You are currently viewing How I Trade (Part 1 of 2)

How I Trade (Part 1 of 2)

Word Count: 1506
Estimated Reading Time: 7-8 minutes

While I’ve written extensively on trading approaches and market movements, I still get many questions like, “How do you trade?” and “What tools do you use?” In this post, I’ll share the frame work that makes it all work for me. In the next post, I’ll share tools that have been most useful for me, over a trading career that has spanned three decades.

This will be a high-level summary since each section could probably become a book on its own!

Discretionary vs Systematic

I have focused heavily on discretionary trading over the years, for the simple reason that this approach appealed to me. I do believe there are some other “Easter eggs”—the human brain is capable of processing patterns and relationships in ways that still confound the best models, AIs, and computer systems we have. The human brain, however, is also subject to confusion and error in some interesting ways.

My focus has been shifting ever more toward systematic approaches, so it’s likely you’ll be seeing more writing and work from me along those lines in the near future. Once again, I’m following my interests and inclinations, allowing my trading to evolve in ways that work for me and my lifestyle.

People tend to draw very strong distinctions between these two approaches and advocate for their way being the right way. I don’t think it’s like that—both approaches have their advantages and weaknesses. The arguments you hear for one approach over the other are at least partially untrue, but they don’t capture the complete picture.

Systematic traders, for instance, will say that systematic approaches remove psychological stress. That’s simply false, as you’ll discover the first time you hit a drawdown. Systematic traders also argue that their approach removes the factor of human decision-making. This is also false: a human decided what and how to research, which approaches were worth executing, how much risk to apply (or what model to use), and a human will decide when to turn the system off if it stops working.

Discretionary traders will argue that they see things systematic traders don’t. This is true, but much of what they see may be faces in clouds—patterns that aren’t there! Discretionary traders think their approach is more robust than a systematic approach, but that’s probably effect of hundreds of tiny, unseen adjustments made along the way. This can also backfire and work against consistency. There’s also the fact that much of what a discretionary trader does defies solid backtesting and quantification, so it can be difficult to assess the validity of an edge.

Both styles have enriched my trading. Systematic approaches have honed my quantitative skills and injected a healthy skepticism towards new trading ideas. They’ve taught me the value of consistency. Conversely, discretionary trading has unlocked insights that defy purely rational analysis. It’s deepened my understanding of trader psychology, including my own.

Learn as much as you can about both styles.

Fundamental and Macro

I am, primarily, a technical trader. That said, we do an extensive amount of fundamental and macro work in-house—equivalent to, if not more than, many dedicated macro traders. Many macro signals simply don’t work the way we expect—much of this is counterintuitive—so it’s important to understand the tools and factors you will use.

There’s another complication: the signals that work generally have few, if any, strong timing implications.  Having a good macro signal is like knowing you probably have an appointment on your calendar, but not knowing if it’s this week, three months from now, or next year.

To me, the most valuable aspect of doing fundamental work is being able to cut through much of the misinformation that’s out there. Financial journalists and social media commentators are incentivized to capture your attention, not to educate you. Most of what you hear and read will not be useful, and much of it could be harmful if taken at face value.

If you track fundamental data carefully, you’ll see that there’s a tremendous amount of variability in the signals. There’s enough variability that we should often ignore the week-to-week or even month-to-month numbers. (Of course, it’s important to know when this is not true.) We should really look at the numbers with measures that grasp the volatility and variability of the data. This requires a nuanced look at the data—another reason to ignore public commentary as we live in an evermore black and white, and less nuanced, world.

To sum it all up… Fundamental and macro factors matter, but they are not timing signals. Most of these measures are incredibly volatile. You can’t trust anything you hear or read about these numbers because the incentives of people commenting are misaligned with your own.


We should first define “technical” here: when I use the term, I mean “making or managing trades based on information in price movements.” This is the broadest definition of technical analysis, and covers everything from trading chart patterns, to fully systematic approaches, to quant analysis of short-term flows. The key distinction between fundamental and technical is whether or not you are incorporating data other than price movements in your work.

There are thousands of rooms in the mansion of technical analysis. It is very hard to figure out where to focus attention, especially for a newer trader. In fact, there’s no way to really answer this question without doing some hard work.

Importance of doing your own work

One of the things that is different in my work is that I do a tremendous amount of statistical and quantitative work. This is not normal for most technical traders, and certainly not for discretionary traders. There’s a reason why most people don’t do this work—it’s hard!

When we look at patterns in markets, there’s really only one question we’re seeking to answer: what has happened when this pattern has occurred in the past? To answer that question, we go through historical data and do a few things:

  1. Find every occurrence of the pattern and assume some consistent entry point.
  2. See what has happened after we entered those trades.
  3. Analyze the results.

Each of these steps is complicated. There are two broad approaches to step #1: either we can go through charts and mark the patterns based on discretion, or we can do it with a rule set and probably let a computer mark them. The first of those choices is subject to a massive amount of error and influence from the researcher.

Even some of the best-known “encyclopedias” of technical patterns show patterns that have never failed, that are 100% reliable. This, of course, is impossible—markets do not work like that. When we discover that there was a discretionary input into which patterns were considered valid (i.e., someone looked at a chart and said, “That pattern is good and that one is not”), it becomes easier to see the issues here. Sadly, work done like this is potentially highly misleading. Many developing traders have run  their trading ships ashore following unreliable maps.

Quantifying patterns does not solve everything, however. It may not catch everything. It is possible for a discretionary trader to have a sense of a pattern that defies quantification. Some patterns may depend on complex factors such as time of day, action (or volatility) in related markets, and other aspects of context. It may be possible to train trading models to recognize these factors, but it’s worth acknowledging that some patterns can’t be explained in a simple rule set.

There is value in the struggle, though. Doing this work has forced me to be precise in my thinking. For discretionary traders, any move toward precision is probably good, especially in the early stages.

Working through steps #2 and #3 are more clear-cut, once you’ve identified the entry points of the patterns. The only real complication here is that you need to compare to a baseline drawn from the asset itself. Your comparison is not zero.

Imagine, for instance, you have an active trading system that is long-only stocks and that this system returns about 6% a year over a long slice of data. Great, but the buy and hold return might have been 7% or 8%. Does your approach even have an edge? (Probably not.) Is it worth actively trading to make less than the buy and hold return? (Probably not, unless your system removes much of the volatility of stocks.) That last answer touches on the issue of volatility-adjusted performance, but that’s a more complex topic for another post.

Can the whole be more than the sum of the parts?

In both music and art, like a stained-glass window illuminated by sunlight, the whole appears to be much more than its individual components. Any one sliver of stained glass appears to be insignificant. Is the same true of a trading approach or pattern?

A trade is composed of many small decisions, and even the entry is likely the result of several factors working together. One of the key questions to ask is if we can strip these ideas apart and test each one individually.

From a very practical perspective, consider this: if you test candlestick patterns, you’ll find there’s no statistical edge. But does this mean that candlestick patterns are useless? Could there be times where we should pay attention to a specific candlestick pattern?

Many of you are emphatically answering, “Of course,” but slow down! Are you sure? If it is true that a candlestick pattern might usually show no edge, but might show an edge at some points, then we should (probably) be able to define conditions that make this true.

Technical traders are guilty of saying “context” and dismissing everything with frantic handwaving: “You have to put it in context, of course. Of course a statistical test will show it doesn’t work, because you must have context.” If this is true, quantifying aspects of this context isn’t just crucial—it’s non-negotiable for anyone serious about understanding their edge.

This is not easy to do, but again, it’s a worthwhile task if you want to truly understand your edge. The whole can be more than the sum of the parts, but this can also be the refuge of a trader that doesn’t want to do the work to quantify an edge. Relying on hopes, dreams, and the authority of books will not get you very far as a trader.

Think rigorously about your trading, and about any pattern that is meaningful to you in the market. See how far you can strip them down into component parts, and if you could examine those parts individually. This is the path to trading wisdom and true insight.

What do we do with all of this?

This post has given you some useful context and ways to think about the market. I’ll be back soon to share some specifics of my trading approach—I’ll show you what I have done with these tools myself.


Adam Grimes has over two decades of experience in the industry as a trader, analyst and system developer. The author of a best-selling trading book, he has traded for his own account, for a top prop firm, and spent several years at the New York Mercantile Exchange. He focuses on the intersection of quantitative analysis and discretionary trading, and has a talent for teaching and helping traders find their own way in the market.