Volatility decay: don’t hold leveraged ETFs long-term

Mar 10

Last updated 9/15/21. Check out my follow-up article Leveraged ETFs and volatility decay revisited for a counter-argument.

If you browse financial blogs, forums, and chatrooms, you’ll eventually find tongue-in-cheek memes that “stocks only go up” everywhere you look. And the funny thing is that in aggregate, and over the long-term, it seems to be true. For example, let’s take a look at the past 95 years of the S&P 500:

The market crashes in 2000 and 2008 are pretty obvious, but you’d have to look carefully to even identify any other events of lasting consequence on this timescale. And if the recent trajectory of the S&P 500 makes you feel uneasy, don’t worry. This is exactly what compound growth actually looks like. If you’re not convinced, I’ve recreated the chart with a logarithmic Y-axis. In this view, there doesn’t appear to be anything terribly abnormal about the past 10 years. Perhaps we are slowly heading towards another bubble, but perhaps not. By the way, the trendline shows about 6% CAGR, but its worth repeating that the price of the index itself doesn’t fully capture its total returns. When dividends are reinvested, the S&P 500 has historical CAGR of about 10%. Sure, it seems like we have bubbles and crashes cyclically, but the overall impression is unmistakable - stocks do always go up. And while nobody knows whether or not they will keep doing so in the future, the belief that they will forms the foundation of investing. Maybe its a self-fulfilling prophecy.

A passive investor can capture the returns of the S&P 500 by simply investing in any number of S&P 500 index funds or ETFs. And in fact they do - these funds have trillions of dollars of assets under management.

A natural question then arises: if we are so sure about the S&P 500 going up in the long run, why not borrow money to invest in it? If we could borrow money on better terms than the S&P 500 returns, we’ll come out ahead. The idea of borrowing money, also known as using leverage, is probably as old as investing itself. Leverage allows us to amplify the returns on investment, but of course, any potential losses are amplified as well. Use of leverage is very common in real estate investing - why pay full price for a rental property when you can pay 20% down instead? Your potential returns are amplified 5-fold.

Your bank probably won’t approve of a personal loan to buy an S&P 500 ETF like VOO, but many brokerages do allow margin trading, which is essentially the same concept. However, there’s an easier way for a retail investor to leverage without using margins or options, and that is through the use of leveraged ETFs. UPRO is one example of a triple-leveraged S&P 500 ETF. Through the use of derivatives and other methods, the ETF does the work of leveraging for you. The result is a product that attempts to track the S&P 500 but triple its movement - if the S&P 500 goes up 1%, UPRO goes up 3%. Sounds simple and enticing.

If you follow the movement of UPRO alongside the S&P 500 on any given day, you’ll see that it does work. For example, on January 6, 2021, the S&P 500 ended up 0.57%. UPRO, on the other hand, followed the S&P 500’s trajectory but amplified its movements, and ended up 1.81%. Note that because of the complexity of leveraged ETFs, they can never perfectly replicate a multiple of the index’s performance, but they usually get pretty close.

S&P 500 on Jan 6, 2021. From Apple Stocks app.

UPRO on Jan 6 2021. From Apple Stocks app.

Of course, leveraged ETFs are active products with high expense ratios. UPRO has an expense ratio of 0.93, which is quite high compared to VOO at 0.03. The reason for such a high expense ratio is that leveraged ETFs incur significant fees from daily rebalancing and interest and transaction fees. But isn’t the expense ratio worth it if UPRO can triple the S&P 500’s performance? In fact, since stocks will almost certainly go up in 30, 20, or even 10 years, shouldn’t long-term investors all invest in a leveraged ETF instead?

The answer is a resounding NO. Leveraged ETFs are designed for short-term trading. Due to a phenomenon called volatility decay, holding a leveraged ETF long-term can be very dangerous. This is the case even with a hypothetical “perfect” leveraged ETF which incurs no expense ratio and perfectly replicates 3x the index every day!

Let’s look at a hypothetical index, the A Frugal Doctor (AFD) index, that starts at 1000. Over the next week of trading, the index posts small but consistent gains daily and ends the week at 1011. Let’s also have a hypothetical ETF, the triple-leveraged AFDx3, which perfectly amplifies every daily movement by 3x and ends the week at 1034. In this scenario, triple leveraging worked as expected.

With a non-volatile investment, leveraging works exactly as expected.

More realistically, however, stocks are volatile and the index fluctuates from trading activity. It gains 3% one day and loses 7% the next, and so on. After a week of trading, the index still ends at 1011. Let’s see how AFDx3 performed compared to AFD itself under these circumstances.

When an investment is volatile, leverage can backfire.

As you can see, although AFD ended up overall for the week, AFDx3 underperformed the index and ended in the red! The amplification of both gains and losses makes AFDx3 much more volatile than AFD itself, and can result in a sequence of returns where the index ends positive but the leveraged version does not. In fact, because any loss is so devastating when amplified, the leveraged ETF can only reliably beat the index if volatility was minimal, and even when it beats the index, it is unlikely to return 3x of the index itself. And remember that AFDx3 is a “perfect” triple leveraged ETF with no expenses! Also, consider the downside risk: if a normal index ends down 34%, you still have 66% of your portfolio value left for a future recovery. But a triple leveraged product will be wiped out completely.

To look at some real world examples, let’s go back to the S&P 500 index and the triple leveraged UPRO. We saw that on any given day, UPRO can amplify the S&P 500 by 3x. How about over the course of 1 year?

As you can see, the S&P 500 peaked on February 19, 2020, crashed, then quickly recovered its losses in March and ended the year significantly higher than it began. It also eclipsed the previous peak on February 19 prior to the crash by August. UPRO, on the other hand, did a remarkable job of triple leveraging gains from the bottom back up, but ended 2020 barely above where it opened the year, and did not surpass its previous February 19 peak until January 2021.

The moral of the story is that while leveraged ETFs sound good on paper, volatility decay affects their performance in unexpected ways. Even for a long-term investor who is certain that “stocks only go up”, it is almost always better to simply buy the index fund itself. More broadly, using any type of leverage on investing can be dangerous for the unwary. As Warren Buffet once said about leverage: “When leverage works, it magnifies your gains. Your spouse thinks you’re clever and your neighbors get envious. But leverage is addictive. Having profited from its wonders, very few people retreat to more conservative practices. And as we all learned in third grade - and some relearned in 2008 - any series of positive numbers, however impressive the numbers may be, evaporates when multiplied by a single zero. History tells us that leverage all too often produces zeroes, even when it is employed by very smart people.“

Happy investing!

Update 9/15/21: As mentioned, I am aware that there is research which shows that leveraged ETFs have in fact beaten the index in a “buy and hold” strategy over the past 10 years or so. I discuss this and more in my follow-up article: Leveraged ETFs and volatility decay revisited. I strongly encourage you to check it out!