Leveraged ETFs like the SKF, UYG, SRS, FAS, FAZ, TBT, DIG and DUG have become very popular trading vehicles.There has been some debate among traders on how to use technical analysis on leveraged ETFs. Some traders argue that technical analysis works the same way on the leveraged ETFs as it does on the non-leveraged ETFs. Others argue that the unleveraged ETF or index corresponding to the leveraged ETF is the better tool.
In this article I will review some topics of relevance when using technical analysis on leveraged ETFs. Henceforth, I will refer to leveraged ETFs as LETF and the underlying ETF or index as the underlying instrument, UI.
Characteristics of Leveraged ETFs
When viewing a LETF I first look at the following characteristics:
1.Duration of compounding: Until now almost all LETFs compound daily; i.e. they aim to achieve the leveraged return based on the previous day’s closing price with the base being reset every day. There is some talk of LETFs which compound on longer durations in the pipe-line; however in this article I will assume daily compounding.
2.The amount of leverage: Most LETFs offer 2x or 3x leverage
3.Trading Volume versus the Underlying Instrument (Follow vs. Lead): In some cases the trading volume on LETFs is not big enough to affect the behavior of the UI; the LETF follows the UI. In some other cases there is speculation that LETFs have become the trading instrument of choice and their price action affects the UI; the LETF leads the UI.
Price Behavior of LETFs vs. the UI: Path Dependency of Results
The fundamental reason behind the debate on technical analysis of LETFs is the fact that LETF returns are path-dependent. In a window covering many days, the final value of the LETF will be affected not only by the absolute movement in the price of the UI, but also by the path taken by the UI to reach that value. So a roundtrip in price of the UI will not only, not translate to a round-trip on the LETF, but the value for the FETF will be different for different price-paths.
For example, suppose a 2x LETF is trading at $100, as it the UI. If the UI moves up by 2% to 102 by the close of the day, the LETF will close at $104. If on the next day the UI moves down to $100, the LETF will close at $99.9216. However, if the price of the UI first went down to $98 and then back to $100, the LETF will close at $96 first, and at $99.9194 next.
Non-Trending Markets: Volatility Decay
The skew between LETFs and UI returns leads to what some call the volatility decay on LETFs when the markets are not trending but trading in a range. The roundtrip example above illustrated that. When the process continues over a larger time-period, this decay accumulates.
In the above example, if the UI completed three roundtrips (100->101->100), the LETF’s value will fall to 99.94. If the size of the oscillations is large, the effect becomes even worse. So three roundtrips to 102 (100->102->100) will reduce the LETFs value to 99.765, a 23bp loss in just six days. The effect is even more pronounced in LETFs with 3x leverage which will have a value of 99.296, a loss of 70bp after three round-trips. The chart below shows the decay over six round-trips.
Path Dependence: When is it Relevant?
The path dependence of LETFs becomes relevant when the technique you are using relies on absolute price levels. Support-Resistance Levels, Gap Analysis, price targets based on patterns, Fibonacci retracements and extension levels, all rely on absolute price values. The behavior here gets significantly altered.
On the other hand, if the technique you are using relies on relative price movements over a given time-frame the path-dependence of the LETF becomes less important. This is typically true of most price-patterns and oscillators among others. Note that the type of pattern observed may change; or the settings of the oscillator and their interpretation will change from the LETF to the UI.
In the next sections, I will address some specific topics of interest.
Time Window of Analysis: Intra-Day versus Multi-Day
If the TA is being performed exclusively within the time-window between the reset of the compounding events (i.e. intra-day), then almost all indicators, including those based on absolute price levels will work. Intra-day support-resistance, trend-lines, chart-patterns on LETF will provide the same information as their counterparts on the UI.
Note that calculated indicators (e.g. oscillators) will have to start with a clean slate at the start of the reset window to get a pure and true result; in practice this level of absolute perfection may not be necessary for the trader to make a decision.
Volatility Decay and Trend Lines
Volatility decay means is that the chart patterns observed will vary between LETF and the UI. The LETF has an inherent negative bias and the price will shift downwards. As a result during periods of consolidation the horizontal support and resistance lines get converted into downward trend lines. In the example earlier, where the UI makes multiple round-trips to the same starting price, the UI will form a horizontal ledge pattern, but the LETF will form a descending channel.
Adjusting Technical Price Levels
In the above example, a trader looking for a breakout will have to consider different price-points. On the UI a break above the upside resistance (102) will constitute a buy point. In the case of the descending channel, the breakout buy point is not that obvious. Is it the highest point of the channel or a break above the previous swing high, prior to the break above the channel?
Volatility Decay and Gaps
Many traders consider price gaps in charts as strong signals, and use the price levels of the gap as key trading points. Due to the decay associated with the LETF, a downside-gap fill on an UI will often not translate into a gap fill on the LETF, since the LETF would still not have recovered to that level. LETF may show an upside-gap fill simply due to volatility decay, when the UI is still trading above the upper side of the gap.
This phenomenon is very well illustrated by comparing the TBT the Ultra-Short Long Bond fund, and the Yield of the 30 Year Bond (TYX). These are good a pair since the price of the long bond varies inversely with its yield. Further the bond market is huge and technical trading in bond ETFs is unlikely to have a meaningful impact on the bond yields.
The TYX gapped down on multiple occasions in late November of 2008. These gaps also appeared in the TBT. The TYX filled the gaps in late April and early May. However, even after the massive rally in TBT this week, it still has not filled all the gaps. TYX though is now at a much higher level.
During the period between February and mid April of 2009, the TYX were range bound and oscillated in a flat channel. During the same period the TBT traded in a downward sloping channel as the volatility decay reduced its value. Any trading decisions based on gaps on the TBT would have been against the spirit of gap analysis.
The Cart before the Horse Hypothesis
There is some talk in the media about how in certain sectors, trading in LETFS is having a significant impact on price movements. If the trading in the LETF becomes the primary driver of price movements, then pure technical analysis on the LETF may become significant. However, most of these hypotheses are primarily based on intra-day price movements which make the issue moot.
So Does TA Work on LETFs?
Without doubt, there is merit to chart reading and technical analysis on LETFs. However, the interpretation of these charts has to be adapted to account for the specific peculiarities introduced by the volatility decay inherent in their structure.
Disclosure: Vikram actively trades many of the ETFs mentioned in this article.