NexusFi: Find Your Edge


Home Menu

 





Use your brain! Mechanical indicator-based systems don't work


Discussion in Traders Hideout

Updated
      Top Posters
    1. looks_one cunparis with 17 posts (63 thanks)
    2. looks_two trade2win with 16 posts (3 thanks)
    3. looks_3 Andrew with 14 posts (6 thanks)
    4. looks_4 Silvester17 with 11 posts (4 thanks)
      Best Posters
    1. looks_one TAJTrades with 17 thanks per post
    2. looks_two FBJS with 5.3 thanks per post
    3. looks_3 Big Mike with 3.8 thanks per post
    4. looks_4 cunparis with 3.7 thanks per post
    1. trending_up 58,086 views
    2. thumb_up 177 thanks given
    3. group 32 followers
    1. forum 111 posts
    2. attach_file 2 attachments




 
Search this Thread

Use your brain! Mechanical indicator-based systems don't work

  #1 (permalink)
 
cunparis's Avatar
 cunparis 
Paris, France
 
Experience: Advanced
Platform: Market Delta & Ninjatrader
Trading: ES
Posts: 2,565 since Jun 2009
Thanks Given: 1,162
Thanks Received: 2,093

In my Learning to use Volume thread, I noticed a lack of participation when I asked people to look at a chart and analyze the volume. I suspected that everyone was waiting for a volume indicator that is currently under development, one that will analyze the volume for them. So then I started writing a little rant about relying on indicators and I thought it would interest everyone so I decided to create this thread.

The majority of posts in these forums are from people looking for indicators that tell them when to enter and when to exit. I do not believe that such a system is possible and have never seen one that that works mechanically. When I say "mechanically" I mean without reading price & volume in a non-mechanical way.

The only way I have been consistently profitable in trading is by reading price & volume. If there is an indicator which can consistently tell me when to enter and when to exit then I would love to see it. If it works then we can all stop learning about reading price & volume and just run an automated strategy and fill our bank accounts while laying on the beach.

This reminds me of a story because I really did run automated systems while I was at the beach!

When I first got Ninjatrader I went to work on a mechanical system. After months (and hundreds of hours) I finally got something that were good. I forward tested for several weeks and results were the same as my backtest (very good). So I started with real money. It just happened to be the week we were on vacation in Orlando. On my 2nd day of real money trading we were at Disneyland. I came home and found out I made $700 while we were out riding rides! I thought I found the holy grail! And so I ran my system. It continued working. So I made another. I spent hundreds of hours on these two systems and they were doing great. Back in Paris I was working my real job and I'd get an email now and then saying I won or lost some money. It was great.

Then one day I had a couple losses. Bigger than usual. Ouch. No big deal, just a little drawdown. I had some wins and some more losses but losses were bigger than wins. I looked at the equity curve. Big drawdown. Both systems. I decided that I had to keep going. More losses. I pulled the plug. Then guess what? They started winning again!!! Ok I learned my lesson, I turned it back on. And you can guess what happens. More losses. More drawdown. Do I pull the plug? Is it not working properly? In? Out? It was all very stressful.

After trading with real money for around 4 months I had given back all my gains and some of my hard earned money!

What happened? The market changed, that's what happened! subtle changes, increase in volatility, going from bull to bear market, leadership changes amongst sectors, news, you name it.

Out of curiosity i just ran some backtests on my two systems for the past few months. For CL it fails miserably. For ES it failed most of the year but has a PF of 1.5 for the past few months.

So here's the deal: Basic human emotion doesn't change. The trading books written back in the early 1900's (Wyckoff, Livermore), mid 90's (Darvas) are just as true now as they were back then when computers didn't even exist.

But the fact that human emotion (fear & greed) doesn't change doesn't mean the markets don't change. They change all the time. Constantly. High Frequency Trading? Just another change in the system, one that had a dramatic increase in volume. Things like HFT are changing all the time, but so slowly that you don't really notice. Unless you've written a mechanical strategy that is.

I will add another experience. I figured my error was in using price data for my systems (bollinger bands, RSI, macd, etc.). So I thought that if I used breadth information (TICK, TRIN, Adv/Dec, PCRatio, etc.) that a mechanical system would work. So I spent hours working on a new mechanical system. This time I read Pardo's excellent book ( Amazon.com: The Evaluation and Optimization of Trading Strategies (Wiley Trading) (9780470128015): Robert Pardo: Books ) and did forward testing (btw that link is to the most recent version, I have the previous version which is out of print, I bought it used on ebay. I appears that the old version may be better, as in the new one he pushes his wares).

Where was I? Oh yeah, walk forward testing. I posted a little about this here:



What I found was if I optimize for the first 3 years (2000 - 2003), it walked forward 1 year very well. Then optimized 2000-2004 and walked forward to 2005 it worked. Cool right? Not so fast! In 2007 it didn't walk forward well. Why? Could be we went from bull to bear market. More volatility. More volume. Who knows. Then I had a problem. I couldn't optimize a period that had both bull & bear periods, and those were the only ones I could identify and easily quantify, we can pick a date to change from bull to bear (trendline break usually). Higher volume is not easy to quantify. Or increased volatility. Or any of the other factors that change in the market. So optimizing 2007-2008 didn't give enough trades to be reliable, and didn't walk forward well either.

And oddly enough the walk foward for the past month has been in drawdown, despite a steady linear curve during the backtesting period!

I mention this because I'm not even using price and it still has problems. Imagine using price to predict price, it'd be even worse.

So my opinion is that indicators just complicate things and if you optimize them (using RSI(10) instead of RSI(14) or use threshold 25 instead of 15) then they fall apart. Use two moving averages and you're history. And lots of these indicators have 3 parameters. The more parameters the more likely the optimal parameters will constantly change.

Look at the turtle system. If you're not familiar with it I highly recommend the book, and he talks about optimization and curve fitting as well:

Amazon.com: Way of the Turtle: The Secret Methods that Turned Ordinary People into Legendary Traders (9780071486644): Curtis Faith: Books

Just watch out cause he's selling his software to help you test your system. So he's not likely to say it'll never work.

So the turtle system made millions. And then guess what happened to it? By now you should be able to guess, yes it stopped working. The most famous system in history stopped. Now you may be thinking that it worked long enough to make millions so it's still worth the pursuit, but they had to accept huge drawdowns to get those millions. Could you take a drawdown of 50%?

Like in my example above, you will think the system stopped working and pull the plug, only to miss huge gains that make it profitable again. OR you won't pull the plug, you'll be patient. In which case the system probably stops working and you loose everything! I don't think any human can tolerate that. too much stress.

So what is the answer? In my opinion one has to learn to trade based on price action, and if you choose, price & volume. When I say price action I am talking about price patterns (H&S, double top, etc.), support & resistance, trendlines, swing pivots, breakouts, etc. Volume can be an additional confirmation (or rejection) but isn't an entry signal in and of itself. And I'm warming up to the idea that volume isn't necessary to be profitable, but I still think can increase your edge.

Using a few indicators to rule out trades is ok, but I think entering on an indicator signal is not a good idea. For example you can say "I'm not taking this long cause there is bearish momentum on the macd". Or "the ema is up so I'm not going short". I think this is ok although I'm not sure it's necessary.

But if you say "the eco bar turned red I'm going short" or "all three of these are the same color I'm going long" I think it will not work. What's interesting is we can't prove that an indicator entry/exit system doesn't exist, but we can prove what that a given indicator entry/exit system doesn't work. So far I haven't seen anything that works.

Think about it, Goldman and the other large firms spend millions of dollars developing trading systems. They hire programmers and pay them astronomical salaries. You can read about the case where Goldman is accusing a trader of stealing their code when he left the company. If it were as simple as eco bars turning the same color as another indicator don't you think they would have found it????

Their current focus is on high frequency trading. They've found something they can do that a human can't. That is place and cancel orders in milliseconds. And that's their edge. If they could just buy and hold for 5-15 minutes and make money I think the millisecond HFT wouldn't be necessary.

But then again maybe Goldman's traders aren't as good as you and you've found the holy grail and can beat them. But I don't think so.

Stop the insanity! Keep it simple. Learn to trade with price & volume. Stop wasting time on indicators, backtesting, and optimizing.

I have some more ideas but this post is long and I want to get some feedback before continuing. So let me know what you think and let me know if you want to hear more. If there is interest I will tell how I lost $20k trading a leading guru's mechanical system.

Follow me on Twitter Started this thread Reply With Quote

Can you help answer these questions
from other members on NexusFi?
My NT8 Volume Profile Split by Asian/Euro/Open
NinjaTrader
NexusFi Journal Challenge - April 2024
Feedback and Announcements
ZombieSqueeze
Platforms and Indicators
Request for MACD with option to use different MAs for fa …
NinjaTrader
 
Best Threads (Most Thanked)
in the last 7 days on NexusFi
Retail Trading As An Industry
58 thanks
Battlestations: Show us your trading desks!
55 thanks
NexusFi site changelog and issues/problem reporting
48 thanks
What percentage per day is possible? [Poll]
31 thanks
GFIs1 1 DAX trade per day journal
29 thanks

  #2 (permalink)
 
Big Mike's Avatar
 Big Mike 
Manta, Ecuador
Site Administrator
Developer
Swing Trader
 
Experience: Advanced
Platform: Custom solution
Broker: IBKR
Trading: Stocks & Futures
Frequency: Every few days
Duration: Weeks
Posts: 50,322 since Jun 2009
Thanks Given: 33,143
Thanks Received: 101,477

I agree 1000% with every word.

Great post, thanks for sharing your thoughts.

I've been saying for a long time that automation doesn't work and indicators aren't the answer. But... it's like this: Water is better for you, but most people prefer a Soda. Just because something works better doesn't mean it is what people want

So I've been contributing to the indicator-industry and giving everyone their fix, but hopefully these same people are always watching the dozens of hours of videos I've recorded and read some blog posts and forum posts, and maybe in the end they realize indicators are not the answer and start to focus on what is --- a good plan, money management, and most important of all, you (the trader).

Mike

We're here to help: just ask the community or contact our Help Desk

Quick Links: Change your Username or Register as a Vendor
Searching for trading reviews? Review this list
Lifetime Elite Membership: Sign-up for only $149 USD
Exclusive money saving offers from our Site Sponsors: Browse Offers
Report problems with the site: Using the NexusFi changelog thread
Follow me on Twitter Visit my NexusFi Trade Journal Reply With Quote
The following 6 users say Thank You to Big Mike for this post:
  #3 (permalink)
 MXASJ 
Asia
 
Experience: Beginner
Platform: NinjaTrader, TOS
Posts: 796 since Jun 2009
Thanks Given: 109
Thanks Received: 800


I guess a good question is can a sucessful discretionary trading method be coded into a sucessful automated trading system... and at what point is one spending more time on coding and infrastructure then one is spending watching price action.

Over the past three, four months or so that I've been coding automated strats for ES, ZN, FESX, and FGBL, have I learned more about those markets and how they are traded or have I learned more about Ninja Trader?

The answer is what the answer is.

Reply With Quote
The following user says Thank You to MXASJ for this post:
  #4 (permalink)
 Eric j 
NY
 
Experience: Advanced
Platform: Esignal, MBT navigator
Trading: Currencies
Posts: 2,507 since Jun 2009
Thanks Given: 1,396
Thanks Received: 2,576

I agree 100 % with all this . Indicators only confirm and quantify price action . I feel that due to the lag inherent to all indicators you cant rely on them ALONE to make consistently profitable entry decisions .

Many of the indicators we use today are either based upon or are the exact indicators that were used and developed before the computer age and were designed to be applied to stocks held for long periods of time . They were computed on paper by hand using delayed data and daily close prices .

Since we know that its impossible to know what will happen next , anything can happen and that an edge only suggests we should act in one way or another , why would anyone believe an indicator, group of indicators, guru or anyone else can know these things ?

Visit my NexusFi Trade Journal Reply With Quote
  #5 (permalink)
 zeller4 
Orlando Florida
 
Experience: Intermediate
Platform: NT8
Trading: CL, NQ, ES, RTY
Posts: 477 since Jun 2009
Thanks Given: 1,416
Thanks Received: 404

thanks for sharing. I would never relish in another's demise or failings and i'm glad you're pulling above the losses of the past... I would be interested to read about the guru you'd followed with the mechanical trading system.

kz

Reply With Quote
  #6 (permalink)
 Eric j 
NY
 
Experience: Advanced
Platform: Esignal, MBT navigator
Trading: Currencies
Posts: 2,507 since Jun 2009
Thanks Given: 1,396
Thanks Received: 2,576

I cant say I've followed a guru but have definitely paid to be taught how to " trade my way to freedom " and " make 1000's every day "

I dont watch the news and havent seen cramer or CNBC in years but I really miss seeing Becky swift and that asian hottie whats her name ? MMMMM !

I know when news is released and check out CMEs site for info and thats the only influence I need .

Visit my NexusFi Trade Journal Reply With Quote
  #7 (permalink)
TAJTrades
Here, GA
 
Posts: 158 since Jun 2009
Thanks Given: 1
Thanks Received: 86

cunparis,

Very interesting perspective. A while back I wrote some software that would capture and log all the bigger trades. The idea was to get an estimate of what the Big Boys were doing, price points where they were buying or selling. Depending on the daily charts I could go back 1,2 3 or more days, sometimes weeks, to determine the zones that the they were buying and selling. It was a thing of beauty for finding Support and Resistance areas. I could see areas that were defended and areas of accumulation / distribution. Then slowly but surely it started to lose its effectiveness until it reached a point of useless. Then a few months later we all started to see the stories about High Frequency Trading. Totally skewed every thing I was doing and my little kingdom of the Holy Grail collapsed. Point of the story is that market conditions change all the time be it technology, economics or the heard mentality.

The only thing that saved my bacon was that I made a decision to spend time and effort on learning how to trade and not rely on my little "Gold Mine". As I have gained more experience and success the only thing I am becoming convinced of is that any combination of a Trend Indicator and a Momentum Indicator (Keeping it Simple) will lead to success if and only if you spend the time and effort to learn all the little nuances of those indicators and price action around them. Indicators are only guides, little hints of what might happen. Price action will tell the story.

If I were to make just three suggestions they would be: 1) Invest in a good printer, lots of paper and ink. Print out your charts every day and review those charts after the market closes. Refer back to yesterdays chart and last week , month, year. Learn the nuances inside and out. Trading is a full time job; treat it as such. 2) Pick one side of the market. Are you going to be a Trend Trader or Counter Trend Trader. Learn one side and learn it well. It will keep you in the game. 3) Stop chasing the next "Latest and Greatest Indicator". Have you ever talked to a successful trader and the tell you that they use indicators A and B. And then you talk to another successful trader and the tell that they are using Indicators C and D. Well how can that be? My guess is that they know the interaction of price with the indicators inside and out. There is a comfort and confidence level from using the same chart setup for a long period of time. Pick a chart setup and do not deviate from it for a year. The results may surprise you.

And as always (should not even need to be said) Money Management is King.

Good Luck.

Reply With Quote
The following 17 users say Thank You to TAJTrades for this post:
  #8 (permalink)
 
record100's Avatar
 record100 
Toronto, CA
 
Experience: Intermediate
Platform: NT
Broker: IB
Trading: YM, ES
Posts: 105 since Jun 2009
Thanks Given: 34
Thanks Received: 44

I still need to verify that money management could do the trick, and make even simple mechanical system profitable?
Diversification by the market, time frame or system?
Optimization, but only for the recent period?

Reply With Quote
  #9 (permalink)
 
Big Mike's Avatar
 Big Mike 
Manta, Ecuador
Site Administrator
Developer
Swing Trader
 
Experience: Advanced
Platform: Custom solution
Broker: IBKR
Trading: Stocks & Futures
Frequency: Every few days
Duration: Weeks
Posts: 50,322 since Jun 2009
Thanks Given: 33,143
Thanks Received: 101,477


cunparis
What happened? The market changed, that's what happened! subtle changes, increase in volatility, going from bull to bear market, leadership changes amongst sectors, news, you name it.

Out of curiosity i just ran some backtests on my two systems for the past few months. For CL it fails miserably. For ES it failed most of the year but has a PF of 1.5 for the past few months.

Playing devil's advocate here... not directed at you... I don't have the answers...

Let's say you have OBV or CCI or Stochastic on your chart and a big MA for some sort of resemblance of trend-finding. Let's say a 100 SMA or EMA.

When you say "the market changed"... did you find that when trading with the 100 MA and with OBV that it used to work but now it doesn't? If you used to be able to take only shorts when the 100 MA was falling, now can you only take longs?

That logic fails for me. Bulls/bears, sure. Who cares though. I'm trading for a handful of ticks. On the CL, my average trade is around 16-24 ticks. CL is trading around $70. 16 ticks on $70 is something near 0.25% (as in 1/4 of 1 percent). It's $10/tick so that's $160 to $240. Average time in trade: 3-5 minutes.

I don't care if we are in a bull market or bear market. Price is going to have no problems moving 0.25% in both directions. Why should I care what kind of market we are in. I am not investing. This is not my Roth IRA.

Now I think you are more of a swing trader. So obviously the longer you hold a position, the more risk, and the more a longer term trend (bull/bear) would have a direct impact.

When you say "the market changed" I am left thinking your strategy was curve fitted.

Let's talk money management. What kind of win percentage did you have? What was your avg win:loser in dollars? What kind of position sizing?

Mike

We're here to help: just ask the community or contact our Help Desk

Quick Links: Change your Username or Register as a Vendor
Searching for trading reviews? Review this list
Lifetime Elite Membership: Sign-up for only $149 USD
Exclusive money saving offers from our Site Sponsors: Browse Offers
Report problems with the site: Using the NexusFi changelog thread
Follow me on Twitter Visit my NexusFi Trade Journal Reply With Quote
The following 6 users say Thank You to Big Mike for this post:
  #10 (permalink)
 wuolong 
Canada
 
Experience: Intermediate
Platform: NinjaTrader
Posts: 76 since Jul 2009
Thanks Given: 175
Thanks Received: 9


Well said. I have been testing some mechanical systems recently, but all turned to losses in the EOD.

Reply With Quote





Last Updated on November 28, 2018


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts