Nearly every investment strategy seems to fall in one of three buckets. There is trend-following--go with the underlying direction of prices--and its opposite, mean reversion, look for the trend to be reversed. Lastly, there are arbitrage strategies. We would put carry trades, where a low yielding currency is borrowed to finance the purchase of a higher yielding currency (or asset).
The Bollinger Bands offer one way to operationalize reversion process. When a currency moves more than two standard deviations from the 20-day moving average, it is often seen as over-extended, signaling a likely loss of momentum and a reversion to the average. It is a dynamic mean.
Another way to implement a mean reversion strategy is almost the equivalent of the Dogs of the Dow, where the worst performing Dow stocks or the next-to-worst Dow stocks are bought. In the currencies, one can buy the under-valued currencies and sell the over-valued currencies.
Valuation is particularly difficult in the currency market, which has become unmoored to gold or silver. What is the value of a fiat currency? Economists have developed a valuation metric: Purchasing Power Parity. In its simple form, it says that the price of an internationally traded basket of goods (Big Mac, Starbucks (NASDAQ:SBUX) coffee, iPhone) should sell for the same price in different currencies. To the extent they do not, implies a currency misalignment.
Like with other tools, PPP is often misused. Here's how. Consider that ant. Reports suggest that depending on the particular species (and what they had for breakfast?), an ant could lift 10, 20 and even 50x their weight. Average people cannot even come close. In this sense, an ant is stronger than a person. Consider a lightweight boxer. Given how much weight she can lift proportioned to her weight, she might be said to be pound-per-pound stronger than a heavy-weight boxer. It does not mean she can bench press or squat more. The heavyweight fighter is still absolutely stronger.
Purchasing Power Parity is the pound-per-pound measure applied to currencies. According to some measures, if the Chinese yuan were to trade levels implied by PPP, the Chinese economy might be bigger than the US. This leads to all kinds of doom and gloom in the US and may spur more aggressive policies. In China, it can lead to over-confidence and adventurism abroad. This is not what PPP means, any more than the lightweight fighter is stronger than the heavyweight. It is a theoretical construct that is still very helpful when used properly.
The OECD has its own model to calculate PPP. What would happen if one sold the currencies that the OECD said were the most over-valued at the end of last year and bought the most under-valued currencies?
At the end of 2017, the Swiss franc, Norwegian krone, and the Danish krone were over-valued by nearly 21%, 19%, and 15.5% respectively. They fell this year against the dollar. The Swiss franc has fallen around 1.3%, the Norwegian krone about 2.3%, and the Danish krone fell by about 2%. The OECD still regards these three as the most overvalued against the dollar (18%, 15%, and 9% respectively).
What about buying the most under-valued currencies? At the end of last year, these were the yen (-12.5%), the euro (-10.5%), and sterling (-5.5%). All three have fallen against the dollar this year. The yen has eased around 1%, the euro 1.5%, and sterling has slipped around 2%. They have fared a little better than the over-valued currencies, but the margin is not meaningful given the volatility.
Another way to illustrate that this mean reversion strategy struggled this year is that the order of most overvalued to the most undervalued has hardly changed. The only change looking at the top ten currencies is that the euro and yen switched so that now the euro is more undervalued (~17%) than the yen (~13.5%). Alongside the euro, yen, and sterling, the Canadian dollar is the other that OECD says is undervalued. At the end of 2017, it was less than 0.5% off PPP. Now it is nearly 7% cheap. This year, the New Zealand dollar and Swedish krona are near fair value.