LCT Blog Post: FOREX Currency Trading Systems Work Until They Don't

Hi, I'm Peter Rose, Founder of Longwood Currency Trading, and welcome to LCT Blog Post 05/10/20 — FOREX Currency Trading Systems Work Until They Don't.

A FOREX trading system is something like: "Enter a trade long when the fast moving average crosses up above the slow moving average."

Well, actually, that's more of a 'rule' than it is a 'system'. But, a system is really just a collection of rules.

Regardless, I could go on then with a whole bunch of stuff to outline a trading system. But that's not the point of this post.

The point of the post is to caution that all such systems work, but at some point they stop working. Yes, yes: they all — at some point — stop... working....

Why?

Well, let's just go back to that example moving average cross over system. Even though it only has one 'rule', it could still be someone's complete trading system, so we'll run with that for discussion purposes.

If you've done any trading at all, you will have tried some form of that stupid moving average cross over system. And you've had success with it. But at some point, it didn't work. When was that? When did it stop working?

Better yet, when does it work? It works in trends, right? Yup. So then what you see when it doesn't work is when price starts to chatter in a channel. That's when it doesn't work.

So, you've got a system that works until it doesn't. Terrific.

So, what do you do when it stops working? Well, you'd formulate another system for trading price bouncing around between the channel's support and resistance boundaries, right?

Yup. And then that system will stop working when price goes back into a trend — or something else. So, then you switch back to your moving average cross over system — or something else. Terrific.

What you're doing here is using indicator based system construction. I had a lot to say about indicators in my post The Myth of Using FOREX Trading Indicators. The thrust of that post was if you only had 5 indicators that you were considering, there would be 325 different permutations those 5 indicators could be sequenced in.

That's just a disaster waiting to happen.

So, what's the solution then? Would you build infinitely increasing complex 'systems' to deal with every permutation? That's hardly practical. In fact, it's not practical at all. And yet, that's what most trading resources imply.

Well, why would that be if they don't work? Why teach that, or why use such systems if they don't work? Great questions....

The answer is that they are taught and used because: How else would you teach or learn something if it's not built on a set of rules based on available information? That's how you'd teach someone to build a house, right?

Yeah, but it doesn't work that way in trading. You can't build a trading 'system' because that system must be based on rules, and — as I've shown — since there can be an almost infinite number of rules such a system would lead to complexity beyond implementation.

The conclusion thus must be to have no 'system'. But you have to have rules! Yes, you do have to have rules.

The solution is to understand that you can't create a system defined by rules, but rather you must have rules that represent the system in what I call 'market dependent trading rules', i.e. rules that are dependent upon the market, not the other way around.

I know, that might sound a little confusing. I'm not sure the following rather detailed discussion of the computer programming concepts of the Object Oriented Principles of Container and Aggregate Relationship will be any clearer, but let me give it a try as the principles are the same. This discussion is taken from one of the trading courses I've compiled called LCT Intermediate Currency Trading Program.

Price Action Within The Currency Market
I retired in 2015 after a 33 year career as a senior software engineer designing and programming enterprise web based business applications using object oriented methodologies. One of the main principles of object oriented design is the concept of a container object as opposed to other objects that are within that container object, called aggregate objects: a container object holds aggregate objects. For example, a shopping bag is a container object whereas the cans of soup in that bag are "contained" within it, they are aggregates of the bag.

The programming "rule" that describes such a system is that the container object must have knowledge of what it contains, whereas the contained objects can not have any knowledge of their container.

The practical explanation of this is that the only thing a can of soup needs to know about is whatever it is that cans of soup need to know. That might be such things as: expiration date, weight, servings per can, calories per serving, etc. But what a can of soup does not need to know is that it is inside a shopping bag.

The shopping bag, however, needs to know not only all of the characteristics of what a shopping bag is, for example, how strong it is, but it needs to know what's inside of it. It can then "ask" each contained can how much it weighs and thus determine how many cans it can hold. The cans don't care about this information, nor should they have to as that's information that's only needed by the bag object.

What's all that have to do with price action and the currency market? If I was going to model the currency market as an object oriented system so that I could program some sort of simulation application, I would cast the market as the container object, and price action – the objects that would represent price itself – would be the contained, or what's called the "aggregate" objects, of that market container.

The immediate affect on the system, from an object oriented standpoint, and in reality, from a practical standpoint as well, is that price action, or an object that represents an open position, has nothing to do with the market. It doesn't even know about the market. However, the market – could be the market makers or large institutional traders – would have intimate knowledge of all of the contained aggregate price action objects, i.e. the big institutions have the ability to see where all of the limit orders are, for example.

If this is your first brush with this theoretical construct of object oriented systems design, it's probably a little confusing. Strip all of that away and just look at the grocery bag and its aggregate cans of soup. Who uses the can? The can can't open itself, for example. It needs some other external object applying an initiating force to do that. In the case of the grocery bag, that might be an action to smash that can against the bottom of the bag hard enough to pop it open. Yeah, stupid example, but you get the point.

Having explained that, can a price action object actually "do" anything by itself? For example, could a price action object execute a command to move itself 3 pips higher? The answer is no. Can an engine contained within a car turn itself on? No. It needs an operator to turn a key on the container car object to do that. The conclusion here is that only the market, i.e. traders, can cause the price to change. What price action is doing at any given time is totally dependent on what the market, i.e. the big institutional traders, are doing right then. And by "right then" the implication is "now", not 15 minutes ago, or last week. Now.

Because of this, looking at chart indicators like moving average lines, stochastics, head and shoulders patterns, a shooting star candle, etc. to make a trading decision is about as useless as basing a trade entry on the shape of a peanut butter smudge on the edge of the screen. You need to look at what the market/traders are doing, not what price is doing. The market, the large institutional traders, is the dog, whereas price action is the tail. You have to understand who wags who....

It might appear that by looking at what the market is doing is implying using fundamental vs. technical analysis. This is not the case. Fundamental analysis is based on looking at the same past information that technical indicators are, except fundamental analysis concerns economic information whereas the technical analysis deals with what price itself did in the past.

Price action is the summation of all of the traders that constitute the market making decisions in real time. The wind creates waves; the waves do not create themselves because the waves are aggregate objects of the environment which controls the wind. Sure, you could run this out to a higher container of this as being the Earth, and then the solar system, and then....

In looking at any system to see why it is behaving the way it is behaving, the immediate container is as far up the chain as you normally need to go. Thus, to understand price action you don't look at what price action has done in the past to determine what it might do now, or in the future, but rather you look at the market, the traders placing orders in real time. What are they doing "now"?

And the question to ask about those traders is: why are they doing what they are doing? The answer is, and will always be: traders do what they do out of a complex interaction between their greed and their fears in order to accomplish just one thing: make money.

Hopefully, that made sense. The principle concept is that it's the current market, the 'dog', that creates current price action, the 'tail' being wagged by the dog, rather than price action showing what the market is doing — because it already 'did it'.... and thus whatever price did in the past is not germane to what it's doing right now.

The market 'system' is thus where the rules need to be applied to, not to the price action created by the market.

Conventional price action modeling is thus invalid if this principle is not the underlying foundation or source of information for the associated rules, i.e. when the rules are fixed to the price action instead of to the container market.

Obviously, a simple blog post such as this can't teach how to do this. The object of the post is to alert you about certain discrepancies in how trading principles are presented, i.e. what the issues are.

If I was working with an advanced private mentoring student I would spend a lot of time going through the computer programming concepts of the Object Oriented Principles of Container and Aggregate Relationship. It's challenging to show a student how to change their view of the relationship between the market and the price action the market creates. But once they 'get it', it's a remarkable asset as to how they then work out trade entry set up analysis.

It's not critical to have this training to be able to trade well anymore so than having the need to take an advanced gourmet cooking course to make a hamburger.

Trading is simple; not complex. Studying all this theoretical container/aggregate relationship crap is not necessary to trade successfully. Later, when you're trading well, then this type of training will be very instructive. But you don't need it to get or be profitable.

However, you need someone who knows what they're talking about as to the full scope of what's going on to help you see and get proficient at the smaller scale stuff.

You're not going to learn to play chess properly, for example, from someone who has never played through all the openings in Reuben Fine's classic book, "Ideas Behind The Chess Openings" because that person just doesn't have the higher level view to help set the stage for you. The teacher that has that background is not going to shove that book in your face, or have you play through any of it. Rather, they'll probably just verbally start you out with something like the Ruy Lopez to lay the foundation of chess opening theory for you.

Trading is simple. You just need to foundation it all properly. A great analogy to this is famous pro golfer Lee Trevino's approach:

"Regardless of where I am, or what the weather is, or what the occasion is, every day I hit 1000 balls. When I step out onto the course, I know where the ball is going, my opponents only think they know where the ball is going."

If I was going to learn how to play golf, that's the attitude and experience I'd want in a mentor. That type of person could explain things simply to me with knowledge that would provide a solid foundation for future study and practice.

I had that in my karate Master (S.A. Brock, St. Louis, MO), in the world class classical guitar teacher I studied a little over 2 years with (Walter Spaulding, Kingston, NH), in learning to cook (Master Chef James Haller of the former Blue Strawberry Restaurant, Portsmouth, NH), in writing (my brother Alan who held a PhD in 19th Century English Literature, taught at the United States Naval Academy in Annapolis, MD, and later Head of the English Department at Frostburg State College, Maryland), and with the real estate broker who I shadowed to learn real estate investing. I was successful in all of those pursuits.

But I didn't have anyone to guide me in currency trading. And I failed miserably.

When I looked back at the differences in the 'systems' I used to learn all of the things I was successful in with that of teaching myself how to trade I immediately understood why I had failed at trading: the folks who mentored me showed me how to implement general principles expressed in simple rules about the activity, whereas I was following conventional trading lore by trying to create a system of rules for the activity, i.e. the currency market.

What a revelation it was for me to not only recognize this, but then to immediately be able to implement the change to my approach to trading. I had been like all the other golfer's instead of following a course of action similar to Trevino's approach.

Thanks for taking your time to read this post,
Peter

Of further interest might be the following featured on my YouTube channel for Longwood Currency Trading:

Video: Problems With Using Technical Analysis in FOREX Trading