Do the laws of physics work in forex? - page 12
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
When searching for patterns in price movements, analogies with various physical phenomena are not uncommon. And with varying degrees of success, different laws of physics, mathematics, geometry, etc. are adapted and applied in trading.
In this topic we will consider the applicability of Newton's and Hooke's laws to price movements.
I will proceed from the fact that any price movement requires a trade or - force application in the language of analogy. The volume of the trade is the modulus of force.
The resulting force can be determined by looking at the chart by the slope angle of the moving average. In this case, the MA period will be an indicator of force duration.
Since there are a lot of force sources and they all differ both in module and duration, several moving averages with different periods can be considered. This will help divide the resultant force into its components.
First of all, let us consider one moving average. In practice, it is not easy to accurately measure the slope of an average due to a large number of random fluctuations. In my opinion, the Hodrick-Prescott filter is the best solution for filtering these fluctuations.
For the calculations, we will take a filtered average with a period of 30. We will get the picture (Fig. 1.) where the slope angle of the average is not so dependent on random fluctuations and clearly shows the trend direction.
Let's calculate the modulus of force in the indicator according to the formula of Newton's second law F=dv/dt (fig. 2.). In the second figure we can see the time of action, modulus and direction of the force acting upon the price.
Now let's look at the deviation of price values from the median line. The constant convergence/divergence of the price vs. the midline suggests an analogy with the elasticity or tension force.
Also we use the formula F = - kx to build an indicator (fig. 3.).
The elasticity coefficient k has been selected so that the values of the forces are of the same order. It is obvious that k must be related to the period of the average as well as in physics this coefficient is related to the properties of a particular material.
Do you have any idea how to calculate it?
Further I have added both forces together in indicator (fig.4).
I have done the same with 3 averages with periods of 10, 50 and 250 and smoothed the result with the period of 15. The result is shown in fig. 5.
Connecting this indicator as a signal source for the EA produced the following picture (Fig. 6).
I suggest discuss how to calculate the significance ratios for averages with different periods as well as the calculation of the elasticity coefficient.
I was working on something similar a long time ago, but I was stopped by the inability to calculate the elasticity. I pondered over this analogy for a long time and came to the following conclusions.
There is no resilient force as such, rather it is present in some places and absent in others. It looks more like a spring which was stretched by bargains pushing the price higher but then when the price is at this level the resilient force weakens gradually and the longer the price is at this level, the less the tension remains. The length of the spring has increased, they have sort of added coils there and now this length is normal, you can still stretch. The analogy works, but with more caveats:
1- the spring length is variable
2- there is a coefficient of elasticity, but it is not clear whether it is constant or variable
3- Each scale is a different spring.
More accurately, we can think of the system as a huge number of (unstable) springs of different lengths and different elasticity, during price movements some of the springs break and new ones appear periodically.
How it relates to the real market: I can open a trade by buying a volume from the market for as long as I like, I can move the price at least 100 pips (theoretically). But then I need to close the position. And here is where the magic happens. If no one has filled orders in the cup, the spread simply widens and I can now with my volume buy out the opposite side of the cup and close the position and the price will fall by 100 points from its initial point (actually the spread will become wider and uncertainty will arise till the moment the order is placed in the empty cup). If the stack is already filled with orders, I need enough liquidity to close my position to stay in profit. So the price can go in 3 extreme cases
1- below buy price (if there is no liquidity at all)
2- return to the buy price (if there was enough liquidity only to close the position and return the price to its original level)
3- to stay at the new level (if liquidity has completely extinguished my needs).
That's the 3 spring options. In the last variant the spring length increased, in the second variant it did not change, in the first variant it decreased....
But there are plenty of entrants on the market, and a different spring for each.
There is no resilient force as such, or rather it is there in places and not there in places. It looks more like a spring which has been stretched by trades, pushing the price higher, but then, as the price is at that level, the resilient force weakens gradually and the longer the price is at that level, the less tension remains. The length of the spring has increased, they have sort of added coils there and now this length is normal, you can still stretch. The analogy works, but with more caveats:
1- the spring length is variable
2- there is a coefficient of elasticity, but it is not clear whether it is constant or variable
3- Each scale is a different spring.
Let's imagine a one dimensional spring in a vacuum :-)
Some analogy in the model: at one end of the spring is a smaller load, at the other end is a larger load. We can only see the smaller load and can act on it along the axis of the spring. However, even in comparison with the smaller load, our action is pejoratively small and can only give something in sum with the others, which do not depend on us in any way. The graph of the smaller load will be remarkably similar to the quotes over time.
But we can't assume anything from this graph. Neither the position of the ends of the spring in the future, nor the relationship between the weights and the elasticity, nor the distance between the weights. Moreover, the model is not physical - it has no relation to the "objective" of our field.
This is the effect of the observer - he is lazy and therefore stretches familiar concepts over unknown entities.
If there are chemists, the trend reversal can probably be explained by "because de Broglie" :-)
Imagine a one dimensional spring in a vacuum :-)
Some analogy in the model: at one end of the spring there is a smaller load, at the other end there is a larger load. We can only see the smaller load and can act on it along the axis of the spring. But even in comparison with the smaller load, our action is pejoratively small and can only give something in sum with the others, which do not depend on us in any way. The graph of the smaller load will be remarkably similar to the quotes over time.
But from this graph we will not be able to guess a damn thing. Neither about position of the ends of the spring in the future, nor about the relations of masses and elasticity, or about the distance between the weights. Moreover, the model is not physical - it has no relation to the "objective" of our field.
This is the effect of the observer - he is lazy and therefore pulls familiar concepts over unknown entities.
If there are chemists, the trend reversal can probably be explained by "because de Broglie" :-)
Yes, that's approximately the same problem we got in the end - a one-dimensional spring in vacuum with two weights). So it turns out that it's not a spring that should be considered but influence of independent events on the weight. If we can calculate this, the spring becomes unnecessary, as well as the physical analogy with the mechanical equations.
The model is too crude and has little to do with the markets, you either have to refine it before... before a common one emerges or develop another, more approximate model right away.
The problem with any physical model (based on mechanics) is that in the end the market consists only of variables, the constants are so few that even if they exist, they can be neglected.
If we develop a model, I think we should start from constants. What do we know about the market and what constants can be there?
I can only suggest one constant, and that is the speed of order execution. It is also not a constant, but the value is predictable and its increase is also predictable.
But there are many participants in the market, and there is a different spring for each.
Here is a screenshot of a chart with two average lines of different period. The standard deviations are plotted for each line.
The distance from the price to each line determines the degree of stretching of the corresponding spring. At the same time, the direction of each of the middle lines determines the vector of the force of inertia.
By adding the moduli and vectors for the inertial and elastic forces we find the resultant force.
The idea is to describe this mathematically and find the resultant force.
Yes, that is about the same problem, a one-dimensional spring in vacuum with two weights). So it turns out that it is not the spring that should be considered, but the effect of independent events on the weight. If we can calculate this, the spring becomes unnecessary, as well as the physical analogy with the mechanical equations.
The model is too crude and has little to do with the markets, you either have to refine it before... The model is too crude and has little to do with the markets, we either have to refine it before the general one arrives or develop another one which is closer to it.
In essence it comes out that we only need shocks to the cargo, because it is these shocks that set the cargo in motion. But the vector of these impacts and the force are not constant and set just the motion of the price. So the cargo can be removed and considered as the impacts on the weightless plate from the opposite sides of some number of objects of different mass and different velocity. These impacts will move the plate in different directions, which side has more momentum, that wins. So elasticity becomes unnecessary.
This model is more like a model of gas with a certain temperature, and the size of gas molecules is different, but it has some average value. This is the model for reflection:
A gas with a certain temperature (average) is on opposite sides of a weightless and infinitely thin partition. The gas temperature is constantly rising (inflation). now we need to create a model of the movement of molecules and hurricanes against this plate, the displacement of the plate will be the price graph.
Parameters of the model: Current temperature (average speed of movement), non-uniformity by volume of temperature, average size of molecule, non-uniformity by size of molecule.
@Alexander, how is such a model?
Obviously, the force of inertia is proportional to the angle of inclination of the mean and the period of the mean, while the force of elasticity is proportional to the amount of deflection.
It is clear that new forces appear and disappear all the time, but when a new force appears, it will automatically take part in the calculation.
This is the point of having an idea of the current resultant at any time.
In fact, it turns out that we need only shocks to the weight, because it is these shocks that set the weight in motion. But the vector of these blows and the force are not constant and set the price motion. So, the load can be removed and be considered as blows on a weightless plate from opposite sides of some number of objects with different mass and different velocity. These impacts will move the plate in different directions, which side has more momentum, that side wins. So elasticity becomes unnecessary.
This model is more like a model of gas with a certain temperature, and the size of gas molecules is different, but it has some average value. This is the model for reflection:
A gas with a certain temperature (average) is on opposite sides of a weightless and infinitely thin partition. The gas temperature is constantly rising (inflation). now we need to create a model of the movement of molecules and hurricanes against this plate, the displacement of the plate will be the price graph.
Parameters of the model: Current temperature (average speed of movement), irregularity by volume of temperature, average size of molecule, irregularity by size of molecule.
@Alexander, how is such a model?
maybe, but it seems to me that your model is more complex in terms of implementation, while I don't see any advantages.
Suppose that the criterion for the correct choice of the average period is the minimum ratio of the average maximum deviation to the standard deviation over the last 200 ticks.
Can you write the code of indicator to check this assumption?
I don't have it (.
I already checked it. There is no consistency. There are completely new variants of analysis. Group behaviour.
maybe, but it seems to me that your model is more complex in terms of implementation, and I don't see any advantages.
so the graph is redrawn