Saturday, February 09, 2019

Some thoughts on defining market states

When people talk about the four market states, typically they refer to these as trending, non-trending, stable (low volatility) and volatile.

However, what you need to consider is there is no definitive answer to how you identify each state, and the answer may differ from trader to trader.

Providing there are no restrictions in buying and selling, then markets will continue to generate 'trends'. At the end of the day, price can still only do one of three things – go up, go down or be stuck in a range. And two of those three have the potential to be categorised as trending.

The difficulty, and why people occasionally jump on the "trend following is dead" bandwagon is that, as Ed Seykota says, there is no such thing as THE trend.

There numerous trends in any market, at any time. As a result, and to further muddy the waters, you may get conflicting signals. 

For instance, you may be looking at a long-term uptrend on a weekly chart using similar long-term parameters, yet on the daily timeframe, and with a different set of parameters, your method may be showing you a downtrend.

We saw in the last quarter of 2018 that, within the equity markets, long-term trend followers were experiencing a pullback within the context of the long-term uptrends. Depending on their parameters, they will have experienced an erosion of open profits, or possibly their stops being hit.

On the flip side, those who trade shorter-term trends may have been able to capitalise and profit from the downtrends which occurred on many stocks in that same period.

While performance is undoubtedly influenced by the markets being traded, we have all seen periods where shorter-term trend following systems outperform longer-term systems, and vice versa. As it happens, with the uptick in general volatility and overall lack of direction, it seems like the shorter-term systems fared better in 2018. 

In other periods, however, the reverse will occur. Longer-term systems may be able to embrace any minor choppiness. And, as I know only too well, shorter-term methods can easily suffer from whipsawing in these phases!

In the past I have said that, should 'trends' start to develop, then all trend followers should be able to benefit. However, that is not strictly true.

You can check this out for yourself by looking at the performance numbers of some of the bigger trend following firms. Depending on the portfolio composition, and whether their focus is geared towards shorter-term or longer-term trends, performance can wildly differ both from month-to-month and year-to-year.

In this regard, you can also refer back to the original Turtles experiment from the 1980's. All the traders involved were looking at the same basket of instruments, so any issues about portfolio composition were eliminated.

However, some traders traded ‘S1’ (20 day breakouts), others ‘S2’ (55 day breakouts) and yet still others traded a combination of the two. And when you look at the monthly performance figures those traders achieved, there were some pretty wild variations!

In 1985, Stig Ostgaard returned an amazing +296.56%. Yet Jim DiMaria 'only' achieved returns of +71.12%. Liz Cheval fared worse, at a pedestrian +51.65%, but she performed better than Tom Shanks, who could only muster returns of +18.1%!

A year later however, Shanks achieved +169.53%, outperforming Ostgaard who returned +108.21%. DiMaria hit +131.68% and Cheval +134.68%.

Taking this a stage further, it is not only the timeframe and parameters which can effect the timing of entries, exits and ultimately performance, but also the nature of the price move in the markets you are trading.

I'm sure you can easily find a chart of a market whereby your own timeframe and parameters are perfectly suited – you get an entry signal near the start of a move, you get an exit signal somewhere near the end of the move, and any minor counter-trend movements do not trigger an exit. 

You will also find plenty of charts where there is no doubt there is a definite directional movement in place. However, your own parameters and entry/exit rules may mean you are unable to profit from that move. The pullbacks or retracements within the overall 'trend' keep triggering your stops, and your entries keep getting you back in at the wrong time. 

With this in mind, it may be better to say that potentially the worst combination of market states is trending and volatile. Generally it is accepted that non-trending and volatile is the worst combination for trend followers, but with well-defined rules the chances are you could be kept out of trading those markets.

So, you need to remember the the definition of each market state is not a blanket 'one-size which fits all'.

What it is better to say, is that your definition of these market states is solely determined by your own timeframe and parameters. 

And as a result of this, how you or I go about defining these market states may result in significantly different performance.

No comments:

Post a Comment