Trend Following With Regression Curves
Understand trends and trade the market with regression curves as indicator for the ongoing trend. Map them out as a forecasting technique to make sense of price action and its overall direction.
- Matthias Hagemann
Posted October 12th 2020
Marcus Tullius Cicero, a statesman and philosopher in ancient Rome, is famous for saying that the man who doesn’t know what happened before he was born goes through life like a child.
Before a new trend is born, it is already trotting on its underlying regression curve. Your task as trend follower is to understand its context and to become aware of the dominating regression curve early.
If you don’t spot the curve, you will trade the market like a child — and earn like a child.
Regression curves are the foundation to trend following. Without them, you lack orientation and you are just poking in the dark. That’s why you must know this technique if you are following trends. No, seriously.
What Is a Regression Curve
A regression curve in trend following is a smooth curve fitted on a price chart according to the principle of least volatility.
What does this all mean? Let me explain.
We analyze where the price regularly moves back to by intentionally blending out its larger gyrations. In other words, you focus on the narrow corridors where the price doesn’t move much. These are price levels where most investors could find a consensus. Trace these narrow corridors and you will see how they shape up a smooth curve. In statistics, this method of regression analysis is called ‘curve fitting’.
Do you get the idea? The curve is a logarithmic line where the price moves with least volatility. It helps us establish a bias in order to follow trends along that curve.
Here’s an illustration of what was said above:
Every trend is confined to its regression curve and it’s our task as trend followers to identify, with the help of our naked eyes, which curve dominates.
I’m certain that there is a way to turn it into algorithm and the experimental technocrat is free to give a try. For the time being, we have to fall back to human intelligence.
The Four Types of Regression Curves
There are four types of regression curves along which all trends evolve and they always take the shape of a logarithmic line. It is vital to detect them early to make sense of price action and to predict its overall direction.
- Accelerating uptrend: Prices are narrowly range-bound with an upward tendency and gradually accelerate to escape velocity.
- Flattening uptrend: Prices bounce with strong upward momentum which gradually fades to a flattening range.
- Accelerating downtrend: Prices are narrowly range-bound with a downward tendency and gradually accelerate to a desperate collapse.
- Flattening downtrend: Prices collapse with strong downward momentum which gradually fades to a flattening range.
How to Analyze Trends in Modular Principle
You can stack one regression curve after another in modular principle to make sense of price action and to predict the overall direction. Trends can even follow two or more curves at the same time if you zoom in and out of time frames.
There can be an accelerating uptrend in the weekly chart while the shorter term, daily chart indicates a flattening uptrend followed by an accelerating downtrend.
A new trend may be shaping up once price action deviates too far from its ordinary price corridor along the current regression curve. The initial move is often enough to pick the most likely regression curve out of the four.
Our previous example analyzed in a little more detail could be like so (it may look bizarre but you get the idea):
What If Our Mappings Differ
Your own analysis could differ from mine and that’s perfectly legitimate. This kind of analysis is generally subject of debate which is why it is best solved either individually as a seasoned market participant or collectively with fellow trend followers.
You can see that, in the broader perspective, most price action was just short-term gyration which can be exploited with varying degrees of experience.
If you prefer a good night’s sleep, the focus should obviously remain on the dominant trend in the bigger picture. You’ll then have to live with the inevitable volatility.
When doing this analysis with individual stocks, I would perform it with its fundamentals in mind. What growth story is the business pursuing? Are the right people at its helm and delivering executional excellence? Do fundamental metrics back up that story?
For stocks to go up, determinant factors are revenue growth and earnings growth. The Boston Consulting Group further cites profitable, “value-creating” revenue growth as primary driver. Growing ‘at all costs’ is where many businesses fail.
A regression curve, after all, is a mere visual reflection of the business’s trajectory. It does not feed on itself.
When coming back to Cicero’s quote, we have to conclude that we cannot look at a regression curve independently from its context. We must bring a good understanding of other disciplines to the table in order to be consistently successful.
That’s why I’m convinced that trend following is an occupation for people with a multidisciplinary background who enjoy intellectual debates, challenges and puzzles — not for the simple-minded chasing instant fortunes.
Now that we have the foundational part of trend following cleared, let’s proceed with trading bounces using moving averages.
Beat the 2021 Stock MarketGet 3 quick strategies by email to invest successfully and to outperform the market. I’ll also send you an invite to my Discord server so you can learn from me in real-time during trading hours. It’s all free!
More Trend Following Posts
The moving average bounce trading strategy gives reliable trading signals in the prevailing direction of the trend. It is best used in trend following systems. This is how you buy and sell stocks with the lowest risk and highest reward potential.
Price action is the trading activity in a pricing process. It is used in trend following to understand how trends evolve and how they can be traded for profit.