Just wanted to share some of my thoughts while developing trading systems.
Among the challenges of trading, is identifying the current volatility conditions we will be trading into.
Volatility refers to how wild are the ups and downs of price over a period.
It’s easy to just look at where the market has been recently, or where the market is now.
But how good is that information as an indication of the future conditions we will be trading into?
Why do we want to identify the volatility of the period we are trading into?
Volatility helps us plan our trade, in terms of the amount of room we will allow the trade to breathe, as well as how to plan our entries and exits.
Higher volatility will usually require wider stops and more duration in drawdown, could also allow better entries and further exit targets.
Indicators like ATR allow you to see recent volatility and how it is changing.
I believe there is some predictive value to observing how the ATR is changing rather than the actual value of the ATR itself.
A more popular way is to observe for a period of low volatility, which tends to precede a period of high volatility.
This type of play is known as a volatility breakout.
Bollinger bands offer a rear-view mirror look at recent volatility.
It also gives us a measure in terms of standard deviations from the mean, where prices would be considered at extremes caused by injection of volatility.
Bollinger squeeze: Similar concept to observing for periods of low volatility which would indicate a period of high volatility coming next.
Price trading at 3rd standard deviation extremes can represent:
A move back to the mean.
An injection of volatility caused by the start of a trend.
(This is when price starts “walking” on the 2 standard deviation.)
These seem to be contradictory instances don’t they?
Something to think about: Can we make a distinction between either scenario?
Another challenge, not confined to just predicting future volatility, is heteroskedasticity.
Heteroskedasticity refers to the phenomenon of inconsistency of error variance across different sample sets of data over time.
In a way, this is the reason all investment products come with the clauses like “past performance is not indicative of future returns” etc.
There is no one “correct” solution to the challenges stated above, if there were, then the markets would not be the markets, it would be an ATM machine!
However, there are ways to adapt to it and even allow for it in our trading, such that over time, we will come out ahead.
One way is to anticipate changes, we can do this by observing the world around us.
For example with Brexit, we knew to expect huge volatility, so we can either try to take advantage of it, or stay out of the markets.
Another way is to play on volatility breakouts.
No matter how we play it, the simple rules apply, when you don’t have a magnificent statistical “edge” which gives you extremely high success rates:
Look to keep our losses small and keep our winners big.
Over time, we will come out ahead.
Good trading folks!