Saturday, September 09, 2006

Blending System and Discretionary Trading

I've recently emphasized that the psychology of the marketplace is as important as the psychology of the trader. If you review my recent AM market updates, you'll see that the purpose of these is to track the market's psychology: the short-term sentiment of the large traders who account for only about 3% of ES trades, but 40% of its volume.

Two trading tools have very recently come out with new products that I'm working on integrating into my own trading (and also into the blog). The first one, Market Delta, is illustrated in my recent post on tracking volume and in my recent Trading Markets article. I find Market Delta to be most helpful in tracking order flow: the quantities of contracts traded, where they trade in the bid-ask matrix, where they trade relative to support/resistance/value area, and how these volume patterns shift over time. A review of the recent AM updates will illustrate this application.

The second tool is Odds Maker from Trade Ideas. To be honest, I've had a harder time integrating this tool into my trading. I'm not by nature a stock picker; I trade equity indices. I very much like using research into historical patterns to frame trade ideas, but have never gravitated to intraday research. Too much noise. As I've noted in my Trader Performance page, I've lengthened my time frame to focus on moves of several days rather than trade intraday patterns exclusively. That's given me the flexibility to do the market updates daily.

Still, Odds Maker is a powerful tool and not one that I wanted to ignore. What it does is go back three weeks (15 trading sessions) in time and scan for all occurrences of a pattern that you identify in the Trade Ideas program. It will then spit out a performance summary to tell you how many trades would have been made during that 15-session period, how many were winners and losers, the average size of winners and losers, and the net profit/loss trading that pattern. The performance summaries can be for the entire market, a selected basket of stocks, or an individual equity. User-controlled features help traders tweak entry and exit criteria and see how these impact profitability.

Major props to the Odds Maker programmers: the program accomplishes a tremendous amount of number crunching very quickly. Once you find a pattern that is promising, you can then track it in real time in Trade-Ideas to see how it performs. The goal, of course, is to not just curve fit patterns to the past 15 days, but find ones that are robust with respect to upcoming periods. (By running reports each day, you can update performance and see if your setups are stable or degrading).

The question, for me, is how do you integrate the Odds Maker setups with a reading of order flow? What I've done is develop Odds Maker setups that test well for equity index ETFs, providing trades lasting 30 minutes in duration. (My next post on the Trader Performance page will describe the setups more specifically). Once the trade is alerted by Trade Ideas, I then go into "order flow monitoring" mode. If the setup fits with the order flow patterns that I'm seeing with Market Delta, I go with the trade idea. If not, I pass it up. Similarly, if order flow shifts during the 30 minute holding period, I'll bail. If not, I'll use the time stop.

What this accomplishes, hopefully, is a combination of one source of edge with another, independent one. By trading a tested setup and then only following the setup when current market conditions are favorable, it may be possible to raise one's win/loss ratio and overall profitability. This blending of system and discretionary trading is what I pursue with my historical studies. Now, with Odds Maker, we can see if intraday historical studies can inform our reading of order flow.

The setups go live next week; stay tuned!


EddieFl said...

Hello Brett,

I stumbled on your blog the other night looking for people that trade using trending techniques, I actually saw some of your work on Michael Covel's site.

I will be getting back into trading full time in Mid-November so I am now researching some blogs that can add to my trading arsenal. Your blog is especially important, i have truly grasped the importance of my trading psychology/attitude that i need to possess to be a great trader.

So I'll be following along from now on. My questions what are the "setups that go live next week", thanks for your response in advance.

D TradeIdeas said...

I think you bring a sound premise to trading with this approach. Combining the best of several tools to mitigate risk and identify trades. I look forward to your results.

Additionally its good to know that others are looking for the kind of metrics that The Odds Maker uses in its summary. I am citing specifically a recent conversation on UglyChart's blog.

Brett Steenbarger, Ph.D. said...


Thanks for your note. The blog really has two components: one dealing with the psychology of the trader, the other with the psychology of the market.

A description of the setups from Odds Maker will be posted later this weekend to the Trader Performance page of my personal site ( and a summary will also be posted to TraderFeed.

Best of luck with your return to trading--


Brett Steenbarger, Ph.D. said...

Hey David,

Thanks for that link to the discussion on Ugly's site. It's an interesting one. For traders really serious about their trading metrics and the evaluation of their performance, I heartily recommend Trader DNA ( It provides a very well-rounded appraisal of a trader's strengths, weaknesses, progress, etc. The upcoming Trading Summit I'll be participating in (sponsored by the CME) will make use of Trader DNA--as do many professional traders.


yinTrader said...

Hi Brett

Through your posts and feedbacks from your readers, I am discovering more in terms of tools and metrics to add to my trading and learning curve, the latest being the Uglychart's blog. (pleasant to me).

Trading performances and trading success is best measured by
-win %(winning trades v losing trades)
-actual W/L ratio
-biggest loser v biggest winner
-expectancy (a derivative of the first two qualities)

And these qualitites should be measured over a big chunk of trades - IMO, 30 trades are a good start, but the more trades, the more reliable these results are, when focusing on measuring consistent success.

This is what my mentor taught me and to think in terms of R-multple to reach my 'grail'.

Brett Steenbarger, Ph.D. said...

Thanks, Yin; I think you're right: it's crucial to be able to think clearly about your trading performance, with a measurement of how much reward you're pulling out of the markets for the risk you're taking.