Seeking Alpha
Research analyst, value, growth, newsletter provider
Profile| Send Message| ()  

Executive Summary

Market lore and disclaimers from ETF companies warn investors to stay away from double reverse ETFs for trades longer than a few days. This caution is unnecessary for investors using stock ETFs to be short the Japanese yen. During the last 140 days of the yen's current declining trend, being short via YCS was twice as lucrative as using FXY. If the yen's declining trend continues, YCS should continue to be the vehicle of choice for stock investors.

Introduction

Shorting the yen has gotten increasingly popular, especially since Japan elected the Abe Government. The Bank of Japan, urged on by the Abe Government, has set out to double the monetary base by doing $80 million in Quantitative Easing (QE) per month. The yen has declined in value against the dollar as a result. The Government of Japan obviously considers the declining yen, a predictable result of their QE policy, a benefit to Japan's exports and the economy, as well as a key factor in their attempt to turn long-term deflation into 2% inflation.

Those investors that wanted to help the Japanese government in their quest, and make some money while doing it, have used a variety of tactics. Mark Cuban borrowed yen, converted it into dollars and paid off his US$ denominated debts.

Now he services his low interest loan in yen that are cheaper each time he makes a payment. There are a number of ways to profit from the yen's decline, like selling yen forward by contract, using the futures market, options or stock ETFs.

Short-yen investors without Foreign Exchange accounts tend to use the listed ETFs. There are two main choices:

1. Guggenheim's CurrencyShares Japanese yen Trust (FXY) seeks to track the actual price of the Japanese yen. They do this by depositing yen with JP Morgan Chase (JPM). The interest earned is credited towards the management fee.

2. The UltraShort ProShares yen ETF (YCS) seeks a return that is minus two times (-2x) the percentage daily price change of the US Dollar-Japanese yen currency cross, before fees and expenses. Note that it attempts to do this on a daily basis, "measured from one NAV calculation to the next" and that this is not the same goal as Guggenheim's FXY.

Dangerous Inaccuracies and Tacking Error?

Some of my clients were concerned that using the Double Bearish YCS would be a worse investment over time, compared to sticking with the simpler FXY, which is little more than the ETF depositing and withdrawing yen from a saving account. They had heard that there were large tracking errors and friction in ETFs held for more than a few days and that the problem was magnified in double ETFs. Reading the ProShares company disclaimer made them even more wary. It warns that, because of "the compounding of daily returns, ProShares' returns over periods other than one day will likely differ in amount and possibly direction from the target return for the same period."

This warning deters investors that want to play the secular, rather than the daily moves of the yen and get increased leverage for larger profits. More scary, the disclaimer warns that the ETF could move in the wrong direction… that an investor could see the yen decline buy lose money being short via YCS. So I decided to investigate to see whether the danger was real. The first step was to look at a Yahoo chart that compared the USD/JPY cross with the FXY and YCS ETFs. (click to enlarge)Ultrashort Yen YCS, vs. Yen ETF FXY vs. USD/Yen cross

A look at this (linear) chart comparison shows that the changes in the FXY are close to identical to those of the yen. The YCS seems to move much more than the yen, as advertised, although whether it moves twice as much is hard to tell by looking at the graph.

To get a more accurate idea, I did some simple statistical comparisons. YCS tries to mimic the daily percentage price change of FXY times minus two. The easiest way to get an idea of whether this is being achieved, more accurate than a chart, is to look at the average daily changes for the dollar/yen cross, the supposed mirror FXY, and the double reverse YCS.

average daily % change

FXY

YCS

USD/JYP

expected

actual

expected

actual

average

0.13%

-0.13%

-0.21%

0.41%

0.41%

standard deviation

0.74%

0.74%

0.77%

1.54%

1.56%

Nov. 14, 2012 through May 17, 2013

The dollar/yen cross moved an average of only 0.13% per day during the 140 day period in which it weakened from 80.2 yen to 102.98 to the dollar. If FXY was a perfect mirror, it would have had an average percentage decline the same size… but it was 65.1% higher than they expected change per day. The variability, represented by the standard deviation, was about the same. It looks like the FXY is more volatile than the actual underlying yen, making it a more lucrative bet than the yen itself, if you bet on a move in the correct direction.

The results for the YCS ETF were almost exactly what they should be if the YCS was managed to mirror the FXY. The mean daily percentage change was twice that of FXY and the variability was also twice as large. Obviously, YCS would be the traders' choice.

I ran a correlation among the daily changes for each ETF and for the yen, to quantify the discrepancy between the observed average daily % moves between the USD/JPY and the ETFs.

Correlations between daily % changes

USD/JYP

FXY

FXY

-78.59%

YCS

77.92%

-99.92%

Nov. 14, 2012 through May 17, 2013

As the data on the averages suggested, the FXY and YCS could not get much more alike, with a correlation of 99.92%, so close to a perfect 100%. The ETFs were strongly correlated with the dollar-yen cross, but nowhere near as high as with each other.

Longer-term traders may also be interested in the absolute value of the changes and not the percentage change on a daily basis. Correlations run between the 3 yen measures revealed that the prices during the 140 days of yen turbulence were almost perfectly correlated with each other.

Correlations between daily prices of

USD/JPY

FXY

FXY

-99.70%

YCS

99.81%

-99.69%

Nov. 14, 2012 through May 17, 2013

Focusing on the relationship between the daily closing prices of YCS, FXY and the number of yen per dollar (USD/JPY), since the start of the yen's decline, the correlation is close enough to 100% to consider it as close to perfect as you can get. Even though the YCS ETF was only trying to mimic twice the daily percentage change in the yen, it mimicked the yen itself better than the FXY did.

Summary

The FXY and YCS ETFs mirror the USD/yen cross but are slightly more volatile. Overall, they are highly correlated. Given that investors take unhedged long or short positions because of an expectation of movement in one direction, the last 7 months of market data indicates that UltraShort ProShares yen ETF will provide twice the leverage with little chance for significant tracking error or a persistent directional mistake. The fear of double-reverse ETF tracking errors have yet to appear in the short yen trade.

Source: YCS Outperforms In A Yen Decline, Without Significant Tracking Error

Additional disclosure: I am long call options on this ETF.