

I looked at my last 4 months of trading and my short trades have been horrible, "Right stock, wrong time" seems to sum up this aspect. Why is it so hard to get shorting correct?

Awful November wiped out my large gains of my last 3 months and back to square one of figuring this out. It was too easy for awhile!

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
[ QUOTE ]
1. Long equities no margin

[/ QUOTE ]

I looked at my last 4 months of trading and my short trades have been horrible, "Right stock, wrong time" seems to sum up this aspect. Why is it so hard to get shorting correct?


[/ QUOTE ]

I think it is quite different to shorting futures or the like due to their upward bias. Even shorting futures is less EV than going long.

What methodology are you using for trading? I find that equities break quicker than they go up and a different technique or method is needed. I tested some long-term trend following systems on a basket of international equity indexes over 10 years or so. I managed an expectancy of $0.30 per dollar risked on the long side but only $0.08 on the short side - and this included the long bear market that started in 2000.

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
What methodology are you using for trading?

[/ QUOTE ]

Over the past few months I have been using the following for my longs(and changing it up; maybe information overload):

-Trading above moving average (from 100-200)
-Trading at 5-10 day lows (I am still trying to find the "magic #"

This seems a good backbone for profitable trading and I have been trying to mix in others RSI, Williams R%, Bollinger Bands, etc., but without much luck yet.

From this criteria, I find stocks I like fundamentally and go from there.

I am finally backtesting starting tonight and will see. On the short side everything flips. I have been slaughtered though, maybe sample size? My timing has been horrendous though too, I missed some good 20%-30% downswings where I mistimed it by less than a week.

Re: A TRADABLE CONCEPT?
 

I have a ton to learn. That being said, there is alot of garbage out there, don't believe the talking heads. Testing some common things you hear, there most likely is not a tradable edge.

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
-Trading at 5-10 day lows (I am still trying to find the "magic #"

[/ QUOTE ]

There really is no magic #. It looks like you're buying some kind of retracement here, but the values you choose for trade set-ups or entries are less important than the amount you decide to bet on your decision or how you are going to exit the trade.

[ QUOTE ]
I am finally backtesting starting tonight and will see.

[/ QUOTE ]

I'd like to know what you are using for backtesting? Are you pouring over historical charts/data or trawling through excel. If you're using software, please say which one.

Re: A TRADABLE CONCEPT?
 

I am using www.stockfetcher.com

The data is only EOD and is $16.95 per month (I got the advanced subscription). You can test 2 year periods over from 2002-2007.

I have been pouring over the data and the most effective trading system produced sell signals on my shorts a few days later than my original entry date and would have saved me a ton of pain!

I am using just a 5 day holding period with no exit strategy currently (to be worked on). So far my data:

Longs: 58% winning rate (average return 0.90% over 5 days not including transaction costs)

Shorts: 51% success rate (average return 0.19% over 5 days) NOTE for clarification: Stock price averages a -0.19% dip over next 5 days.

For the shorts there is only about 20% of the sell signals my longs produce and it is a very small edge, although it is in a bull market).

My short technical analysis had to be extreme #'s to get any edge, had to go more overbought than I did on oversold on the long side.

FYI- I am using RSI(2), RSI(5),MA(50), MA(200), and 5 and 10 day high (or low), and a $5 stock price and 30,000 avg. volume in my analysis to far.

Interesting thread, the combination of this and a good pummeling in November, especially the final week in November, is making me really do some much needed work and self reflection after 3 very easy months! I thought I was the next Warren Buffet [img]/images/graemlins/tongue.gif[/img]

Re: A TRADABLE CONCEPT?
 

Here are the instruments generated by the screen described, as of Friday's closing prices:

NOV PTR PKX PCU LFC

Re: A TRADABLE CONCEPT?
 

Sorry for the somewhat sloppy nature of my post, I see we are both commodity plays although different ones. How does your screen backtest?

For my relative strength #'s, I am using below 2 for my long, the higher the # the less profitable, but still good #'s.

Currently the long screen is really quiet with only a few commodity plays: USO, CFC, OIL

Close reached a new 5 Day Low
and Close above MA(50)
and Close above MA(200)
and RSI(2) below 2
and Close above 10
and average volume(30) is above 25000

---------------------------------------------------

For shorts I have alot more, the most I think I have ever seen going back a few years! The last 9 times (as far as I went back) that there have been 5 or more that met the criteria, 8 times the S&P was down 4 days at an average of 0.74%

CRH
GVA
HIX
HUBG
IMKTA
MINI
MQY
MUS
NAD
NIF
NPT
EAD
JPC
JPS
BMTI
NZF
FHY
HYT
HPS
FHI
FRB
BAC-W
KEY-E

Re: A TRADABLE CONCEPT?
 

Here is the Scottrade chart for OIL (a long):

http://img517.imageshack.us/img517/3141/stockas0.png

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
Currently the long screen is really quiet with only a few commodity plays: USO, CFC, OIL

[/ QUOTE ]

Sorry, I have Countrywide on my mind for some reason. CFC is not a long The three are:

USO: US OIL FUND ETF
GSG: Ishr S&P Gsti Cm
OIL: iPath S&P GSCI Crude Oil Tot Ret Idx ETN

Re: A TRADABLE CONCEPT?
 

[ QUOTE ]
I am using www.stockfetcher.com

The data is only EOD and is $16.95 per month (I got the advanced subscription). You can test 2 year periods over from 2002-2007.

[/ QUOTE ]

Looks like a useful tool. You might also want to try www.prorealtime.com it's a free charting/screener/backtesting platform with EOD data. I've been using this mostly to try some backtesting ideas, but I haven't worked out how to program in proper exits and money management yet. Also, my regular platform is LSE only. I'll have a look at stockfetcher.com too.


[ QUOTE ]
Longs: 58% winning rate (average return 0.90% over 5 days not including transaction costs)

Shorts: 51% success rate (average return 0.19% over 5 days) NOTE for clarification: Stock price averages a -0.19% dip over next 5 days.

[/ QUOTE ]

I wouldn't spend too much time worrying about success rate. This stat is only really useful when calculating the EV of your system. After all, no winning poker player wins more hands than they lose, yet can be +EV. Obviously a system with a higher success rate might be psychologically easier to trade - a very important point.

I think your main job should be to try to make low risk, high reward trades and worry less about your success rate. Being right about a trade is very different from trading right, imo.