The Limits of Technical Analysis: What It Can and Cannot Do
After fifty posts on tools and patterns, an honest summary is overdue. Technical analysis is a way to organize information about price. It is not a crystal ball.
What it does well is describe context: where price has been, where participants reacted, and how volatile the market is. It offers objective reference points for stops and targets. And it gives a framework for managing risk, since a stop placed beyond a logical level has a reason for being there.
What it cannot do is predict with certainty. No indicator or pattern converts the future into a known quantity. The same setup produces wins and losses, and a portion of any outcome is luck.
There is also the question of self-fulfilling behavior. Widely watched levels attract orders, which makes them matter. But when everyone sees the same thing, crowded trades can reverse violently.
This leads to the real foundation of trading: probabilities, risk control, and consistency. A method with a modest edge, applied without deviation and sized carefully, beats a brilliant method applied erratically.
The analysis should answer three questions: what do I expect, where am I wrong, and what is the reward relative to the risk. If a tool does not help answer one of them, it is decoration.
Practical step: review your toolset and write which of these three questions each tool answers. Remove those that answer none.
The Limits of Technical Analysis: What It Can and Cannot Do
Re: The Limits of Technical Analysis: What It Can and Cannot Do
One thing I'd add to the summary: technical analysis is most useful for risk management. Even if it can't predict, it gives you a logical place for a stop and a target. That alone makes it worth using. Without a chart, you'd have no reason to place a stop at one price rather than another. With one, you can say that if price goes past this level, the idea is wrong. That clarity is valuable even if your entries are only slightly better than random. The traders who do best with it tend to treat charts as a map of where reactions happened, and their real edge comes from consistent sizing and exits, not from forecasts.
It’s Fairman 
-
LondonNewsTrader
- Posts: 421
- Joined: Mon Sep 21, 2026 9:30 am
Re: The Limits of Technical Analysis: What It Can and Cannot Do
I'd add a fourth that technical analysis can't answer on its own: when could something outside the chart change all three? The calendar is the most obvious part of that. A level, a stop and a target can all be sound at 13:25, and by 13:31 the release has made the stop meaningless. Knowing the timing of scheduled events doesn't predict anything either, but it tells you when the chart-based answers are least reliable.
Re: The Limits of Technical Analysis: What It Can and Cannot Do
Where it struggles is with sudden fundamental shocks. No chart pattern predicts a surprise rate cut. That's why I always check the calendar and keep risk sensible, even when the chart looks perfect. The SNB dropping the franc floor in 2015 and the sharp yen rally of summer 2024 both blew through levels that had looked solid for months. A chart could tell you where stops were clustered, but not that the trigger was coming. So I treat technical levels as the plan for normal days, and position size as the plan for the abnormal ones, since no level protects you when the news is big enough.
It’s Fairman 
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: The Limits of Technical Analysis: What It Can and Cannot Do
A simple way to see how much of a method comes from the entry signal is a random entry test. Keep the stops, targets, sizing and session rules exactly as they are, and replace the entry signal with a coin flip at the same times of day.
Say the real entries give plus 0.12R per trade over 300 trades and the coin flip version gives plus 0.08R. Then most of the edge sits in the exits and the sizing, and the entry adds about a third.
If the coin flip version is clearly negative while the real one is positive, the entries are doing real work. Both results are useful, because they tell you where to spend the next month of testing.
I am less sure that every stop needs a drawn level behind it. A stop set at one and a half times the average range of recent bars has a reason too, and it ignores levels completely.
That kind of stop tends to sit away from the obvious swing points where others cluster theirs. On pairs that like to run those points before turning, it can survive moves that clip the level based version.
Say the real entries give plus 0.12R per trade over 300 trades and the coin flip version gives plus 0.08R. Then most of the edge sits in the exits and the sizing, and the entry adds about a third.
If the coin flip version is clearly negative while the real one is positive, the entries are doing real work. Both results are useful, because they tell you where to spend the next month of testing.
I am less sure that every stop needs a drawn level behind it. A stop set at one and a half times the average range of recent bars has a reason too, and it ignores levels completely.
That kind of stop tends to sit away from the obvious swing points where others cluster theirs. On pairs that like to run those points before turning, it can survive moves that clip the level based version.
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: The Limits of Technical Analysis: What It Can and Cannot Do
Technical analysis is good at showing where price is relative to recent levels and trends, and less good at predicting what happens next.
Combining it with clear risk rules and realistic expectations is what makes it work. A level does not have to hold every time, it only needs to hold often enough that the small losses are covered by the wins.
Combining it with clear risk rules and realistic expectations is what makes it work. A level does not have to hold every time, it only needs to hold often enough that the small losses are covered by the wins.
FXS Official Team | The Idea Lab & Strategy
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: The Limits of Technical Analysis: What It Can and Cannot Do
Computer clocks drift. A typical PC can lose or gain a few seconds a week if it isn't syncing regularly. On a VPS, drift can be worse, since virtual machines share hardware time.
For most activities that doesn't matter. For scalping analysis, it does. When comparing the timestamp of a trade with the timestamp of a news release or a fill, a drift of 2 seconds distorts the picture. The trade looks like it was placed 2 seconds before the news, when it was really after.
Similarly, scripts that log latency by comparing the local time with the broker's timestamps produce nonsense if the local clock is off.
The remedy is simple. Windows syncs with time.windows.com by default, but only once a week, and sometimes less. Setting a shorter interval, or installing a tool that syncs every few minutes, keeps the error under a few milliseconds.
A quick check: visit a time-comparison website and see how far the computer's clock deviates. If it's more than half a second, it's worth fixing before trusting any timing data.
How accurate is the clock on the machine used for trading, and has a drift ever distorted an analysis?
For most activities that doesn't matter. For scalping analysis, it does. When comparing the timestamp of a trade with the timestamp of a news release or a fill, a drift of 2 seconds distorts the picture. The trade looks like it was placed 2 seconds before the news, when it was really after.
Similarly, scripts that log latency by comparing the local time with the broker's timestamps produce nonsense if the local clock is off.
The remedy is simple. Windows syncs with time.windows.com by default, but only once a week, and sometimes less. Setting a shorter interval, or installing a tool that syncs every few minutes, keeps the error under a few milliseconds.
A quick check: visit a time-comparison website and see how far the computer's clock deviates. If it's more than half a second, it's worth fixing before trusting any timing data.
How accurate is the clock on the machine used for trading, and has a drift ever distorted an analysis?
Re: The Limits of Technical Analysis: What It Can and Cannot Do
The fourth question is the one I miss most often, honestly. I can have a level, a stop and a target all sorted by 13:20, and then forget there's a release at 13:30. A stop 8 pips away and a spread that's 1 pip normally can turn into a 25 pip slip when the number prints, which turns the planned 40 dollar risk on 0.5 lots into over 100 dollars.
So my routine now starts with the calendar before the chart. I mark anything medium or high impact for the day and treat a 30 minute window either side as no-entry. If I'm in a trade at that point, I close it or move to breakeven if it's already moved at least 1R.
I agree with the point that knowing the time doesn't predict the outcome. All it tells you is that the chart's answers are weakest right then, so the sensible move is to reduce size or stand aside, not to guess direction.
Do you include the speakers, like central bank members, in that window? Those are harder to see in advance and I've been caught a couple of times. And how long after the release do you wait before trusting the levels again?
So my routine now starts with the calendar before the chart. I mark anything medium or high impact for the day and treat a 30 minute window either side as no-entry. If I'm in a trade at that point, I close it or move to breakeven if it's already moved at least 1R.
I agree with the point that knowing the time doesn't predict the outcome. All it tells you is that the chart's answers are weakest right then, so the sensible move is to reduce size or stand aside, not to guess direction.
Do you include the speakers, like central bank members, in that window? Those are harder to see in advance and I've been caught a couple of times. And how long after the release do you wait before trusting the levels again?
It’s Fairman 