I don’t really see a forum dedicated to pure trading theory / methods, so I opened this thread (would be useful). If this threads lives, then so be it.
I have never been a strong believer in technical analysis, because to me, things like Gann, Elliot Waves, pitchforks, Heads & Shoulders, Japanese candlestick and most technical analysis methods are rarely subjected to comprehensive, rigorous testing.
I pulled some hard numbers about a number of standard tech analysis techniques, over the course of 1 or more years, as applied to some futures contracts. Unfortunately, the C2 forum doesn’t allow for much formatting…
Here is the results of the S&P 500 June contract trading one contract:
Indicator…Total Number of Trades…Average Days per Trade…Total Profit from Trades
Trend Spotter ™ 18 Trades 29 Days/Trade -34525.00
7 Day Average Directional Indicator 63 Trades 11 Days/Trade -12850.00
10 - 8 Day Moving Average Hilo Channel 85 Trades 6 Days/Trade -26425.00
20 Day Moving Average vs Price 65 Trades 11 Days/Trade -34575.00
20 - 50 Day MACD Oscillator 13 Trades 50 Days/Trade -35275.00
20 Day Bollinger Bands 29 Trades 4 Days/Trade -2775.00
40 Day Commodity Channel Index 40 Trades 7 Days/Trade -45550.00
50 Day Moving Average vs Price 37 Trades 18 Days/Trade -30925.00
20 - 100 Day MACD Oscillator 6 Trades 96 Days/Trade -17650.00
50 Day Parabolic Time/Price 52 Trades 14 Days/Trade -19475.00
60 Day Commodity Channel Index 28 Trades 9 Days/Trade -17825.00
100 Day Moving Average vs Price 10 Trades 58 Days/Trade 20200.00
50 - 100 Day MACD Oscillator 5 Trades 115 Days/Trade -48300.00
The final number is the total over the real-world trading period. It doesn’t bode well for tech analysis in this instance.
If there is interest, I can pull similar results for most other futures contracts (don’t have access to individual stocks or forex pairs, but I can pull tech analysis results for the major currency futures like Swiss Franc, Yen, Canadian Dollar, I believe).
To a certain extent, you are correct. Each technical analysis method has its strength and weaknesses and thus, no single method is adequate for trading. However, I disagree in the usefulness. Combining the right kind of techniques can enhance predictability to a high degree. The trick is in knowing what methods to combine and what parameters to use for each method. In the latter, simply using default settings doesn’t always work well but a tweak in the numbers can definitely show results. One really needs to understand market character and how each indicator reacts to changing market conditions. This way, one can understand how to change settings on indicators to be more useful as well as apply the right filtering techniques to generate more positives in trade signals. Combing techniques is a must though…if you can cross validate signals through multiple methods, the odds of success can be greatly increased. Technical analysis should not be used in the cut and dry format you find in most charting sources.
"I have never been a strong believer in technical analysis, because to me, things like Gann, Elliot Waves, pitchforks, {snip} are rarely subjected to comprehensive, rigorous testing."
Actually, Prechter using pure Elliott had a spectacular 10 year run well documented in real money trading and real time picks. Granted his performance since semi retirement in 1989 has been on again off again, but trading is a tough job. If a pro basketball player averages 30 points for 10 years, but can’t play when he’s 60, do we say he just got lucky for a while when he was young? FWIW, I think trading is MORE difficult than pro basketball <G>. The games last longer (all day, every day) and once a trader has a few hundred million it’s hard to hold that edge.
Here is the results of the S&P 500 June contract trading one contract:
"Indicator…Total Number of Trades…Average Days per Trade…Total Profit from Trades
7 Day Average Directional Indicator 63 Trades 11 Days/Trade -12850.00
100 Day Moving Average vs Price 10 Trades 58 Days/Trade 20200.00
50 - 100 Day MACD Oscillator 5 Trades 115 Days/Trade -48300.00
The final number is the total over the real-world trading period. It doesn’t bode well for tech analysis in this instance."
I agree simple indicators in and of themselves are often of little value, but there are trend following technical ideas that have stood the test of time. Year after year, for several decades now, there are several names (Dunn, Henry, etc, (who are mostly mechanical traders)) that are near the top of the CTA list:
http://www.autumngold.com/Performance/YTD%20Ranking.pdf
How do you get lucky every year for 20 years? No doubt, it’s not easy,
but to say it can’t be done just isn’t true.
Also, I’m guessing your stats above are for daily data on the S&P. My research shows it’s tough to make money off the daily time frame with these types of timing indicators particularly with an always in trading style. The markets are fast and they trade 24 hours now. Daily data is poorly suited for short term turns. I can show you shorter time frames where pure mechanical methods (using the ADX along with short term trends) have done very well over the last year, as well as the last decade. Some indicators (like the ADX, MA, channels, etc) may not work as a “method” in and of themselves, but can be used as filters or in combination with each other.
simply using default settings doesn’t always work well but a tweak in the numbers can definitely show results
There is likely truth in this, and obviously, market geometry/structure can influence the value of indicators.
I often wondered about how much relationship exists between the theory that over 85% of traders use tech analysis, and the fact that over 90% of futures traders lose their money.
I am mainly hoping to open dialogue, since what works and doesn’t work to me is far more important than “the psychology of trading.” My psyche is a lot happier with profitable algorithms, than feeling good while using losing algorithms.
"I’m guessing your stats above are for daily data on the S&P. My research shows it’s tough to make money off the daily time frame with these types of timing indicators particularly with an always in trading style"
Yes, this is the current results based on daily charts (OHLC), for the June 2006 S&P futures contract.
I read Futures, and Tech Analysis of S&C magazines with a very simple rule. If someone discusses the merits of a particular method, and uses backtested results over a reasonable time frame (year plus, not 3 days), I will study it (as a minority of their articles do). If they only use a couple of examples, it goes into the trash.
"I often wondered about how much relationship exists between the theory that over 85% of traders use tech analysis, and the fact that over 90% of futures traders lose their money."
I don’t know any traders who made money on fundamentals. I know some that lost a lot. I think most traders lose money because they haven’t done their homework and/or they don’t stick to their plan. Even
the best systems have difficult times (please see below).
"I am mainly hoping to open dialogue, since what works and doesn’t work to me is far more important than “the psychology of trading.” My psyche is a lot happier with profitable algorithms, than feeling good while using losing algorithms."
The most robust methods (see Dunn, Henry, etc. in the CTA list) are often the most difficult to stick with. Trend following systems can have low % of winners, sometimes big DD’s, etc. Please note these guys have $10-20 million MINIMUMS… But with 20 year track records and $300-$500 million in equity, you’d better believe they have been doing something right. Can regular guys use these methods? Yes, but it’s not as easy as you might think to keep your psyche happy. A lot of people jump ship on that first DD.
Just BTW I know some of the other top guys on the CTA list. They haven’t been around as long, but they are also using trend following/breakout methodologies. Variations on the breakout theme have also been used for short term trading (see Toby Crabel).
Because a canned channel breakout or moving average doesn’t make money doesn’t prove a thing. If all you needed to do was plot a MA and have money pour into your bank account everyone one would be rich. It’s like believing you can buy a magic wrench for $19.95 from a TV commercial and become a great auto mechanic. It just isn’t that easy.
Good tools will help, but you better know what you are doing with them.
"it’s not as easy as you might think to keep your psyche happy. A lot of people jump ship on that first DD."
On that note please check out recent comments. The two best systems on C2 just had subscribers bail out and post. One complained there was a drawdown since Tuesday night… another that the DD was too large. In both cases the systems have used pretty heavy leverage to achieve their spectacular returns. As we like to say: “leverage is a double edged sword”. You can’t enjoy the gains if you can’t handle the pain.
The sad thing is, the hype and attention people give to tech indicators. There are a stream of articles and tutorials in these trade mags and on the internet, as if they are can’t-miss. And many advisories seem to think that throwing a lot of indicators at the problem will guarantee success. Reminds me of the 90s, when a survey showed that most Americans thought the market would keep going up 20% a year ad nauseum…
A lot of us would probably not give away our best methods, but you are right that breakouts and following the trend are in the right direction. One of the universal things I have found is, when filtering an advisor’s trades, the signals with the trends were almost always significantly better than those against.
the systems have used pretty heavy leverage to achieve their spectacular returns
Larry Williams became famous from his World Cup $10,000 to over a million run the year he won (he still holds the Robbins World Cup record. This was overleveraging to the hilt. One wonders what running a Monte Carlo simulation on his track record would look like!
Very interesting reading here.
Ill declare my trading style bias straight away ; I trade a purely mechanical system based on technical analysis and Ill explain why.
Trading methods such as technical and fundamental analysis often attract a lot of debate. Most people will strictly adhere to one method and swear by it. In my experience, the only way to find an edge in the market is to research different approaches to find a statistical edge, which is why I find the trading results posted by Ross very interesting its always good to see an argument backed by some numbers. Firstly, Ross is right, these numbers dont bode well for TA at all; however, numbers in isolation generally mean very little. To me, these numbers demonstrate just how difficult is it to find an edge in the market, whether you use TA or some other approach. As Sam said, If all you needed to do was plot a MA and have money pour into your bank account everyone one would be rich.
I think Sam hit the nail on the head with this: Good tools will help, but you better know what you are doing with them. Technical analysis is just a tool. Giving it more credit than it is due is dangerous to your account balance. All TA essentially measures one variable: price. One system on C2 proudly states that they employ 10 TA indicators. Thats hilarious as theyre all measuring the same thing.
My biggest problem with fundamental analysis is this: every moment in the market is totally unique. Every minute of every day, a collection of buyers and sellers are coming together, making decisions based on all sorts of reasons that ultimately move price. The chance of these buyers and sellers making the exact same decisions at the same time is very unlikely indeed. Pure fundamental analysis rests upon the assumption that you can predict what buyers and sellers are thinking. Remember that these thoughts are translated into price, and Im there finding a statistically probable result with my TA
Peoples decision making processes are (in general) just too complex to be reliably captured by a fundamental approach. Some people are very good at it, but this is a skill, not a quantifiable method. If anyone is interested, they should read a book called The Winners Curse, or an article called the Paradox of Choice. Food for thought
Ok, Im now going to divulge my trading secret: money management. This is the key to consistency and profitable trading over the long-term. I personally use a program called Market System Analyser to measure drawdown, equity curve and dependency with every single trade. If my equity curve drops off too much from its moving average, I adjust my position sizing accordingly. If I see that some trades have a strong dependency, I change. I adapt to a changing market. Its a fantastic program and well worth the price (and no, Im not the developer, and I dont get a cut from recommending the software!).
I always laugh at the best seller lists on websites selling trading books. Most of the titles have the word secret in them. Everyone is searching for the holy grail. The truth is that the vast majority of traders really dont understand the game. Sam notes that some subscribers bailed out from a C2 system because of a drawdown. It is hard to handle the tough times in trading, and no amount of technical or fundamental analysis is going to help you sleep at night. Whenever I read subscriber reviews that complain of a loss I really feel that these people would be better off putting their money in a mutual fund. New traders need to be educated, which I why Ive written a small intro to trading for my clients. You need to understand that losing is part of winning. This is the psychological aspect of trading. Perhaps its better left to another post
"Most of the titles have the word secret in them. Everyone is searching for the holy grail.“
If I had a nickel for everytime a book said
"the secret that Wall Street doesn’t want you to know”
…or even better,
"secrets that the market makers don’t know."
Like Merrill, Bear Stearns and others can’t afford to spend the $29 for the book that Joe New Trader can buy at Borders?
"The sad thing is, the hype and attention people give to tech indicators. There are a stream of articles and tutorials in these trade mags and on the internet, as if they are can’t-miss."
On the one hand I can understand this: sometimes a small edge can make a big difference. OTOH, I’m acquainted with one of the market wizards. He charges $500 an hour or a $2500 retainer to become your trading coach. He once said: "After I’ve told them everything, I don’t know what else to tell them, so I just make stuff up…"
Still though you can’t just blame the Wizard or the vendors: if there is a
market for pies in the sky someone will sell those pies. Or as the saying goes: you can’t cheat an honest man.
So, in conclusion, technical analysis tends to be useful hype.
Are we back to square one yet?
"So, in conclusion, technical analysis tends to be useful hype.
Are we back to square one yet?"
Some TA tools ARE useful, but more than likely no one will sell it to you for
$19.95 + S/H or give it away as the default setting on a free website.
Correct – and one should keep in mind that said tools are only useful under the market conditions that they’re meant to analyze. Using a screwdriver as a level will likely result in a crooked project.
Too bad so many TA systems over price themselves purposely, knowing that many traders associate price with quality.
"Too bad so many TA systems over price themselves purposely, knowing that many traders associate price with quality."
Oh yeah, that too! There is the famous W.W. (RSI inventor) story about the “Delta Secret” (I think that was what it was called). It went on the market for $125,000, then $35,000, then $999, then $99, and finally $35 or something like that. My details are surely off, but someone showed me a copy of it several years back, it wasn’t worth a dime.
And as I said: "a $2500 retainer to become your trading coach. He once said: “After I’ve told them everything, I don’t know what else to tell them, so I just make stuff up…”
Reminds me of the guy who invented Fixed Ratio - Ryan Jones, I think?
I kept getting his spam emails for about 5-6 years, selling variations of advisories (PowerTrade and others), and always offered you the "chance" to buy the secrets to Fixed Ratio, usually for about $3000.
Too bad it has been available on the net for years, for free…
Given the interest in forex, I extracted the following technical analysis indicators on the Euro (the commodities version).
Trend Spotter ™ 9 Trades 45 Days/Trade 5762.50
7 Day Average Directional Indicator 56 Trades 10 Days/Trade -13162.50
10 - 8 Day Moving Average Hilo Channel 63 Trades 7 Days/Trade -2287.50
20 Day Moving Average vs Price 51 Trades 10 Days/Trade -8800.00
20 - 50 Day MACD Oscillator 8 Trades 59 Days/Trade -4237.50
20 Day Bollinger Bands 23 Trades 3 Days/Trade 2700.00
40 Day Commodity Channel Index 22 Trades 9 Days/Trade -3425.00
50 Day Moving Average vs Price 17 Trades 28 Days/Trade 3587.50
20 - 100 Day MACD Oscillator 4 Trades 100 Days/Trade 8937.50
50 Day Parabolic Time/Price 37 Trades 15 Days/Trade -26812.50
60 Day Commodity Channel Index 19 Trades 10 Days/Trade -4775.00
100 Day Moving Average vs Price 12 Trades 33 Days/Trade 4037.50
50 - 100 Day MACD Oscillator 4 Trades 100 Days/Trade 26500.00
Overall Performance Summary 25 Trades 15 Days/Trade -921.15
Ross – I’ve found that in life, people would rather believe their own opinions rather than the facts presented to them. In fact, you might be viewed as rude for interrupting this intellectual discussion with quantitative evidence. How dare you blind us with knowledge.
So true.
Somehow, I found trading became a lot more successful, after demanding evidence-based advisories.
What you have said about willing to spend a couple of thousand a month for a profitable, stable advisory is true. It is worth its weight in gold.
As I said once before, I read Futures, TechAnal of S&C, SFO & other trade journals with a knife. When the authors tout the value of ADX, RSIs, MAs and other indicators, I study whether they either give a few examples or a year plus worth of backtesting to prove it. "example" articles are cut & put into the trash. "evidence/backtesting" articles go into a pile for further study. I refuse to trade based on what seems right vs. what works.
Deliver us from Elliot Wave, Gannsters and Fibonacci preachers. In fact, I remember several well-done articles that showed no evidence that 62% or other Fibonacci levels were any better than random levels, and that Candlesticks demonstrated little evidence of value. Yet the books, advisories and others still swear by them. I buy the concept that some things like breakouts and following existing trends have value.
I am a strong believer in the updated Random Walk theory - that markets are "random" but with a drift (trend) component. There are some seasonal and intermarket effects as well. But none of this is easy, as the fact that there appear to be far more active Collective2 advisories under the 100K mark than above it
"I am a strong believer in the updated Random Walk theory - that markets are “random” but with a drift (trend) component."
If you believe in Random Walk, why do you trade? Why would you bother to spend so much time and effort here? Why are you going to offer your own system?
"But none of this is easy, as the fact that there appear to be far more active Collective2 advisories under the 100K mark than above it."
No doubt, it isn’t easy, but the fact that most people lose here doesn’t
prove the random walk. You mentioned the Market Wizards. When Marty Schwartz averaged 800% a year, over a ten year period, in real money public contests, he made more money than all of the other contestants combined. This 60%+ per month average was done with a worst drawdown of less than 3%. Someone once suggested as you do that this is just random luck, that .00x% of swimming pool contractors find gold while digging holes for swimming pools. Well Schwartz “struck gold” in 19 pools out of 20 over a ten year period. Do you really believe this is random? BTW, he started making money (he lost for ten years) when he switched to technical analysis. On the top of his tech rules is simply comparing the market to a moving average.
Another example from wizards is Seykota. He has a documented real money account that made 250,000% from 1972 to 1988 on a cash to cash basis (several MILLION% accounting for withdrawals). His original system was an exponential MA system. Please see:
http://www.autumngold.com/Performance/12%20Month%20Rankings.pdf
These guys are Ed’s younger students:
Dekker Capital Mgt - Global Diversified 3X
"Deliver us from Elliot Wave…"
Dreiss Research Corp - Diversified claims to use “fractals” which may or may not have anything to do with Elliott specifically, but the Elliott theory is all about fractal progressions.
That system shows a real money net of 111% in 12 months vs a 2.5% drawdown. How can you make 111% against 2.5% risk in a random market, or with a slight “drift”?
I don’t “view (you) as rude for interrupting this intellectual discussion with quantitative evidence”, but your “quantitative evidence” doesn’t really carry any weight. Do you know what an ADX measures? I’m not sure why anyone would ever suggest an ADX system in and of itself could ever make money. This would be like putting spark plugs, tools, and a car in the garage overnight, and expecting to wake up with a tuneup.
Um, didn’t work? Let’s put more tools in the box and see if we can get a tuneup tomorrow night…
The fact that some (six) of the indicators you show, using nearly random default settings, make money surprises me. That the sum of these indicators looses money shouldn’t surprise anyone. Nor should we be surprised that free articles in free magazines offer zilch. It often takes years for brilliant minds to come out with a viable system, and then even more time to learn how to follow it. Why expect to find it sitting on the coffee table in the waiting room at Les Schwab tires or Super Cuts?
Do you expect to “lose twenty pounds and get six pack abs in 14 days” by reading the article in the magazine next to T&A of S&C?
Anyway, I’ll repost my “ADX, Moving Average” system results here. Yes, it uses near standard ADX and Moving Averages. No, it’s not that easy, there is a little more involved. But, it is really very simple. Just a few lines of non-optimized code. I hope you’ll pardon the interuption and not think I’m being rude.
Here you go. S&P daytrade. Single unit. 5 point stop. Assume $10K
margin. DD is near 1% of gross profit. Not a “back test” as this variation has been live for a few years. Profit, DD, win % is about the same long or short.
TheTruth SP-60 min 01/22/2002 - 06/05/2006
Performance Summary: All Trades
Total net profit $ 524250.00 Open position P/L $ 0.00
Gross profit $ 811500.00 Gross loss $-287250.00
Total # of trades 1445 Percent profitable 77%
Number winning trades 1119 Number losing trades 326
Largest winning trade $ 10850.00 Largest losing trade $ -1250.00
Average winning trade $ 725.20 Average losing trade $ -881.13
Ratio avg win/avg loss 0.82 Avg trade(win & loss) $ 362.80
Max consec. winners 31 Max consec. losers 3
Avg # bars in winners 2 Avg # bars in losers 1
Max intraday drawdown $ -5950.00
Profit factor 2.83 Max # contracts held 1
Account size required $ 15950.00 Return on account 3287%
Performance Summary: Long Trades
Total net profit $ 263225.00 Open position P/L $ 0.00
Gross profit $ 407525.00 Gross loss $-144300.00
Total # of trades 738 Percent profitable 77%
Number winning trades 571 Number losing trades 167
Largest winning trade $ 10850.00 Largest losing trade $ -1250.00
Average winning trade $ 713.70 Average losing trade $ -864.07
Ratio avg win/avg loss 0.83 Avg trade(win & loss) $ 356.67
Max consec. winners 26 Max consec. losers 3
Avg # bars in winners 2 Avg # bars in losers 1
Max intraday drawdown $ -7250.00
Profit factor 2.82 Max # contracts held 1
Account size required $ 17250.00 Return on account 1526%
Performance Summary: Short Trades
Total net profit $ 261025.00 Open position P/L $ 0.00
Gross profit $ 403975.00 Gross loss $-142950.00
Total # of trades 707 Percent profitable 78%
Number winning trades 548 Number losing trades 159
Largest winning trade $ 9100.00 Largest losing trade $ -1250.00
Average winning trade $ 737.18 Average losing trade $ -899.06
Ratio avg win/avg loss 0.82 Avg trade(win & loss) $ 369.20
Max consec. winners 19 Max consec. losers 4
Avg # bars in winners 2 Avg # bars in losers 1
Max intraday drawdown $ -9175.00
Profit factor 2.83 Max # contracts held 1
Account size required $ 19175.00 Return on account 1361%
Here’s the previous 4 years:
TheTruth SP-60 min 01/22/1998 - 06/05/2002
Performance Summary: All Trades
Total net profit $ 681675.00 Open position P/L $ 0.00
Gross profit $ 996050.00 Gross loss $-314375.00
Total # of trades 1393 Percent profitable 80%
Number winning trades 1111 Number losing trades 282
Largest winning trade $ 15025.00 Largest losing trade $ -1250.00
Average winning trade $ 896.53 Average losing trade $ -1114.80
Ratio avg win/avg loss 0.80 Avg trade(win & loss) $ 489.36
Max consec. winners 26 Max consec. losers 3
Avg # bars in winners 1 Avg # bars in losers 1
Max intraday drawdown $ -6550.00
Profit factor 3.17 Max # contracts held 1
Account size required $ 16550.00 Return on account 4119%
Performance Summary: Long Trades
Total net profit $ 337550.00 Open position P/L $ 0.00
Gross profit $ 494800.00 Gross loss $-157250.00
Total # of trades 652 Percent profitable 79%
Number winning trades 513 Number losing trades 139
Largest winning trade $ 15025.00 Largest losing trade $ -1250.00
Average winning trade $ 964.52 Average losing trade $ -1131.29
Ratio avg win/avg loss 0.85 Avg trade(win & loss) $ 517.71
Max consec. winners 19 Max consec. losers 3
Avg # bars in winners 1 Avg # bars in losers 1
Max intraday drawdown $ -8275.00
Profit factor 3.15 Max # contracts held 1
Account size required $ 18275.00 Return on account 1847%
Performance Summary: Short Trades
Total net profit $ 344125.00 Open position P/L $ 0.00
Gross profit $ 501250.00 Gross loss $-157125.00
Total # of trades 741 Percent profitable 81%
Number winning trades 598 Number losing trades 143
Largest winning trade $ 10900.00 Largest losing trade $ -1250.00
Average winning trade $ 838.21 Average losing trade $ -1098.78
Ratio avg win/avg loss 0.76 Avg trade(win & loss) $ 464.41
Max consec. winners 25 Max consec. losers 4
Avg # bars in winners 1 Avg # bars in losers 1
Max intraday drawdown $ -6900.00
Profit factor 3.19 Max # contracts held 1
Account size required $ 16900.00 Return on account 2036%
