Extending The Frontiers Of Technical Analysis: The Application Of W. D. Gann’s Forecasting Method To The Currency Markets

Article originally published in "Market Technician" No 84 - March 2018

 

Introduction

William Delbert Gann (1878-1955) was a successful stocks and commodities trader. He also wrote seven books and numerous short courses on how to trade. Gann’s success as a trader was based on his forecasting method which he discovered from his research between 1902 and 1908.

This paper examines the application of Gann’s forecasting method to the currency markets. The evidence suggests that Gann never traded currencies. I examine the reason for this and why his forecasting method can nonetheless be applied to the currency markets today, and also look at the potential practical problems.

 

Why Did Gann Not Trade The Currency Markets?

Gann does not mention currency trading in his books and courses, nor in his various advisory services or in his personal trading records. Thus it is reasonable to assume that he never traded them. To identify the reason for this, it is first necessary to examine his forecasting method.

Cycles of time form the basis of Gann’s forecasting method (Smithson 2016) and hence he referred to cycles throughout his writings, including:

“Time is the most important factor in determining market movements because the future is a repetition of the past and each market movement is working out time in relation to some previous time cycle” (Gann 1946, p4) and

“My experience has taught me that nothing can stop the trend as long as the time cycle shows up-trend. Nothing can stop its decline as long as the time cycle shows down. Stocks can and do go up on bad news and go down on good news” (Gann 1949, p3).

He summarised his forecasting method in the following statement:

“In making my calculations on the stock market, or any future event, I get the past history and find out what cycle we are in and then predict the curve for the future, which is a repetition of past market movements” (Gann 1927, p76).

Thus Gann’s forecasting method essentially has three stages:

1) Acquire a detailed price history of the financial instrument;

2) Analyse that price history to identify the underlying cycles driving the financial instrument; and

3) Forecast the future prices of that financial instrument from the future progression of those underlying cycles.

Therefore a necessary condition for the application of Gann’s forecasting method is that a financial instrument’s prices should fluctuate freely in accordance with their underlying cycles; both historically (i.e. during the period of the price history) and into the future.

However, the evidence suggests that between 1908 (when Gann completed the discovery of his forecasting method) and 1955 (when he died) this condition was not generally present in the currency markets. To be specific, the major currencies typically did not fluctuate freely.

A detailed history of currencies from 1908 to 1955 is beyond the scope of this paper; but here is a summary of the exchange-rate regimes of the leading industrial countries during that period which indicates the constraints and controls on currency prices:

1908-14: Fixed-exchange-rate gold standard

1914-18: First World War: gold standard suspended and capital controls

1919-27: Interwar managed-floating-exchange-rate period

1928-31: The major currencies returned to gold standard by 1928 and from 1931 started leaving gold standard in response to the Great Depression                               

1931-39: Interwar managed-floating-exchange-rate period

1939-45: Second World War

1945-55: Bretton Woods system of fixed exchange-rates

In summary, it would appear that Gann never traded currencies because, from 1908-55, the major currencies were typically prevented from fluctuating freely; which nullified his forecasting method.

 

Application Of Gann’s Forecasting Method To The Currency Markets Today

It is more than 60 years since Gann died and the Bretton Woods system of fixed exchange-rates that was in force for the last decade of his life eventually collapsed in 1971-73. Today the major currencies (viz. Australian dollar, British pound, Canadian dollar, euro, Japanese yen, Swiss franc and U.S. dollar) ostensibly float freely. They are monitored by their respective central banks, all of which are mandated to maintain economic stability and hence can intervene in currency markets when they deem necessary. For example, the Swiss central bank pegged the Swiss franc to the euro between September 2011 and January 2015.

The key question, which I shall now investigate, is: Can Gann’s forecasting method be applied to the currency markets today?

As stated above, the first stage is to acquire a detailed price history of the financial instrument. I decided to examine first the U.S. dollar and therefore obtained a series of daily price charts (open/high/low/close bar charts with arithmetic price scale) of the U.S. dollar index (i.e. measured against a trade-weighted basket of currencies) from 1988 to 2017. I selected a 30-year period because this is long enough to include a range of market conditions. Finally, I selected a starting-date of January 1 1988 because that was 15 years after the end of the Bretton Woods system of fixed exchange-rates. I hoped that by this time the U.S. dollar and the other major currencies would have been fluctuating freely in accordance with their underlying cycles.

As stated above, the second stage is to analyse the price history to identify the underlying cycles. I found it quite straightforward to identify the set of cycles driving the U.S. dollar and therefore I concluded that Gann’s forecasting method can be applied to the currency markets today.

Unfortunately my conclusion was premature. More specifically, when I then applied this same methodology to the other major currencies (as stated above, the Australian dollar, British pound, Canadian dollar, euro, Japanese yen and Swiss franc) I was unable to identify clearly any underlying cycles; whether I analysed a currency index or the various currency pairs.

My initial thought was that I may have been mistaken in my analysis of the U.S. dollar, that proponents of the Efficient Markets Hypothesis were correct after all and that market prices are essentially random, making looking for underlying cycles in a price history a naïve and futile activity. However, I rapidly came to my senses, remembering that I had already successfully applied Gann’s forecasting method to a range of stocks and commodities.

I therefore hypothesised that currencies present a specific problem: individual stocks and commodities are essentially single financial instruments that are measured in a currency that is relatively stable compared with the stock or commodity. In contrast, a currency pair is essentially two financial instruments fluctuating simultaneously in accordance with their underlying cycles and therefore it is difficult to identify those cycles.

I then hypothesised that I had been able to identify the underlying cycles of the U.S. dollar from its index price history because its cycles dominate those of the basket of currencies it is measured against. In contrast, the underlying cycles of the other major currencies are relatively evenly-balanced. Therefore the cycles driving a particular currency are obscured, whether one is examining a currency index or currency pair.

I therefore concluded that in order to identify the underlying cycles driving a particular currency it is necessary to measure that currency in terms of a second currency, one that is relatively inert. Consequently, I set out to find a currency issued by a country both economically and politically very stable and with few natural resources (e.g. mineral or energy assets, whose cycles may influence its currency). The New Zealand dollar appeared to be a suitable currency.

I then re-examined the major currencies, employing the New Zealand dollar as the reference currency. In each case the underlying cycles driving that currency could finally be identified. I then re-examined the U.S. dollar, but measured in New Zealand dollars, and confirmed my earlier findings regarding its underlying cycles.

I then examined a selection of minor currencies, the Brazilian real, Indian rupee, Mexican peso, Norwegian krone, Russian rouble, South African rand, South Korean won, Swedish krona and Turkish lira. Once again, I employed the New Zealand dollar as the reference currency and, in each case, the underlying cycles driving that currency could be identified.

I excluded some currencies from this analysis because they are linked to others, once again nullifying Gann’s forecasting method. These were the Chinese yuan (partially-pegged to a basket of currencies), Danish krone (pegged to the euro), Hong Kong dollar (allowed to trade in a range linked to the U.S. dollar) and Singapore dollar (stabilised against a concealed basket of currencies).

In summary, my research encompassed the 21 most actively-traded currencies by value, which constitute approximately 97% of foreign exchange turnover (BIS 2016). More specifically, the underlying cycles of 16 currencies were identified. Four were excluded because they are linked to other currencies and it was not possible to identify the underlying cycles of the New Zealand dollar due to the apparent absence of a more stable reference currency.

I then concluded that Gann’s forecasting method can be applied to the currency markets today. However, in order to identify the underlying cycles driving a particular currency it is necessary to analyse the price history of that currency measured in New Zealand dollars.

At this point I should apologise for not providing a detailed exposition on how to identify the underlying cycles driving a currency from a price history of that currency. I am however continuing a practice started by Gann himself: “Mr. Gann has refused to disclose his method at any price” (Wyckoff 1909, p55) and “It is not my aim to explain the cause of cycles” (Gann 1927, p78).