ETF Complete Strategy Insights: The Ins and Outs of Leveraged ETFs (Part 2)

James Kimball | October 13, 2019

The ETF Complete model closed the week down -0.3% compared to the SPY which closed up 0.7%.

Stay tuned for the daily updates and log into the website to see holdings and additional performance data.

image

This Week's Strategy Lesson: The Ins and Outs of Leveraged ETFs (Part 2)

image

Last week we started a series looking at the factors that can affect leveraged ETF performance. The relationship between percentage gains and losses is one of the key factors that can cause some of the “decay” in leveraged instruments. We covered this in detail in the first part of the series. It may be helpful for some to review that article.

As we saw in the first part, the loss/recovery percentage relationship has an interesting effect on leveraged instruments. The percentage gain needed to offset a loss increases and an increasing rate the larger the loss. Simple 2x or 3x leveraged instruments do not account for this effect causing “decay” in certain market conditions.

The amount of the decay or rate of decay is determined by three key factors: The number of periods, the amount of the leverage, and the volatility of the underlying instrument. In the first article, we demonstrated how the amount of leverage and number of periods make a difference. Here we are going to show the effects of the third factor, volatility, using another experiment.

In this scenario, we are going to use a random number generator to generate two different groups of 1000 numbers based on a normal distribution (bell curve). Both groups will have the same average value of zero but one will have a standard deviation of 0.5% and the other will have a standard deviation of 1.0%.

(For comparison, the standard deviation in the SPY currently is around 0.45%)

Or stated another way, both groups of numbers should average out to about zero, but the group with the larger standard deviation will tend to have larger positive and negative numbers while the group with the smaller standard deviation will tend to have smaller positive and negative numbers. This experiment allows us to hold everything constant except volatility so that we can isolate its effects. We would expect the group with higher standard deviation (volatility) to show more decay than the other group.

image

In the chart above we run the smaller 0.5% standard deviation scenario. We picked out a result where the “no leverage” symbol basically ended the 1000 days flat. The “3x leverage” took the same daily numbers and just multiplied them by 3. It ended down -7%. The “5x leverage” ended the period using the same daily number multiplied by 5 at -22%.

image

If we repeat the same process to generate another list of 1000 random numbers with the same zero average but now with larger 1.0% standard deviation, we can see the direct effect of the increased volatility.

In this scenario, the “no leverage” based on the random numbers happened to end the period mostly flat, down -2.8%. While, off the same baseline numbers, the “3x leverage” ended down -30% and the “5x leverage” ended down -65%.

While this isn’t a perfect “apples to apples” comparison, the increased volatility of going from 0.5% standard deviation to 1.0% standard deviation in our two scenarios increased the “decay” in the “3x leverage” from -7% to -30%. For the “5x leverage” it went from -22% to -65%.

We intentionally used large values in terms of the amount of leverage, the amount of volatility, and the number of compounded periods to help demonstrate the underlying mechanics. Scenarios with small differences in these values can make it harder to spot the effects. You have to use some combination of lots of leverage, or lots of volatility, or lots of periods to get these drastic differences.

Different Effects in Different Markets

So far we have looked at how the application of volatility can affect decay rates in scenarios where the market ends mostly flat. However, when we have directional markets, the leveraged daily compounding can have different effects.

In down trending markets, the leveraged compounding tends to reduce losses to lower than what you might expect with a simple multiple applied to the final return in the non-leveraged portfolio (though you still have to be careful about leveraged portfolios having exaggerated or critical drawdowns that can’t be weathered while the drawdown in the non-leveraged portfolio is quite reasonable).

In up trending markets, the leveraged daily compounding can dramatically increase our return over a simple multiple of the non-leveraged return value.

Next week we will walk through a few of these examples to see the real-world positive and negative effects of leverage and compounding.