An Introduction to Cycles in the Forex Markets

It is now generally accepted that cycles of all kinds occur in nature and that they also occur in financial markets although with less formality than in the natural world.

The 4-year (41-month) Kitchin cycle is probably the best known financial cycle and it measures highs and lows in business activity. Other cycles include the 54-year Kondratieff wave and the 18 and 9.2 year cycles in equities.

At the end of the 20^th and beginning of the 21^st century some of the accepted cycle models failed to work as accurately as before because markets experienced sustained periods of growth and speculation which led to the excesses of the credit bubble and the financial crisis. This was much like the extended period of prosperity which bucked the trend in the 1920s and was then followed by the Great Depression.

In this first introductory article I have steered away from testing these common market cycle models in the foreign exchange markets and instead tried to look for completely new cycles. Future articles may attempt to integrate current cycle theory and new approaches to arrive at a better understanding of the link between foreign exchange and other markets and existing cyclical models.

It is important to understand that not all cycles work all the time – technical analysis is not an exact science – but they can be a useful confirmation or early warning tool for the ends of trends or beginning of new trends. Another principle to remember is that useful cycles are good at measuring turning points in markets but not always the direction of the turning point.

The 67-day cycle in the EUR/USD pair

There is evidence of a 67-day cycle in EUR/USD with mainly lows at the cycle trough intervals. This equates quite accurately to the actual number of working days in any 3-month period – or a quarter. It is not surprising that there should be a quarterly cycle in the foreign exchange markets given the importance of major fundamental data releases such as GDP which represent economic activity on a quarterly basis. The cycle measures market turns of intermediate and large degree mainly. It is also quite good at marking the end of sideways consolidation ranges and the point at which trending activity is likely to resume. The diagram below illustrates this cycle:

Observing the diagram above the cycle does seem to pick major highs and lows reasonably effectively. A cycle trough occurred at almost exactly the bottom of the recent Jul ’10 bear movement and it also picked the Nov ’09 high quite accurately. The cycle also signalled the ’08 highs and lows of the financial crisis very accurately encapsulating in a complete cycle almost the entire bear market of a year. Unfortunately, like all cycles there are times where it fails and it was first a little early for the March ‘09 low and then completely failed to mark the April ’09 low – or was very late in marking it.

I have analysed 39 occurrences of this 67-day cycle in the history of the euro-dollar and have found that it has had an 85% success rate at marking significant turning points. Cycle points which occurred at the end of consolidation moves were not included as ‘failures’ even where they did not occur at significant highs or lows, because, they nevertheless marked major tuning points in activity – ie from sideways to directional – and this was sufficient to validate them. Only those cycle lows which did not mark any change in the direction of the market at all were counted as failures.

13-15 points occurred within a 5-day range of a significant market turning point or a week in non-market time. This was quite a high number or 33% – 38% of the total cycle troughs. Out of the 39 cycles 4 marked the end of consolidation periods very accurately. For those who wish to test this cycle the last cycle low occurred on the 29^th November 2010.

The Daily Volatility Cycle in the EUR/USD

There is a fairly reliable daily cycle in the EUR/USD pair which measures changes in volatility. This cycle makes a trough during the low volatility measured in the early hours of the morning between 5 and 7am GMT and then rises to peak in the high volatility of the afternoon sessions between 12am and 5pm in the afternoon.

The cycle is based on the common sense fact that raw market activity tends to be concentrated during ‘Anglo-Saxon’ trading hours when London and New York are trading – particularly when they overlap – because the foreign exchange industry is concentrated in these cities.

The diagram below shows the broad cycle on a 30-minute chart of EUR/USD using Average True Range on an 8-period setting as a measuring method of volatility. As can be seen from the diagram the cycle marks highs and lows in volatility with unerring regularity.

Given many traders tend to trade volatility and not direction this cyclical phenomenon could be a helpful device for such operations. A simple ‘box’ strategy where a buy order is placed above the low volatility consolidation range and an order to sell short below might work quite well to take advantage of this cycle. However, a cursory glance at the data shows that one problem with such a trading methodology would be the inevitable false breakouts which would ensue during the early stages of the high volatility part of the cycle when it would be quite common for price spikes to trigger trades in the wrong direction to the day’s main move. To avoid this problem would require the placing of orders with great care, probably quite far above and below the consolidation range which would inevitably erode the risk/reward ratio for the strategy.

One way of possibly getting round this problem and also avoiding subjective order placement might be to use Bollinger Bands. The high and low bands could be used to pin-point trade entry points at the moment in the cycle in the early morning when they would be squeezing together. The opposite band used for entry could provide a stop and visa-versa. The use of OCO (Order Cancels Order) might be another useful technique in avoiding the necessity of forecasting direction and minimising risk of damage from whipsawing.

A trailing stoploss could be used to maximise returns although this might suffer as a result of whipsawing in the highly volatile climate of midday and afternoon trading and so another alternative might be to use a time-based stop to close the trade. Whilst the range between 5 and 7am seem to provide the optimum low volatility time for entry, the time around 5pm tends to be when volatility peaks, the trend for the day is exhausted and volatility then begins falling. Such a time could provide a useful time-based stop.

Obviously for certain strategies such as those used by option’s traders, which profit from lack of volatility, the reverse cycle points could be used with the low volatility early mornings providing the perfect environment for profiting from lack of directional change.

A possible 17-year cycle in USD/CHF

The dollar-Swiss franc pair could have a roughly 17-year cycle which began in October 1978. However, given how long the cycle is and the lack of available historical data going back further, at this stage this is still only a speculative cycle. Nevertheless it has to say that the chart makes a compelling case it (see picture below).

The first cycle low was in 1978 after which the market rallied and then fell to a new low in April 1995 – 17-years latter. We are at the latter stages of the next cycle with the next trough forecast to occur in October 2011.

The recent dollar lows have already suggested a long term trough-like phenomenon in the dollar but it remains to be seen whether the exchange rate will fall any further and make fresh lows before the cycle low is expected in 2011 or whether perhaps we have already reach those lows and the cycle is slightly late this time. If the dollar begins a strong rally either now or next year then the cycle may indicate a much stronger dollar for quite a long period in time – perhaps the following 5 – 8 years, if not longer.

Posted in: FOREX