Exchange Rate Forecasts
Forecasting exchange rates, is it possible?
Many people believe that it is possible to accurately forecast exchange rates using fundamental analysis, whether they can or not I do not know, but in this short article I am going to see if there is any evidence that exchange rates can be forecast using technical analysis. As with all methods of technical analysis, the only information we have to use for our forecast is the market price, or rather the market's past price.
In this experiment, I am going to try and use technical analysis to forecast the closing price 30 days from now and see how several simple methods of technical analysis compare with just assuming that the closing price in 30 days time will be the same as it is now. The methods of technical analysis are as follows –
1. I will compare today's closing price with the closing price 30 days ago and assume that the price will move the same amount and in the same direction over the next 30 days as it did in the 30 days previous. If the closing price is greater than the closing price 30 days ago then the exchange rate forecast is to be: the closing price + (the closing price – the closing price 30 days ago). If the closing price is not greater than the closing price 30 days ago then the exchange rate forecast is to be: the closing price – (the closing price 30 days ago – the closing price).
2. I will use two simple moving averages of the closing price, a 30-day SMA and a 60-day SMA. I will assume that the closing price in 30 days time will be the difference between the two moving averages. If the SMA 30 is above the SMA 60 then the exchange rate forecast is to be: SMA 30 + (SMA 30 – SMA 60), if the SMA 30 is not above the SMA 60 then the exchange rate forecast is to be: SMA 30 - (SMA 60 –SMA 30).
3. I will compare today's closing price with a 30 day simple moving average of the closing price. I will assume that the closing price in 30 days time will be the difference between today's price and the average price of the previous 30 days (the 30-day SMA). If the closing price is above the 30-day SMA then the exchange rate forecast is to be: closing price + (the closing price – the 30-day SMA). If the closing price is not above the 30-day SMA then the exchange rate forecast is to be: closing price – (the 30-day SMA – the closing price).
I will measure how accurate these methods are and how close they come to predicting the closing price in 30 days time at the close of each day, and take an average of their accuracy over a long period of time. I will compare these accuracy averages with the average accuracy we get when we assume that the price will not change at all.
The logic behind this is simple; the logic is that if technical analysis can be used to forecast future prices then we should be able to find evidence of this in even the crudest forms of technical analysis producing better currency price 'predictions' than simply assuming that there will be no change at all. I will test this on the EUR/USD and the GBP/USD from the beginning of the year 2000 to the end of 2010. The results are as follows –
EUR/USD –
Experiment #1: Current price compared to price 30-days ago: 3.865%
Experiment #2: 30-day average price compared to 60-day average price: 3.375%
Experiment #3: Current price compared to 30-day average price: 3.125%
Assuming no change: 2.822%
GBP/USD –
Experiment #1: Current price compared to price 30-days ago: 3.467%
Experiment #2: 30-day average price compared to 60-day average price: 2.817%
Experiment #3: Current price compared to 30-day average price: 2.780%
Assuming no change: 2.391
Using technical analysis to forecast exchange rates: my conclusion,
The percentages in the results above show the average of how much each of the daily predictions was off by. Predicting prices daily using technical analysis actually seems to have performed worse than just assuming that the price wouldn't move at all. I am doubtful about the ability of technical analysis to predict future prices just as I am doubtful that analysts using market fundamentals can actually predict future exchange rates. Nevertheless, technical analysis does have some uses. On markets that trend well, technical analysis can be used to determine how strong the trend is and therefore how likely it is to continue – but it can't predict when the trend will finally end. On markets that are driven by speculation, technical analysis is very useful at identifying likely support and resistance levels and whether the market is over bought or oversold. And on markets that are prone to sudden bursts of movement, technical analysis can be very useful for detecting momentum. But I can see no evidence that technical analysis can be used to predict the future closing price on any given day in the distant (or even the not to distant) future; and I'm my experience, the 'gurus' using fundamental analysis can't do this either, most of the TV pundits are however really good at telling you what has just happened.
|
