I have a pet peeve that has been bothering me for a while. It bothers me because I largely think it does a massive disservice to our industry. You see it all the time, someone reads a book and then starts posting charts full of trendlines and indicators. They pass this off as technical analysis, but it is really “charting.” Unfortunately, people think that "technical analysis" and "charting" are the same thing.
They're not.
One is a discipline rooted in evidence and the scientific method. The other is an “art” form built on subjective interpretation. To be fair, both have a place, but they are not remotely close to the same thing.
David Aronson made this distinction the key point of his 2007 book, Evidence-Based Technical Analysis. If you haven't read it, you should because in it he describes how he came to realize, through his own experience putting capital at risk, that what was being passed off as “analysis” was really just random nonsense.
The Core Problem Aronson Identified
Aronson's argument is simple at its foundation. Much of what passes for technical analysis is just storytelling. It's a practitioner looking at a chart, seeing a pattern, and weaving a narrative around it after the fact.
The problem is that we are human, we need to find patterns and order. If we think we can make money from that order, all the better. Show us random noise and we'll find a head-and-shoulders top, a cup-and-handle, a hidden message. Our brains are wired to see structure where none exists. Then we think we are smart enough to profit from that perceived structure. Aronson leaned heavily on cognitive science to make this point, but that can be a whole other post.
So the question becomes: how do you separate what works from what only appears to work?
What Is Subjective Charting?
Let me be clear here, a chart is simply a graphical representation of data. That's it. I use charts constantly for presentation purposes, to frame a story, to show a client what price is doing at a glance. But I would never make a trading decision based on charting alone. The chart shows me the data, that’s it.
Subjective charting becomes a problem when interpretation gets dressed up as analysis. Think about the classic methods:
Trendline placement that varies from analyst to analyst
Elliott Wave counts that can be redrawn whenever price disagrees
Hand-drawn pattern recognition with no fixed rules
"It looks like it's forming a bottom here"
Could you imagine if we put our client’s capital at risk because I saw a “rounding bottom” on a chart?
The real issue is that these methods cannot produce a falsifiable claim. If two skilled chartists look at the same chart and draw two different trendlines, who's right? There's no objective answer. And if a method can't be proven wrong, it can't be proven right either.
This is not a knock on charting. Charting can build intuition and frame a market story. In many cases it is also the starting point for idea generation. If you look at enough charts, you begin to come up with theories that can be tested. But until you test, you do not have evidence that what you see on the chart works.
Enter Evidence-Based Technical Analysis
Aronson defined evidence-based technical analysis around one requirement: a method must be objective enough to be tested. That means a signal must be defined by precise, repeatable rules. No discretion.
If you can write the rule as code and run it across thousands of historical bars, you can measure whether it has predictive value. This is the scientific method applied to market data:
Form a hypothesis. ("A 50-day breakout produces excess returns.")
Define the rule objectively so it can be coded.
Define Entries, Exits, and Position Size
Test it across a large, representative sample.
Measure the result against a benchmark of random chance.
Try to disprove it before you trust it.
That last point matters a lot. A claim has value only if it's falsifiable. You don't prove a strategy works by finding charts where it worked. You prove it by failing to disprove it across a rigorous, out-of-sample test.
But, just because we are testing something, does not mean that there are not pitfalls still…
The Data-Mining Trap
Here's where most investors with good intentions are at risk. Especially with the proliferation of “back-testing” capabilities now being made available.
If you test enough rules against enough data, some will look brilliant by pure luck. I can torture market data until I get a strategy with a 3+ Sharpe Ratio. That doesn't make it viable in the real world.
Aronson dedicated a large portion of the book to this problem, what statisticians call data-mining bias. He argued that without proper statistical correction, most back tested results are worthless. The returns are an illusion created by the testing process itself.
His solution involved formal significance testing, things like Monte Carlo, to estimate how likely a result could have occurred by chance. If your strategy's edge can't clear that hurdle, there is likely no edge.
Investors should ask not just whether a strategy worked historically, but whether the result is robust enough to inspire confidence going forward.
Where I Land on This
I'll be honest. I love charts. On a Sunday morning while my family is still sleeping (because I get up at 4:00am) I love to sit and scroll through charts. I find it therapeutic in a way, me, my charts, and my espresso. So, you might expect me to defend subjective charting.
I don't, not as an effective way to manage money. I think Aronson was right about the core issue, and our industry is better for having his book in the world.
The chart is a presentation tool. It's how I show the data and frame the market's story for much of the content that we produce. But the trading decisions are made based on objective, testable rules. They are based on code!
Aronson’s concept forces discipline. It makes you write your rule down. It makes you test it against chance. It makes you accept a result you don't like. Here is a harsh reality, most ideas don’t work. Finding an edge in the market is hard. But the work is worth it!
Final Thoughts
The difference between technical analysis and charting comes down to one word: evidence.
Evidence-based technical analysis demands rules. It demands testing. It demands intellectual honesty about whether an edge is real, luck, or produced by datamining.
Use charts to visualize and communicate. Use evidence to allocate capital.
But don’t confuse the two.
