Discretionary Trading vs Automated Trading

[syswizard 10]

You rarely sell at the ask or buy at the bid. Certainly nothing you can depend on. This has to be taken into account when programming. Tradestation/Easy Language does not take this into account, you need to program it yourself. Tradestations ProftTarget strategy will fill limit orders on a touch which is not realistic at all.

I disagree here....for one, e-mini futures fills with limit orders (buy@bid,sell@ask) have been filled about 80% of the time. However, my experience with Nasdaq stocks was about the opposite.

For another, one thing to do is to backtest with the Limit orders, and then in real live trading, make them MIT orders.....most platforms are supporting those now.

[torero 10]

80% of the time pretty much screws the reliability of the backtest results wouldn't you say? You want to test as close to 100% real world environment as possible. Those 20% could well mean all the winning trades that never happen. IMO I wouldn't even continue testing if I cannot find a 99% of the results resembling live trading.

[DrEvil 10]

With the advancement in technology and improvements in trading softwares, many intraday traders are now automating their trading. There are advantages to this... emotions can be eliminated (not completely but more than discretionary) and traders can use mechanical exit points. Example: scaling out at +10pts on every trade.

 

As a pure discretionary trader, I still believe there is an edge in discretionary trading. Intuition comes into play as well as better entries. Why better entries? If one is a good tape reader he can pinpoint his entries and take minimal heat. This can save a trader a ton of points in the long run.

 

I have a couple questions regarding this topic and could use your guys opinion. Do you think discretionary traders enter before automated traders? What are your thoughts on discretionary vs automation? What type of trader are you and which one do you prefer?

 

 

 

 

 

 

 

Soultrader, I'm relatively new to trading (less than a year). I have made a decision to be a discretionary trader. I trade forex, using only price, candlesticks, support and resistance. Looking for breakouts of channels, consolidations and trendlines on longer term charts, then taking the signal in same direction on an hourly chart. More or less this is my style of trading so far.

 

I read everywhere that backtesting is essential to succeed. As I mostly trade off price patterns, how can I backtest? If I can't backtest how can I put together statistics like, percentage winners, losers, drawdown etc.

 

Any answers greatly appreciated.

 

Thanks,

 

Soultrader

[wsam29 10]

I'm not Soultrader but I will explain this thanks to Jason Alan Jankovsky's book.

 

Volume is what drives prices higher or lower. Volume is caused by some other trader who is either getting into or out of a trade.

 

Spot FX being an unregulated market does not have a centralized exchange to post volume. eSignal provides volume to their FX data.

 

Price patterns alone do not tell the full story, what you need to be aware of is the volume in relation to the price pattern to get a higher probabiltiy of one thing happening more than the other.

 

I have tried trading FX just off price patterns because "thats all the FX market has to look at." Lets just say I am not a big fan of the FX market, not because I cannot trade it but rather the lack of infomation.

 

I have found that understanding trading psychology and behavior yields much greater results than relying on technical patterns and indicators to decide what course of action or inaction to take.

 

The way to see that behaviour is through the analysis of just price and volume.

 

There is alway some discretion to your trading, you still need to decide to take the trade or not when you think about it.

[Soultrader 12]

Soultrader, I'm relatively new to trading (less than a year). I have made a decision to be a discretionary trader. I trade forex, using only price, candlesticks, support and resistance. Looking for breakouts of channels, consolidations and trendlines on longer term charts, then taking the signal in same direction on an hourly chart. More or less this is my style of trading so far.

 

I read everywhere that backtesting is essential to succeed. As I mostly trade off price patterns, how can I backtest? If I can't backtest how can I put together statistics like, percentage winners, losers, drawdown etc.

 

Any answers greatly appreciated.

 

Thanks,

 

Soultrader

 

Hi DrEvil,

 

I trade discretionary but I have rules for each trading setup. I dont use a backtesting software though. What I do is keep track of each setup on a spreadsheet everyday and keep up with the data. If I see some setups failing to work after a while, I will observe it and see how I can improve it. It would be pretty much impossible to backtest most of my setups as I rely heavily on internals and tape. I also like to think in terms of market concept such as asking Why or What.

 

Soultrader

[DrEvil 10]

Thanks for the the answers guys. I'n my limited experience I find that on daily forex charts, a combination of classic price patterns and candlesticks are quite reliable. The problems I have are mainly trying to day trade by using hourly charts after receiving a signal on a daily chart ... the patterns certainly don't behave as well intraday. I am not at this point trading big enough to trade off daily charts. I find that I get a lot of information from looking at daily closing prices, which I assume filters out the day trading noise. For example if I can get a breakout of an congestion area on a daily chart, confirmed by a close on the daily bar, then I will always take a hourly trading on ideally a flag or pullback to a trend line, or m.a.

 

I agree that lack of volume info is a handicap, but isn't it a handicap for everyone? ... thus leveling the playing field?

 

I wasted months trying to figure out indicators and since I have removed them and focussed on price, I think I have raised my game. I hope I am on the right path.

 

Good trading!

[wsam29 10]

The problems I have are mainly trying to day trade by using hourly charts after receiving a signal on a daily chart ... the patterns certainly don't behave as well intraday.

 

I find that I get a lot of information from looking at daily closing prices, which I assume filters out the day trading noise.

 

I agree that lack of volume info is a handicap, but isn't it a handicap for everyone? ... thus leveling the playing field?

 

I wasted months trying to figure out indicators and since I have removed them and focussed on price, I think I have raised my game. I hope I am on the right path.

 

Good trading!

 

You're helping (analyzing) yourself already and you don't even know it.

 

Day trading may not suit your personality. Don't force a style of trading onto yourself if it is not working. If you lose money, stop doing it. Remember what I said before, "in order to make money, you need to stop losing money" or somewhere along those lines. How you come to terms of stopping the "harmful" actions is up to you, but that is the key to being a net winning trader.

 

Level playing fields in FX, I think not, there are some who do get FX volume, you can spot where the "big money" is coming into the market to take your money. All volume does is confirm if there are players or not because if you get more volume in your direction, you have a pretty good chance your trade will work out. It's not a shoe in, but you will have the edge.

 

I am so glad to here you junked the technical indicators that are mathematical based. You definately are on the right path.

 

Just remember you will not succeed at this game if you quit.

[torero 10]

I think there is never level playing in every market you trade in, i.e. the locals will almost always be one step ahead of you, the market makers will always be one step ahead of you. Given, that forex is the same way. Volume is important but somehow I've managed to stay profitable without volume. One change or adaptation I've made from futures to forex is I scrutinize price action more, be more picky with S/R areas and relied on Fibs, Pivots more to really make sure the setup is overwhelmingly in my favor. Basically I made sure the breakout is TRULY confirmed before making that entry and stay away from ranges, trend-following specifically. All this without volume, a miracle in some ways. The other observation I've made is that my R/R ratio is high 3:1 minimum or I don't trade. Surprisingly this is what has made the equity curve climb. I've managed to get many 3:1 wins when it was a winning trade.

[TinGull 10]

I'm about to embark on implementing a semi-automated way of trading for myself. Basically, I need things to be as mechanized as I can make it, or else I may not take certain trades. My trades will all be based on Market Profile, and I'll basically just make my calls pre market, and setup my working orders so they're out there to get filled, and the odds are in my favor that the market will do as I plan on. Based on my analysis of the market I trade over the past while, the setups are pretty well defined. When the first setup fails and I lose some points, then I know right where the next setup will be to get that lost profit back and then some.

 

That's basically how I will be doing it for this entire week, and if it works in my favor, then I'll be doing that from here on out. I'm basically posting it up here so I have some accountability and know I need to follow through with it, instead of sayin...ah, nevermind. I can go purely discretionary since it's not making me much money, and thats what I love, right? not making money? ummmm, nope.

[Nick1984 10]

Well I would assume that most discretionary traders are such because they either a: dont know how to code and b: they love to have the "hands on" approach to their trading.

 

If you love to trade why would you take away the enjoyment of it through using an automated system?

 

Soultrader asked before what would become of discretinary traders as more and more ppl become reliant on automatic systems. Well I may not be a coding genius (i dont know diddly by the way) but i would assume that these systems are all reactinary in nature i.e. they react to what the market does. I wouldn't think an automated system could ever make the market move! If everything was automated 100%, where would the momentum in a market come from?

 

I migth be wrong about my analysis of automated sytems but in the end I think that yes there are still more discretinary traders and these people are the real market makers.

 

If you love to trade, and love the wheeling and dealings of the markets why would you contemplate giving that up in the first place?

 

Good luck on your trading mates!

[notouch 10]

I read somewhere that most of the liquidity on the NYSE comes from automated systems and the percentage is growing on Globex and Eurex. That doesn't mean the market will be untradeable for discretionary traders although the nature of the market will change as it always does.

[Nick1984 10]

really? Where did u see that?

 

I must not be able to understand how these automated systems work to well, cause I can only conceptualise them as bein 100% reactionary to what the prices are doing. In a market with low volume and price basically flat how would an automated system get the ball rolling?

 

If these systems are profitable in the long run, it's pretty interesting and I'd like to learn more about how one would go about setting one up especially if one is coding illiterate!

[Nick1984 10]

Another pertinent question that could be asked would be: Is automated trading likely to overtake discretionary trading in the future or make it obsolete?

 

Personally I can't see that hapenning. Very "terminator" to me.

[drk1t 10]

Sounds like some of you are looking for the holy grail of trading again. I'd love to have a box (software) that you just turn the crank and out comes the cash but it hasn't happened yet and I don't think it's going to. The markets change to much. What moves it today isn't going to do it in the same amount and the same way tomorrow. All I see in this magic software is another level of work and confusion. Instead of using our human brain for basic primary decisions I see it being used for making decisions of factors that may or may not make a difference, and those factors my change from plus or minus and in varying degrees. Gosh, I can programs to make programs just to establish values to put into trading software. The market is the market because of the human nature, when you can plug in all 8 billion people and their thinking at the time automated systems might have a chance. All this said, if anyone has one that works and will always work, let me know! I'd love to go drink beer and know the old computer is cranking out cash. Right now I'd settle for a program that would automatically do my taxes, use creative credits and deductions, print out receipts for them and guarantee I'd stay out of trouble with the IRS.

[Nick1984 10]

amen brother

[nelo 10]

Nick1984, every system is 100% reactionary - you always react to something, otherwise it is random. Not all automated systems are price-based. Reuters (an I expect other institutional data providers) already feed their clients with fundamental data in an "algorithm-friendly" form. So automated systems can process reports generated by FED and other agencies. Goldman-Sachs is throwing out 60+ traders and replacing them by automated systems. What else is necesary to say?

 

Personally, I am both automated system trader and discretionary. Both, if done right and after spending enormous amount of effort can be profitable. Single whatever great automated system is always less profitable than good discretionary trader due to timing (and I do not accept the graph posted here because the guy was fooling himself by buying at bid, selling at ask, if not mentioning slippage). On the other side, you can auto-trade several systems at one and fully utilize kelly ratio which discretionary trader cannot. In the end you end up with similar results. No way is easier or better than the other, it is about what suits you more.

 

Nelo

[syswizard 10]

How would one go about programming tape reading? Pretty impossible I'd say.

Penn-Lehman SOBI ATS project

Never say never.....the big brokers are into this stuff in a huge way (have you seen their earnings recently ?)

The smart kids at Penn/Wharton came-up with a pretty novel approach to reading the order book and this strategy (after modifications of course) eventually went live at Lehman Brothers.

[brownsfan019 17]

Interesting thread...

 

As I mentioned in another post, I suppose I am a discretionary automated trader... In other words, I am discretionary b/c I am pulling the trigger myself and automated b/c I simply follow my rules.

 

In my time as a trader and broker, I have never seen a successful automated system that can stand the test of time. I won't argue that some systems are very profitable in a given period of time (but who's system can't do that, right?). The key in my opinion for the automated traders is knowing when to turn your system 'off' or when to leave it alone. It's easy in hindsight to say that based on your backtesting, you would have turned it off at XYZ point, but making that call in real time is not so easy. If you turn it off and you simply were in a small drawdown, you lose. If you let it run and turn your account into half in one month (like the blog post pasted here), you lose. And as that blog post said, in essence, no matter what - your emotions can/will drive your ultimate decisions even if you are a 'hands off automated' trader.

 

And of course the devil's advocate in me says that if there are successful automated traders out there, you won't find them on the internet message boards. I think they in fact have too much to lose by even hinting to their setups. I can post my charts all day here and each person would interpret them differently. I post a trading code and everyone has the exact same thing. I would think that's a potential problem...

[notouch 10]

Automating tape reading would be pretty straight forward. What are you looking for when you're tape reading? Large orders hitting the bid and a fast moving tape to give you a buy signal (in combination with other aspects of your system of course)? That could easily be incorporated into an automated system. I don't think there's any aspect of a good discretionary system that couldn't be automated.

[syswizard 10]

I don't think there's any aspect of a good discretionary system that couldn't be automated.

Hmmm...you sure about that ? You can program the emotional and psychological aspects of market behavior ? Also, your program can anticipate and interpret the "news" events ?

Omigosh, we are at least 10 years away from software being able to do this.

If we were "there" today, IBM's Big Blue program that crushed the best chess players would be applied to the markets. Guess what ? They tried it and it FAILED MISERABLY.

That was at least a 5 MAN-YEAR software effort...probably much more.

[brownsfan019 17]

I don't think there's any aspect of a good discretionary system that couldn't be automated.

 

Notouch - I think some aspects would be very difficult to code. For example, when trading candlesticks as I do, I can easily code what a hammer is by textbook definitions. But in real time, I am much more lenient on what I call a hammer. In intra-day trading, I think you have to be. So, while an automated system may be looking for picture perfect hammers, I may have taken other trades simply by 'eyeing' the charts up. And I wouldn't call one hammer better than the other, so to bypass the non-textbook hammer is a big mistake in my opinion.

 

I think it really boils down to what you are using in your trading. As you said, tape reading may be easily put into code. I have no idea if that's true, but it sounds like it could be. Your textbook candlesticks can easily be put into code as well, but that is one part of trading candlesticks in my opinion. A hammer that appears in chop is nothing to me.

 

I would love the ability to code my trading system in such a way that I know it will do what I can do on my own, if not better. And since I am unable to do this myself, there is strong hesitation to try to find a programmer that is willing and able to do it.

[Soultrader 12]

Automating tape reading would be pretty straight forward. What are you looking for when you're tape reading? Large orders hitting the bid and a fast moving tape to give you a buy signal (in combination with other aspects of your system of course)? That could easily be incorporated into an automated system. I don't think there's any aspect of a good discretionary system that couldn't be automated.

 

I highly disagree and have to say it is impossible to code such a thing. Tape reading is not all about watching for big lots come across the tape. That is just the basics of it. I would love to elaborate on tape reading but I find it extremely difficult to explain in words. If everything can be automated... why is it that the majority of successful intraday traders are discretionary traders?

[walterw 10]

Market Climate ¡¡¡ Market Climate ¡¡¡ my dear fellow traders.... only a human trained neural net is in conditions to understand it.... sistems and computers will miserably fail.... lol and lol... there is no greater pleasure than to feel the market... once you feel it (click) cant live without it... its almost adictive... cheers Walter.

[notouch 10]

In reply to syswizard: "emotional and psychological aspects of market behavior" are presumably reflected in the charts and tape or how else would a discretionary trader see them? So if they're reflected in the charts and the tape then they could be automated. And yes it is possible to analyse and interpret news events. In fact it's extremely simple to code an Excel macro that enters an order on the basis of an economic data release (released real time by Bloomberg and Reuters). Reuters, Bloomberg and Dow Jones also use XML tags that enable automated systems to react instantaneously to news events. Why do you think the market reaction to news events is instantaneous? Because automated systems are reacting within milliseconds - much faster than any discretionary trader could react.

 

Brownsfan: you could be as flexible in coding an automated system as you are in implementing a discretionary system. What are you looking for in interpreting your candles? For price to swiftly reverse on high volume? That can be coded into an automated system.

 

Soultrader: I'm sure it would be difficult to code a system that automates tape reading but definitely it's possible. I personally don't think the majority of successful intraday traders are discretionary traders. I think the majority of successful intraday retail traders are discretionary traders because it takes a high level of programming skills to write a successful automated system. I think the majority of intraday trades are executed by automated systems run by investment banks and other institutional traders. Definitely successful institutions are moving away from discretionary traders toward automated systems.

 

Just to clarify, I'm personally a discretionary trader but only because I don't have the programming abilities to put all my trading rules into an automated system.

[TheBramble 10]

[if] I post a trading code and everyone has the exact same thing. I would think that's a potential problem...

But that's the thing. The major charting software vendors do exactly that. They sell their product with default settings for all the indicators and they pretty much all set them at the same levels.

 

Guess what, most traders will take these packages on-board and will rarely think to examine the defaults or change them.

 

It's a potential problem all right, but only for those that don't think outside the shrink-wrap.

 

If you posted a piece of code I guarantee most would use it 'as is' without thinking to modify or test modifications of it. And you could then work out a pretty useful fade to what you'd supplied.