Thoughts on trading the financial markets - updated 2010-0708-1145

The following commentary is associated with trading financial markets. It is not a trading suggestion or recommendation. It is compiled as a flow of thoughts, suggestions, and knowledge I have derived from others and from my own years of experience. I update and use it to keep a current accounting of my approach to trading the markets and to apprise those with whom I collaborate.

Giving back: I have been the beneficiary of a wealth of effort freely given by many generous fellow traders to help me and others. Most but not all of it was of value to my trading. And aside from a few con artists, it was all put forth with the very best of intentions. It is with this in mind that I submit the following, that some benefit I have received may be passed on to others. Again, this journal is dedicated to my wife who has put up with, and supported me through many years of education as a trader. May she be forever blessed by my continued success.

As for any additional rationale for all this, there is one: I find it helpful in my trading (you may as well) to act and reason as if I am describing to an apprentice, the proper psychology and methodology of trading. This helps to keep me on track and avoid deviating from my plan. If I am explaining to someone else, the proper psychology and methodology, I had better follow it myself. So I compile my thoughts. Much of the first part this is common knowledge. You are welcome to read . . . if you find it useful, great! If you wish, let me know either way.

The first portion below is a mishmash of mostly disorganized thoughts, anecdotes, and bits and pieces that I have written in response to other's inquiries . . . all of which I have given up trying to organize further. I do, however, read it routinely to reconfirm its validity and edit as necessary. If you wish to skip the introductory and philosophical discourse and go directly to reading about methodology, use the menu above.

As noted in my introduction, learning to successfully trade in the financial markets will likely be the most difficult thing you'll ever do and it should be viewed and approached as if training for a new career. Plan to put in grueling hours of study, tests, years of education, tuition costs, and much frustration. The additional issue I mentioned but is worth repeating, is that you aren't graded on a curve. Doing better than 95% of other traders doesn't mean you pass. The odds are still against you. Losing less than our peers won't cut it. Consistent profit is your only hope. You will have losses but the winning can mitigate the losses and sustain you.

Okay so trading's not easy, but we have to try anyway . . .

Here are some Trading Principles:
Define success.
Find a PLACE to concentrate without interruption.
Be patient.
Be disciplined - leave all emotions outside that place when entering.
Have a plan.
Test ideas simulated before before committing real money.
Own your strategies. Don't rely on others.
Be prepared to adjust when the market changes. (This one is important)
If a strategy is not working, shut it down.
Manage risk sensibly.
Prepare for technical catastrophes.
Never ever allow a loosing trade to turn into a long-term investment.
Give good advice to yourself, just as if you are advising another how to proceed properly.
Walk away from your monitor - sometimes you need to clear your head, take a break.
Know your demons.

What is your idea of success. Is it becoming filthy rich off the market? Earning a comfortable living? Supplementing your current income? Simply breaking even because you enjoy trading so much? Think about it; change it as you see fit. My goal is to earn enough for my wife and I to live comfortably and give away 25% or more of my earnings. I still have some work to do.

If you plan to trade real money, most people need a quiet and private place from which to operate. You should be able to eliminate disturbances, and be willing to ignore phone calls, etc. Interuptions can cause you to miss both positive and negative incidents including major technical glitches, all of which can destroy your profits and your account.

Impatience is a serious stumbling block, interruptions can be a big problem. There are many hours and sometimes days, that we must be willing to do nothing, sit quietly on our hands, observe and aggressively remain idle. But EMOTIONS are the real enemy. Success will be very difficult with any of these affecting trade decisions, especially emotions. I have walked out of my office wanting to either scream or cry, throw my monitor through the window, and throw in the towel. I have let my ego direct me to take the same bad trade many times in a row because I KNOW I AM RIGHT or I want revenge from the market. I should quit but stubbornly I don't, so maybe I get past the current roadblock, find a solution to make an idea work, profit potential is looking good, and I think I might actually have "the market" figured out, elation settles in . . . then "the market" changes and I wreck again. In time I begin to control and overcome emotion and focus on discipline.
So what do you need to do? You need to survive because this business is all about survival, not this minute, today, tomorrow, next week, or even next month, but long term survival. You may need to start again at what you perceive to be square one, where it seems you have been many times. But thankfully, unless you slept through all your prior efforts, the new square one is actually several squares ahead of the real square one.

Make a plan which lets you survive. Survival means that initially you must manage risk by testing simulated before trading real. There is no way around this unless you have nearly unlimited funds to burn through. So your plan includes a reasonable, achievable description of initial success and survival, testing, slow improvement and losing only pretend money, because you will lose. Stick to the plan. Learn, test, practice, find what things work, what things don't. Change your plan if necessary, but don't deviate from the plan until you decide it truly needs changing. Then change it and follow it again. And one more thing . . . sweat the loss of pretend money just as if it is real. Give yourself a reasonable account and don't reset it every day. Force yourself to be financially prudent. And one more thing about your plan; if you cannot manage to describe and maintain a plan in your head write one down. EXAMPLE TRADING PLAN

Own your strategies, build and formulate them, know what they are doing, know how to change them. Don't indiscriminately trust other's strategies. Trading is subjective - what works for others may not work for you or me. There are so many variables in trading: account size, risk tolerance, attitude and personality, emotional self-control, trading instruments, execution technique, market perception/bias, etc. But wait . . . if a strategy is 100% automated, shouldn't one expect the same results to be had by all users. Logical, but I know from experience that differences in platforms, data feeds, etc. are more common than one might imagine. Even within the same platform, data feed, and brokerage you can have different results on different computers. They can be caused by many factors including internal clock time, chart startup time, chart/data errors, bandwidth, CPU speed, the way money management and margin/leverage is applied in a strategy based on your broker's rules regarding account size, etc. A few seconds variation in time can be enough to trigger a trade on one system and bypass the same trade on another. So automated strategies that live on the edge, as many do because of over-optimization, can be quite fickle and subjective as well. So own and develop your own strategies in your brain and on your system. That way you can't blame others.

Because the market is always changing we must be willing to change as well which in turn makes it very important to own and understand all our strategies - to adjust and adapt. It is often said that a good strategy can be depended on to work over time. And years of backtesting may bear that out. There are strategies that work that way but are you willing and able to go through the extended draw-downs they may cause me to suffer. I am not.

Anyway, I approach the market in this way. Strategy losing? It's not working because it's not working, so don't throw more money at it. Nothing works all the time, or even for very long, especially in this rapidly changing technical age when sophisticated computerized trading algorithms can adjust in a millisecond. So do you need to adjust in a millisecond? No, but you CAN monitor the situation and adapt as needed. Sometimes those algorithms are controlling vast amounts of money which can move the market, drawing other automated systems in with them, potentially causing a cascading effect, chipping away at profitable systems and stops. This can go on for minutes, hours, and much longer. If you are hemorrhaging money, shut down your trading and stop the bleeding when this occurs. Turn it back on when it returns to normal or search for an alternate strategy that works now. This approach is different than you will hear from many traders and books, but it is my approach and it works for me.

Manage risk by determining how many losses in a row can you sustain and still reasonably expect to trade your way back to a profit. Account size, leverage and delta, and many other factors come into play to determine our optimum position size and amount to risk. The important thing is to keep your level of risk manageable and understood - don't risk collapsing your trading business by taking too big a position or holding a bad trade for too long. I monitor and adjust position size manually based on my account condition, equity curve direction, my risk tolerance and the current market behavior. I am usually averse to scaling out because of the added complication, additional risk of errors, and additional difficulty in testing and maintaining records. I will definitely scale in, to increase position size if a trade is positive and I get a second signal. Once again, testing is very important - potential losses must be viewed in the context of today's environment because market behavior changes. Risk management, trade position sizing and money management are subjects for in depth personal study - too much complication for this commentary. There are many good books and articles written with workable approaches.

Be prepared for technical problems such as power failures, internet connection failures, computer freeze-ups, system failures at the brokerage or at the exchange, and just as important, problems caused by erroneous entries. How can YOU avoid taking a huge hit if one or more of these things happen at once. Figure it out and prepare for it. A frequent cause or errors - don't rest your hand on your mouse over a buy/sell button. Mistakes happen . . . glitches happen.

Exit a bad trade. Admit when you are wrong. Be objective. If you are holding a position and you would advise someone else to get out, that means that you should too.

Give your best possible advice to yourself. If you've not tried it, you may be surprised how much it helps your objectivity, especially on those difficult days, to pretend you are responsible for showing a good example and training someone else in the proper trading methodology and psychology. Be sure you listen to your own advice.

Just as with exiting bad trades, sometimes you need to exit a bad state of mind. If you are tired and you are not thinking clearly, or if your emotions are starting to build, walk away from your office to clear your head. Take a break, talk a walk, get some fresh air and exercise. Take a day or two off if necessary. If you miss a big trade so what - there will always be another one - your survival had better not depend on the next trade.

Know your demons. We all have demons - if we don't know them in advance, trading will likely bring them out. Try to learn and understand your physical and psychological shortcomings and how to overcome them in times of stress, panic, failure, and just as important, success and elation.

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

A little posturing and background:
My approach to trading financial markets is difficult and tedious but compulsory for me. The lessons I learned from skilled and seasoned traders came after initial overconfidence and many tuition dollars lost. My current attitude came not as a revelation, but rather as an evolution resulting from hearing and reading proper perspectives and rules repeatedly and learning from my mistakes.

Accomplished traders told me it would take many years to learn to trade successfully. Wrong - I thought, not me - I'll just do what they do. After the first year of struggling I worked up to believing maybe it could take two or three years. Later I realized why some professionals have losing months even after they have been trading for 15, 20, or 30+ years. Even the best strategies don't work all the time.

Many traders will tell you to backtest something over years of data. If it makes a profit over time, forward-test it, then turn it loose and trust it. I won't argue against that method. I have seen proof that it works. But it is not for me. Nothing in the market ever stays the same. The question is, how will you will find a way to survive when things are not working according to your plan. For that reason I need a variety of ammunition in my arsenal - alternate strategies. When something that was working is now failing miserably . . . it is time to shut it down and look for something else that WILL work NOW. Despite historical evidence that a strategy should work, draw-downs can go on for a very long time. Don't beat your account to death just because some strategy "should work". Examine some things that worked previously. I discover frequently that my supposition - that a certain strategy that quit working a while back WILL NOT accommodate the market's current phase - is wrong. If I keep looking, something frequently will jump out and work even when things seem crazy-hopeless.

Something very important I know (besides learning patience and leaving emotions outside the office door) is that I need to create and write my own strategies. I must clearly understand what they are doing - I must "own my strategies". I don't depend on codes from others. Since I am a technical-based trader and programming strategies, I must be able to look at a chart, clearly see the patterns, and be able to describe what I see as a trade opportunity. Then perhaps I can program it successfully. Most of us don't have the sophistication of knowledge, trade platform power, or deep enough pockets to outwit the big guys, so we have to ride their coat-tails by observing market behavior.

When the big trading institutions are trying to outsmart and out-trade each other with their super-computers and host of analysts and programmers, how hard can it be to out-trade me? Pretty damned easy - I am not a highly educated software engineer, nor do I have a staff of programmers and a super-computer-network linked up to the floor of an exchange. What I do have is the patience to wait until they show their hand and everyone lines up for the same feeding frenzy.

Fortunately, I am willing to grind through this, day after day. The odds are heavy against me. My attitude and approach discussed above is what works for me. Others have to figure out through their own trials and effort what works for them. I will repeat . . . be very, very patient and don't commit real money to anything until, through trusted simulation, you have proved to yourself that it works, and not until you understand the conditions under which it doesn't work. Lose your emotions - not your money.

Trading Instruments:
I began trading stocks, then moved to options, index options, index futures, etc. Each move looked easier then the previous and appeared as if it was the answer to my problems. But the market was unobliging, and years later I still toil. But now it's routine rather than mentally savaging. I prefer spot/cash currencies, also known as forex (foreign exchange), but not because they are any easier than other markets - in truth they may be more difficult. What I like is the liquidity of a true 24 hour market (the biggest world-wide) with trillions crossing the wires every day. They never expire. They are easy to understand. I like the variable leverage, and especially I like the almost infinite capacity to control position size.

I find it appealing that there is no single exchange where everyone's currency trade clears but rather hundreds of banks and networks to which the world's biggest institutions (and I) demand and receive instantaneous access. This helps to make technical failures less likely and manipulation more difficult. However, every financial instrument including forex is traded by people who try to, and to some extent, succeed in manipulating the market.

With currencies, I can take a very small position and hold it indefinitely or take a large position for a few seconds, or anything in between. Currencies can sometimes trend seemingly forever, but there are also long periods when they will churn painfully sideways or trend very slowly with sickening, unpredictable volatility. I use those difficult periods for testing new strategy ideas and retesting and revising old ones.

If I want to trade, I have to trade what moves, so in addition to currencies, I will trade gold, oil, a handful of other commodity futures, and equity index and bond futures. I will trade the others when the currencies are behaving badly. For my style of trading these instruments behave better than individual stocks, which are more easily manipulated and influenced by external, internal, and fraudulent actions. In testament to that belief, due to my earlier stock trading history, I have a file drawer full of legal documents (which I am ignoring) inviting me to claim my share in class action litigation against various market entities that have indirectly and fraudulently drawn me and others into stock trades. The effort and time needed for documenting each and every transaction would be better used elsewhere, so the attorneys are winning that round.

Trade Service Vendors:
For market data and brokerage account I use Interactive Brokers (IB). Although IBs trade management interface is very good, for me, their charting is inadequate and their platform is deficient in auto-trading capabilities. But IB has a low commission "universal" account where I can trade virtually any instrument from the same acct, and they have an open architecture which allows third parties like Sierra Chart, to access their system. I have heard complpaints about IB's lack of concern and marginal customer service, but I have had no major problems since I began with them in 2002. I use Sierra Chart for charting and trading and have been with them almost as long as IB. Sierra's response to inquiries and requests for system improvements has been very impressive, both accommodating and fast. I have used other vendors including Tradestation and NinjaTrader, but I find IB and Sierra to be very versatile and favorable for my style of trading. They are also reasonable in cost, enough so that if I want to take a few months off trading, it does not bother me. Sierra and IB together work quite well. For auto-trading and indicator programming Sierra offers a C++ environment, ACSIL (their own proprietary language), and an Excel spreadsheet style programming interface (Sierra calls it a Workbook containing one or more Worksheets). IB offers both a Mac and PC interface, but Sierra is limited to operation on a PC.

I elect to use Sierra's Worksheet Trading System because worksheets are fast and easy to program and because I don't want to spend time learning to program C++ or other complicated languages. Anyone with a basic understanding of Excel operators and functions can easily program trades and custom indicators using these worksheets. With Sierra's Worksheets I can quickly create auto-trades and create my own custom indicators or studies. Sierra also provides a couple hundred standard and proprietary indicators. Through Sierra, I can simultaneously run auto-trades (entries and/or exits), trade directly from Sierra's charts, trade from a price-ladder/DOM interface, or from a very simple trade window which all link solidly into IB's trader-workstation. I neither work for, nor receive any benefits from Sierra; I just promote them because I feel they deserve it.

It is good practice to employ an additional brokerage(s) to handle backup for all the instruments traded. In the event your primary brokerage becomes unavailable, a position can be neutralized by taking the opposite position with an alternate brokerage. For this service, a no-fixed-cost or very low-cost but trustworthy vendor with online and telephone service should be selected.

Modus Operandi:
Note: The information below is compiled here for my use and is not a trading suggestion or recommendation. You are welcome to read it. If it helps anyone, I will be pleased.
My technique makes me a short term trader, usually intra-day, although due to the 24 hour liquidity of currencies, I will hold currency positions overnight, and up to several days if all is going well. With the leveraged risk of currencies, however, I will not hold a large or medium position through a weekend due to potentially being trapped during a catastrophic economic or political event. I prefer to trade with-trend which is of course, subject to what one uses to measure it and to the timeframe in which observations are made. Those don't matter much to me, since I am always adjusting to changing conditions. I also cautiously experiment with methods that may allow trading both directions during sideways markets.

I am flat, out of the market most of the time, but time-off works into my plan. While flat, I forward-test at least one trade strategy continuously (simulated) on multiple charts/instruments. I don't discard strategies that aren't working - there is/may be a condition for which I wrote them and that condition may come to life again some day. While a strategy is not working (i.e. Sim trades are failing), my system remains in a simulated mode. This can last for minutes, hours, or days. I sim-test various strategies, different chart periods/ranges/ticks/time frames, switch from close bar entries to intrabar, test and revise old strategies into new ones, and simply fish for something that works. Nothing I have ever found works all the time. For that reason, I avoid continuing a strategy that is in an account-drawdown phase. Why suffer. While I am backtesting and fishing for something that works, I am obviously also identifying strategies that are not working. If a strategy has not worked leading into the current market phase/condition and up to the current moment, why should I employ it and expect it to miraculously work now. Yes, the market can suddenly change, but I will let the market tell me it has changed rather than trying to predict when, while hoping I'm right.

As the market goes in my favor and I get additional entry signals I usually add to my position, often tripling or quadrupling my initial position. This places multiple stops to back-up each additional entry. This can work against me if the market is volatile and choppy due to large pullbacks. When I sense that condition I will move my stops to breakeven or better and when parabolic moves occur, I will often move one or more stops right under the market, allowing the taking of partial or all profit.

For exits I use various combinations of fixed stops, BE stops, indicator based exits, time/price-based programmed exits, and manual intervention, but no predetermined profit targets because I don't want to place a limit on the upside.

Although my charts give me clear visual and audio signals, depending on market conditions, I can be either overwhelmed or sometimes bored watching eight or more charts tick along. For that reason I try to use automatic entries as much as possible. With automation I don't miss trades. Only my system/strategy may miss trades due to its imperfections or the imperfections of my dataflow. During simulation I monitor the Maximum Open Position Profit vs. the Maximum Open Position Loss. When the the Max Losses are greater than the Max Profits, the strategy is failing. If the pattern begins a shift to Max Open Profit becoming greater, I assume that that trend-following strategy is beginning to work. If I happen to be at my trade desk at the that time, I switch the chart(s) from Sim to Automatic-Live. I take into consideration the difference between Max OP Profit vs. Max OP Loss. The ratio should allow for slippage and commission, and stops should be determined accordingly along with a consideration for ATR. Due to various correlations, typically, several instruments or currency pairs start to generate profits at the same time. The immediately preceding Max Open Profits and Losses give me an idea of what levels of profit to expect on subsequent trades.

Sometimes a price pattern/trend will continue, repeating long enough for the trade series to yield a good profit before reverting back to losses again, at which time I switch back to Sim. Sometimes not, and the trade series yields a loss. A good series of profitable trades can make make up for many losers.

When the pattern goes negative, I switch back to Sim and revert to testing. Often the trend is short-lived which will produce losses because I go Live after it starts and quit after it ends. But if the trend lasts several hours or more, my profits are far greater than losses. Sometimes the trend tricks me, appears to end but really doesn't, and I jump back in. Sometimes I shut down the Live trades early because I can see the move is unsustainable and weakening on the charts and I want to avoid churning away a nice profit. Yes, this switching is discretionary; perhaps some day I will have a clearly definable plan when to automatically switch off and on. Anytime - in the middle or at the end of each day, I can look back at all the Live and Sim trades by loading the recent records (both Sim and Live) into my database. I compare Max Open Profits vs. Losses, trade duration, and the other obvious statistics, and make adjustments for future trades, either mentally, in my strategy code or both. Virtually every kind of statistical test can generate spurious conclusions including my technique. But it's fine for the time being and I continue trying to improve it because it works and I truly enjoy the challenge.

Market phases:
The market is typically either IN a phase or CHANGING to a new phase. Phases can last a long in terms of short-term trades, sometimes for several hours and considerably more. Phases can be trending, sideways, consolidating, distributing/acquiring, whatever - it doesn't matter. What matters is that some phases are tradable and some are not. If you can identify a tradable phase you can take advantage of it. Quick identification and evaluation to find a strategy that works sets you up for a profit. The advantage of a system like this is that I am always adjusting to current conditions rather than spending a lot of time under drawdown conditions, hoping that my long-running strategy will come back to historical profits.

Sierra Chart offers a backtest feature I use to help me with this. Their backtesting can be run at speeds up to 10,000x. Although running that kind of speed backtesting eight or so charts might be too inaccurate and probably too taxing on my CPU as well, there are reasonably fast speeds at which I can test with acceptable accuracy. To prepare for this I check my backtest accuracy by running identical tests at different speeds and comparing results. I pick the highest speed I am comfortable with. I note that the market has recognizable conditions of trend and volatility. If the market is in a recognizable and repeating pattern, I look backwards over various range/time/tick-bar periods. If I recognize a condition where a strategy may work in the current market, I quickly backtest it and some similar ones over the current pattern's period. If the recent test proves profitable, I apply the strategy forward in real-time . . . . .

I religiously monitor US and international economic news schedules and place audio alerts on my calendar prior to any potential market-moving news releases. I don't particularly care about receiving/learning the news immediately because someone else will learn and analyze the news faster and more accurately than me, either legally or illegally. I just need to know the schedule in advance. If there are news leaks, they tend to reveal themselves a few minutes early in the price action.

Occasionally, I may choose to take a position immediately after a news release if it appears that the move will sustain. Often however, I don't need to manually enter because one or more of my auto-trades which I turn on after the announcement has done it for me.

I also do longer-term strategy backtests, but not while I am trading live, so as not to disrupt the market data flow and analysis. Sometimes I forward-test a strategy for several hours, all day, overnight, several days . . . whatever. I save all statistics using Sierra's activity log and load them into an evolving database that I created so I can analyze the data. I evaluate trades by time of day, Maximum Open Position Profit and Loss, individual strategy success, instrument(s), depth of losses (stops), etc. I note interesting tidbits such as . . . Certain hours of the day or night and/or days of the week will yield significant profits vs. other times which routinely incur big losses. Since currency markets run 24 hours I observe middle of the night stats. When North America is closed, having dinner, or sleeping, the other side of the world is releasing economic news, interest rate changes, opening exchanges, businesses, banks, etc. I might also note that certain currency pairs are more consistent winners than others, even when averaging all my strategies together. EUR/JPY, AUD/JPY, AUD/USD and EUR/USD are current examples.

I save strategies I have written that were successful as well as those not successful. Regarding those have stopped working, I spent hours, sometimes days compiling and revising the strategies for a reason, because they made money. The market changes and the strategies stop working, but it may come around full circle and return to conditions that supported previously profitably trading methods. Make note of the market's condition when a strategy is working so you can recognize it and use it again. And keep something else in mind, when you tweak a strategy, you are really writing a new strategy, so any statistics from the previous version are completely invalid and have no association with the revised version. So technically we are not tweaking strategies but writing new ones each revision.

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

Strategies and Methodology:
There are dozens of strategies that work in various markets at various times and I will site several examples below. But are there strategies that stand out worthy of being called the "holy grail" that everyone is looking for? Clearly, none that I have ever found. Apparently there are some computer algorithms that are sophisticated enough to identify and self-adjust to any market condition, but those are the work of a programming staff at Goldman Sachs and the like and not in my realm. Maybe some of you can write code at that level. If so, I don't know why you are bothering to look at this web site, but I hope you enjoy it anyway.

Charts and Indicators:
Range Bar Charts: Range bars are my favorite way of plotting price action, for setting up and triggering trades, and for general viewing of financial markets. They offer a predictability that no other chart plotting method offers, excluding perhaps a few highly specialized studies such as Renko or Three-line-break charts which for me, seem to distort or hide price action. With range bars you know way ahead of time the limits of where the the current bar will close and where the next bar will open. Range bars enable you to easily identify a sideways market. Range bars will with many strategies, greatly reduce trade-triggering in sideways conditions for obvious reasons. They give excellent opportunities for price action based strategies. With range bars you NEVER have large "data-spike" bars - something necessary to avoid if your strategies are designed to enter and exit at bar close. Range Bars also enable you to take advantage of quickly changing momentum.

A note about Fibonacci - I like to use Fibonacci (or "golden ratio") increments when experimenting with and adjusting tick bar or time periods, standard deviations, channel lengths, moving average lengths, etc., not because I believe the numbers hold any magic in the financial market, but because there is a beautiful logic to the way Fibs increment themselves up or down and I like both the pace and ratio at which they move. One also has to consider that they are well known as a major component in nature, used in architecture, number systems, geometric shape sequences, etc., and most importantly in many technical analysis techniques, meaning other traders employ them as well. So why ignore such a practical system.

Many of my preferred strategies involve determining a trend by using from two to five linear regression channels/lines of different lengths, typically measured in Fib increments. If the currently (usually at least two) employed LR Channels confirm the trend direction, then I trigger a trade entry at the failure of a countertrend trade. See example below. For a trigger I use a chart-based indicator with an Average True Range (ATR) input that continuously simulates a trailing stop in either direction for any arbitrary/imaginary counter-trend trade. When that trailing stop is triggered (meaning the counter-trend trade has failed) and all other above factors align, my strategy automatically enters a trade in the opposite direction, i.e. with the trend. Typical inputs or settings for the strategy are as follows. Bar periods/ranges are based on your preferred risk tolerance and trade time frame. I typically use something between 144 and 610 tick bars. For range bars I use an increment of 3-10 times the value of each individual tick. Linear regression channel (4) lengths are 34, 55, 89, 144 for the low end, or 55, 89, 144, 233 for the mid-range, moving up to about 987 maximum ticks for the high range. I usually move in fibonacci increments and may skip a level, jumping from 144 to 377 for example if I want to spread things out a bit.

Notes for the above chart: Blue colors indicate downward bias and aqua = upward. Linear Regression Channels lengths are 55, 89, 144, and 377 (magenta). The price bar colors are controlled by the simulated 1.5xATR 21 period trail-stop. In this example a short trade could be triggered where the bars turn from aqua to blue (a few bars from the right edge) indicating a failed move upward. The price in the upper half of the LR Channel is a plus. The mid-chart plot is Time & Sales with Linear Regression lines on the T&S bid. This example shows the condition where all T&S Lin Reg is negative, indicating a strong immediate downward bias. This condition usually only lasts for a few seconds but may return frequently if momentum is strong. The lower region of the chart (not cited in this example) shows Linear Regressive Slope conditions of four varying lengths and the bottom line shows the condition of a 2.0xATR 21 period trail-stop.

I expect rapid favorable momentum, which is a good indicator of whether or not the strategy is working and the trend momentum is sufficient a trade. If the position is down $100 and the Maximum Open Profit during the trade is $10, something is wrong. If this happens several times in a row something is very wrong. My expectations . . . it should be the other way around.

Regarding the above methodology, if the market becomes choppy but still is trending, I may require that the entry occurs only if the current or previous 1-3 bars are within a certain distance, e.g. upper 25% of the LR Channel range for a short or lower 25% for long. I pick a channel deviation between 0.6 and 1.6 depending on volatility. Because I like to use price action, I may employ Donchian Channel (along with or in lieu of Lin Reg) and require that at least one high of the last three or four bars came within 10% on the top line/highest high for a sell and vice versa for a buy. This may be combined with the approach described above.

Another channel that I like, mentioned above - the Donchian Channel (DC) - simply shows recent highs and lows for a given number of bars back, so for someone wanting to create some trades from price action, it offers a convenient vehicle for simple (I love simple) price patterns. I like to combine the Linear Regression, slope/direction and proximity with a simple move away from recent highs/lows for short/long, perhaps also with the trailing-stop simulator on my side.

Anchored Linear Regression: I occasionally anchor the left side of one of my LRCs if warranted. On Sunday evenings when the Fx market opens, I look to see if there is a significant gap from Friday evening's close, indicating a change in sentiment over the weekend. A gap may may prompt me to anchor the left side of a trend-identifying Linear Regression Channel to the Fx open or perhaps the previous Friday's close. If there is no sentiment change, I will either anchor the LR Channel to a recent significant high or low or allow the LRC to maintain a specified length. From that point on, until there is clear change in trend, that LR slope direction is the trend.

Another filter I may add to any strategy is measuring Time & Sales (T&S) - shown in the above chart) to which I apply multiple lengths of Linear Regression (LR). This filter only allows a trade to be taken when all LR channels slope in the same direction indicating a T&S momentum consensus. I apply LRs to either T&S bid or ask at the following period increments: 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, and 987. This works fine for the major currencies and liquid equity, bond, and commodity futures. When trading less liquid instruments, one or more of the longer periods can be omitted. Above strategies filtered in this manner along with really tight stops typically allows me to take very small losses and jump in at points of strong momentum.

My automatic exits may be triggered by the above-described simulated trailing stops or by the slope-reversal of short term Linear Regression Channels (34-55 bars). I may also decide to adjust stops manually, as described above, to minimize losses or move to break-even, overriding my automatic exits. Doing this when it becomes choppy can turn potentially big losing days into small losers and occasionally into small profits. Depending on market conditions, I may use an automatic exit strategy that contains an instruction to take me out of a trade within a prescribed time limit if P/L is not positive; this is why I pay attention to trade duration.

I also like to employ price patterns as triggers typically with trend identifying indicators as described above, the simpler the better. Sometimes a very simple 3-bar or 4-bar pattern may trigger profitable trades for a period of time. Broader price patterns like head & shoulders, wedges, and other geometric patterns can tip me off to a change in trend or a return to previous trend, but are not used as the main triggers in trade strategies.

Below is one example of a simple price pattern that can work surprisingly well, in fact this pattern has been working on and off for many months now. This simple pattern and similar ones were brought to my attention in a book by Suri Duddella. As mentioned above, watch for it to start working in simulation, then run it live until it starts to fail. I have found that this and many other simple entry triggers combined with a short term momentum filter like T&S-LR above will yield positive results in a majority of trades. You just need to determine when to exit with your profit. You need to test it on your own chosen instruments, play with bar period and combine linear regression proximity and direction to filter entries and determine trend.

If the High 3 bars ago > High 2 bars ago, and Close 2 bars ago < Close 1 bar ago, and Low 1 bar ago > Low 2 bars ago, and High 1 bar ago > High 2 bars ago, and High of current bar > High 3 bars ago then enter a long position.

Tradestation style formula is: If H[3]>H[2] and C[2]<C[1] and L[1]>L[2] and H[1]>H[2] and H[0]>H[3]

S-1 PATTERN EXAMPLE

Breakouts are also one of my favorite methods of trading. I like to wait for a pullback after a breakout and then look for strong momentum in the direction of the breakout to enter.

Directional changes in a very fast moving average like Sine Wave can also be used to trigger trades, provided other filters are passed through.

So . . . I nearly always enter trades automatically and I add to my position as the market goes in my favor and I get additional entry signals. Depending on conditions I may or may not exit manually. When I started trading, automated trading was typically unstable and in early development stages, at least for retail traders. It would have been easy back then to run successful automated strategies if the software/hardware was available. Now, with so many computerized, algorithmic, extremely fast and high volume trades occurring through the big institutions, hedge funds, and on-location exchange-based traders, you never know which market condition will prevail at any given moment. You have to be quick with your reactions and be ready to change your diagnosis.

I have learned to be ready for anything. Technical glitches are our second worst enemy after our own shortcomings. Both my Mac and PC systems, including modems, routers, monitors and speakers are on a UPS that can handle a power failure long enough for me to to exit all positions. Opening an account with a second backup broker is a valuable insurance policy. I keep phone numbers handy for brokers and for others I know and trust who trade the same system or are at least comfortable enough with computers to log-in and exit positions on my behalf. I use stop limit orders when logical to force placement of stops outside my system onto a broker's or exchange's server.

Many people have asked me to post my ideas. Above are some details and an overview of my technique. I realize that there is much ambiguity in the information, but that is because of my continual adjustment to current market conditions. I will add to and edit this journal whenever I have the time and find new ideas, so if you have visited here before, it is a good idea to refresh your browser to display updated material. I hope this helps anyone who takes the time to read it. Email me if you care to discuss the above.

Home | Coal Futures | Baltic Dry Freight Index | TS Code | Trading home | Dave home | Email me