Concepts for creating models and hypothesis of price movements

To try and explain how prices move without real data, purely based off of trading and observations on markets, assets and the change in value relative to other prices seems to be an almost impossible exercise.

Yet - experience shows that small amounts of money through trading assets, accepting some losses, and make more profits is possible.

It is also the case that asset value can violently change, leading to widescale losses if not acting on the warning signs. We can define these phases as being;

• Up phases
• Down phases
• Flat phases

Our assertion is that these phases are the only three situations which occur.

Ignoring shorting

We can ignore the idea of shorting assets because, for most people - retail investors, the idea of shorting is violently risky and the potential losses infinite (nothing is infinite?). We can do physical shorting by exiting a long position and being short in the expectation of the price dropping.

Focusing on the three conditions

For this reason, we seek to;

• Avoid losses during down phases.
• Make small gains during flat phases.
• Avoid losses during flat phases.
• Profit on up phases.

Objective outcome definition is therefore;

1. Avoid losses.
2. Make small gains.
3. Make large gains.
4. Save gains.

Safe havens (Save gains)

We can think of safe havens as being the mechanism to lock in profits. To have the ability to leave value in an asset which is inert to volatility whilst everything else is failing seems to be a great mechanism to "live to fight another day". It buys time to determine where to put value next.

The paradox of safe havens

A safe haven is nearly always losing value to another asset's price increase. To be able to determine a safe haven, we have to accept it will never be a true safe haven.

Genuine examples of cycles in markets

Stock Market 2017-2018

It is clear you could make profits by buying stocks where they had (not complete);

• A lowish p/e (Below 30).
• A current ratio above 1.
• Low debt to equity ratios.
• A price operating at a low end of the 52 or 26 or even 13 week high.

You could make a profit.

Then, everything turned.

Cryptocurrency pre January 2018

You could buy assets and they would increase in value. Selling into super-exponential rises in value relative to fiat seemed not worth doing.

Cryptocurrency post March 2018

We have seen a steady decline of value to a point where the highest value is now 80% lower than the high.

Gold market post 1971

Behavioural tendencies leading to apprehension and losing profits

Fear of missing out (FOMO)

FOMO causes people to;

• Avoid selling when an asset keeps increasing in value.
• Avoid buying when asset has already increased in value - how can it go higher?
• Avoid taking profits gradually as prices continue to rise.
• Fear of selling all position in an asset for fear of not profiting on super potential gains - i.e., Bitcoin could be worth 1 million so why sell?
• Panic - buy and sell. We suddenly take extreme action to try and protect revenue.

Lack of long term understanding

• Assets can go to zero value.
• Governments can destroy markets and assets.
• New assets can make other assets irrelevant.

Impatience

Markets can take longer than humans have to profit from them.

Misunderstanding of volatility and time

Not the same as impatience. stretching and contracting the time horizon magnifies, nullifies, and reduces volatility.

Lack of appreciation for efficiency

Taxes significantly reduce profits and yet many assets are protected by tax free schemes.

Complete inability to analyse data and take informed action

Without; viable data, automated intelligent systems, intelligent agents to act, educated guesses.

Inability to lock in profits

Hypothesis - The snake index

When a snake moves, it's head seems to determine the overall direction the rest of its body will take and yet, certain segments of its skeleton move up and down seemingly in opposite directions to other parts of the snake at some time, and the same direction to other parts of the snake at another time.

We may be able to state that the head leads the body always, yet other observers would state that a large change in the direction of a part of its body is responsible for the change in direction of the snake and other parts of the snake.

Copernicus discovered the sun was fixed and the earth rotated around the sun. In much the same way, once we know a snake has a brain we can pretty much rule out the idea that parts of the body determine where the head should go. This is how markets are, they concoct elaborate theories which seem as ridiculous as the sun rotating around the earth.

The snake index is a metaphor for how we can try to track asset values in relation to others. If we assume an index to be comprised of 10 assets, it is likely there will be a dominant asset or group of assets which we can state the head to be, a other set of assets we can state the body to be, and a third set of assets we can class the tail to be. In our ten asset example;

• Two assets are the head, based upon their market capitalization.
• Six assets are the body.
• Two assets are the tail.

At this point, it is important to understand the following;

• The determination of head, body and tail are configurable and dynamic.
• There is no need to see an equal number of assets at the tail or the footer.
• The assets can exist anywhere. They don’t have to be a portfolio of cryptocurrencies, blue chip stocks, a set of industrial stocks or a group of municipal bonds.

Validating the snake index

Our only goal is to be able to manage positions to maximize profit and minimize loss. The proofs are that profit keeps occurring and original opening stakes are not lost. This may sound like we are begging the question, but if our model works and keeps working, it seems a reasonable validation.

Our offline attempt to build a snake index trading tool

This is covered elsewhere, but to recap, a number of libraries and applications were built to;

• Execute trades when trigger levels were hit.
• Generate synthetic prices.
• Collect asset prices from cinmarketcap.com on regular intervals.

Time ordinals

• If we checked every ten minutes, there may have been n trades at price x.
• Time ordinal offset (lag,lead)
• To be able to act on a change of one assets price in relation to another, there needs to be enough time to do so.
• Ticks
• We define ticks as being sequential time periods. Ticks means something else in stock prices.
• Direction
• Within an ordinal, stocks at the beginning of a tick which have a lower price than at the end of a tick have moved up. We can also see asset prices which moves down or are fairly flat.
• Chart measures
• We consider correlation, goodness of fit to determine how consistent price movements are within a tick.
• Slope
• In linear algebra, we know that a slope can be used to determine the direction a price is moving in. We found the m or gradient to be ineffective and was looking to use tangent. This allowed us to measure slopes in terms of gradient.
• Tile/rank
• By tiling a gradient we can compare completely different assets by a scale of increase or decrease.
• Other approaches?
• Could we have used percentages?

Wrapping up

This article is purely a thinking space. It is not financial advice, do undertake your own research before attempting to use an approach such as this. We will be seeking to create models as our platform matures and test these hypotheses on different platforms such as trading view via Pine Scripts.

Written with StackEdit.