Trend Following Vs Mean Reversion: Trading Regimes

[Do Or Die 10]

Almost everyone acquainted with technical analysis knows that trend following indicators work great in trending markets while oscillators work great in range bound markets. What if we could predict accurately whether market will be range bound or trending tomorrow? Well, that would literally allow printing money by shifting from (say) MA crossovers and RSI signals. However, such holy grail does not exists. Mr. Market does not foretells, “Hey, I’m going to continue moving in this direction for next 3 weeks, so please carry your trade; Or I’m going to reverse trend tomorrow from such level so please book your profit and reverse your trade.”

 

Trading Regime analysis means trying to identify similar periods of low volatility or price range contraction; and rising to high volatility or price range expansion. Regime analysis is looking to identify periods when conditions will calm down, at least in the direction of the immediately preceding trend, then this is similar to other so-called mean reversion.

Market moves in phases of low volatility to high volatility. Phases of low volatility can be associated with prices stuck in a trading range, congestion area or a very slow trend. Similarly, phases of high volatility can be associated with clear, larger, and directional moves.

 

Volatility is a very generic term used in reference to trading so let me elaborate what I’m referring by ‘volatility’ here. To clarify, I’m not referring to volatility as end-of-day price returns; but as the price changes relevant to the time frame you are trading. For a swing trader, a stock which hardly moves by 5% at end-of-month will be less volatile than a stock which moves 15% in a month. For a daytrader, a stock which moves by 0.5% will be less volatile at end-of-day than a stock which moves by 4%.

 

I’m using ‘volatility’ in this article in reference to the expansion and contraction of price ranges for a particular time frame rather than the generic speed of change in the prices. Trading Regime refers to whether a market is in a trending mode, range-trading mode or whether it is doing a bit of both. My strategy focus is to identify when such trading regimes exist and when, more importantly, they are likely to change or switch. This can help choose the right trading or investment techniques that go with the particular trading regime and not try to persist with something that would obviously not produce good results in that regime.

 

Let us get started with an example. The Standard Deviation on price changes on a rolling window of last 20 days is a crude way to measure volatility.

 

 

The vertical lines on the chart indicate where the standard deviation has reached what could be thought of as historically low (hi) levels and we can see that the drop in the standard deviation to these low levels usually, not always but usually, precedes a strong trend like move in the price of the OIH. The calculation properties of the standard deviation simple support this observation. The standard deviation is calculating the distribution of price values over a specified time period and so if the standard deviation is low then it suggests that the distribution of price values has been occurring in a tight range. That is, the market has been range trading. Given the mean reverting (cyclical) nature of the standard deviation then, after a period of range trading, it is reasonable to expect the market to enter a period of higher standard deviation and that usually means a trending period. Sometimes, of course, the resulting price action is not a strong trend and the analyst has to be open to this possibility. However, the probability lies with the fact that after a period of price range contraction comes a period of price range expansion.

 

In addition to a low standard deviation preceding a trending move in the market, a high standard deviation can precede a range-trading move in the market.

 

Just as low volatility will probably be associated with a coiling price pattern such as a triangle and high volatility will probably be associated with a trend exhaustion turning phase perhaps with momentum divergence, candlestick reversals, volume considerations, gap theory and most other methodologies will all be cousins of each other. All these linkages are inevitable because we are, after all, analyzing the same underlying variable, the price; it is just that there are different ways of expressing the analysis.

 

Technical techniques in relation to volatility breakouts have become popular whereby most people tend to think of these analyses as useful only for short-term market movements. This, is unjustified because whatever time frame (or fractal) is being examined, the market is being driven by the same underlying psychology that comes from human emotions. Price patterns repeat themselves at every degree of the market. A time series chart of one-minute closing prices will look similar in certain circumstances to a chart showing one month closing prices because the underlying driver of the market is the same human psychology that causes price trends and reversals of those trends. Trading regime analysis, can be broadened out considerably to look at both the long time frames as well as intraday time frames.

 

As you must have already inferred from above discussion, for regime analysis we use components of market behavior which are cyclical in nature. The components could be any data related to price, volume, market internals or their derivatives.

 

Among market data, the most useful is Volume, Hi/Lo Index, TICK and TRIN indicators.

 

For price analysis, Elliott Wave can be particularly useful. Among price derivatives, the most useful are: Standard Deviation, ATR, Bollinger Band Width, Donchian Channels and ADX.

 

I will soon elaborate techniques (usual technical analysis indications) on indentifying regimes through some examples.

 

thanks,

DD

[SIUYA 10]

Hi DoorDie....

as an old options trader and to help this thread or readers of it with a socially responsible message from the surgeon general ...not to hijack or attack it.

 

- the measures of volatility often bandied about in trading rooms have little to do with how volatility is measured and traded by those actually trading it. This is an often missued and abused term.

 

Both congestion and trends can have periods of both high and low volatility periods.

Often a better suggestion is to use/think in terms of - momentum, continuation, reversal - as we are often trading these things and their direction - not the volatility.

 

While you do attempt to define it and establish that volatility is often a generic term it is this statement as an example that can cause issue .....

"For a daytrader, a stock which moves by 0.5% will be less volatile at end-of-day than a stock which moves by 4%" - really?

I assume you mean the range of the stock was 0.5% or 4% for that day.

 

Within that day the 0.5% stock could have actually been less/more/just as volatile intraday, even if the range was smaller.

Just an alert for those when talking volatility.

[Do Or Die 10]

Hi,

 

I very much appreciate your response, but I think there has been some misunderstanding here.

 

My short answer is "Context is king..."

 

I will like you to take note that volatility has a totally different perception for a stocks trader vs a options trader.

 

Say I'm a short-term stocks trader (holding one week to 2 months). I buy AOL in April around 19.00. It may have great daily range (ATR EOD (15)= 50 cent, 2.6% avg daily range) but its not moving around my buy price since last few months. I will be happy to dump it as low volatile.

 

Compare it with KO which has daily movement of around 1.4% BUT it has been moving in the time frame relevant to me. I will consider KO as more volatile than AOL as a short-term trader with history of past few months.

 

 

the measures of volatility often bandied about in trading rooms have little to do with how volatility is measured and traded by those actually trading it. This is an often missued and abused term.

Terms are misused when they are used out of context. Terms are abused when they are used in a totally irrelevant context.

I’m using ‘volatility’ in this article in reference to the expansion and contraction of price ranges for a particular time frame rather than the generic speed of change in the prices.

 

There is a lot to volatility itself and perhaps it deserves a separate article. BUT what I'm talking here about is Trading Regimes. Beginners say all the time that MAs do not work, RSI do not works and so on. The fact is every trading technique/indicator works in a suitable regime.

 

I will elaborate more on trading regimes soon.

[phantom 11]

In addition to a low standard deviation preceding a trending move in the market, a high standard deviation can precede a trading range in the market.

 

As an aside, a high standard deviation can also precede a market reversal.

 

 

Luv,

Phantom

[SIUYA 10]

Hi,

been moving in the time frame relevant to me.

 

this is the important element in tracking volatility.

 

(sorry if there was any misunderstanding, I was not meaning to come across as provacative, more thought provoking as volatility is often misused and taken out of context and more often becomes part of the vernacular without such a discussion as this pointing out some crucial context elements)

[Do Or Die 10]

Thanks Phantom and SIUYA.

 

To all: please refer to Introduction to Understanding Volatility for continued discussion.