The Truth About Leveraged ETFs transcript

$10,000 invested in the ProShares Ultra Pro S&P 500 3X leveraged ETF when it launched in 2009 would be worth $1.3 million today, more than 10 times a regular S&P 500 ETF. Leverage, using any form of financing to obtain asset exposure greater than your own equity, can play a sensible role in asset allocation. It can also be a disaster. Leveraged ETFs are a straightforward way to get leveraged exposure to stocks. I can't tell you whether they make sense for you, but I will give you a framework to decide for yourself.

I'm Ben Felix, Chief Investment Officer at PWL Capital, and in this video I'm going to tell you when leverage makes sense, at least in theory, and whether leveraged ETFs are a good way to get it. I haven't nerded out like this in a while. Feels good. While it's normal to borrow a lot of money to buy a house, a mortgage, leverage has an overall bad reputation when it comes to the stock market, and honestly for good reason.

It's often used irresponsibly by overconfident traders, which leads to bad outcomes, but I don't think it needs to be dismissed as a tool entirely. Leverage amplifies the amount of exposure that you get to an asset for each dollar that you invest in it. One of the most straightforward ways to get access to leverage is using leveraged ETFs. For example, investing $1,000 in a 2X leveraged index ETF gives you exposure to $2,000 of the index's daily returns.

Leverage also increases volatility, and it can increase or decrease returns depending on the cost of leverage and on what the underlying asset does. This makes for a wild ride, but it can have a lot of upside. Whether leverage makes sense for your situation specifically is similar to asking whether you should invest in stocks or in bonds. I can't give you the answer, but I will give you a framework for thinking about it.

Then, given the decision to use leverage, whether leveraged ETFs are a good way to obtain it is the next question I'll address. One of the strongest cases for leverage is that, at least in theory, young people with lots of human capital, that's lots of future earnings potential from working, and low financial wealth, meaning they don't have much money to invest today, should be borrowing to invest because their human capital is like a big, safe bond.

The idea is that if, and that's a big if that I'm going to come back to you later, if your human capital is a safe bond-like asset, and you have low financial assets, meaning investments in things like stocks and bonds, your overall asset allocation early in life is concentrated in safe assets, in your human capital, while lacking exposure to risky assets like stocks. It then follows that you should be borrowing money to invest in stocks.

This idea was given its formal footing by Paul Samuelson and Robert Merton in 1969 and 1971, respectively, then further refined in a 1992 paper. The authors of that paper treat human capital as an asset in the portfolio and showed that a young worker with little financial wealth, whose future earnings are relatively safe, optimally uses leverage to reach their target equity exposure. This effect is even larger if their labor supply, that just means their ability to earn income by working, is flexible.

You can imagine a situation where the stock market crashes, but you can pick up some overtime hours to keep your income stable. You can therefore take more risk, increasing your optimal use of leverage in the model. They also show that the more your income is correlated with the stock market, the less leverage the model is going to call for. A lot of this comes down to a simple question posed in the title of a book by Canadian finance professor and researcher Moshe Milevsky, are you a stock or a bond?

If your human capital is stock-like, you probably want to take less risk in financial markets all else equal. If it's bond-like, you can afford to take more risk. While this can all sound exciting, and I know from talking to people that young investors will hear this message and want to get their hands on as much leverage as possible, the assumption of safe human capital needs to be interrogated. One potentially critical issue to point out here is that young people specifically may have riskier human capital.

There's also evidence that many high-income households' wage and business incomes do decline at the same time as the stock market, and that they sell some of their liquid assets, including their stock holdings, to make up the income shortfall during these bad times. This leads to them selling stocks after they have fallen in price, which is not good. The intuition here is very simple. Households have large fixed expenses that are hard to adjust.

Think like mortgage payments or rent or car payments, daycare costs. If your income falls, you can imagine selling some of those assets to fill the shortfall. You're not going to not pay your mortgage or stop paying for daycare. The evidence for high-income US households suggests this is exactly what often happens. With that caution in mind, relying on the assumption of safe human capital, which I want to emphasize once more is a big assumption worth thinking carefully about before applying the information, the 2013 paper Diversification Across Time argues in favor of leverage for young investors.

For practical reasons, the authors suggest using 2x leverage early on, declining over time as you near retirement. To illustrate the benefits of this approach, they compare 2x leverage stocks declining to 50% stocks and 50% bonds at retirement to a portfolio with a non-leveraged constant lifetime allocation of 74% stocks and 26% bonds. The reason for that specific benchmark allocation is that in their back test it produces the same mean retirement accumulation outcome as the leveraged strategy, but has a higher standard deviation of retirement wealth.

I do want to say again that there are multiple cautions and warning labels that come with this finding. The research assumes that human capital is a bond-like safe asset. But, as I mentioned earlier, real human capital can, in some cases, behave more like a stock. And this may be especially true for young people. The paper does acknowledge this to an extent. The target allocation in the paper is defined by the Merton Samuelson share.

That comes from the papers I mentioned earlier. The Merton Samuelson share is the ratio of the expected equity risk premium divided by the expected variance of stock returns multiplied by the investor's constant relative risk aversion. I warned you guys, we're nerding out in this video. This can get messy fast in practice. We don't know what future returns or the variance of future returns will be, and defining your constant relative risk aversion can itself be a noisy exercise.

If you are highly risk-averse, your optimal equity allocation is lower, and your need to use leverage to reach your target allocation is reduced. Depending on your expectations and your risk aversion, your target asset allocation could swing wildly. Kind of like how your hand should swing wildly toward the subscribe button and click it if you haven't already. I'll link to a web app that we built based on a different academic paper that takes risk aversion as an input, so you can see the effect on asset allocation.

That paper did not allow leverage, so neither does the app that we built, but you can see how sensitive the equity share is to risk aversion. Another recent working paper models the life cycle of an American household and runs a million simulations rooted in long-term historical data to identify optimal life cycle asset allocation strategies. This model does include labor income risk as part of the optimization process.

When the leverage is allowed, and the cost of borrowing is 1.4% above the risk-free rate, which is a realistic cost of leverage for most households, the household optimally applies 2x leverage early on, decreasing over time while investing in a portfolio of approximately 34% domestic stocks and 66% international stocks. I know that the theoretical and data-driven case for leverage can sound very compelling. But the other thing to keep in mind is that leverage can result in a lot of volatility and total wipeouts of your financial assets.

That is objectively not very fun to live through, and while it may still look good in a model, even if you include the fact that there can be wipeouts, I don't know that most investors will get right back on the investing horse after a total wipeout of their assets. I hope I've made it clear that leverage can be useful on paper, in theory, when it's used the right way for the right situations, but it comes with risks that you need to be aware of before committing to it.

You also need to be able to live with the volatility. A 2x leveraged S&P 500 ETF fell nearly 80% in the Great Financial Crisis. A 3x leveraged S&P 500 ETF fell nearly 80% at the beginning of COVID. We didn't know at the time that stocks would recover the way that they did after the drops. Those are serious psychological events for an investor in these products. As journalist Jason Zweig described it, buying a leveraged fund and holding it for the long run is appropriate only for people who can spend the rest of their lives in a sensory deprivation tank.

It's not even just about behavior in extreme short-term downside events, though. There is risk of long-term underperformance, too. In markets that have not performed as the US, like emerging markets, leveraged ETFs have made you worse off after volatility drag, which I'll talk more about in a minute, fees, and financing costs compared to just investing in the underlying assets. Even looking a little further back in US market history, a live 2x leveraged S&P 500 mutual fund underperformed an S&P 500 ETF, non-leveraged, for decades before pulling ahead only recently, and that was due to a combination of catching the dot-com bust, the fund's slightly higher fee compared to the ETFs we have today, and higher financing costs through its earlier history, which is something I'll come back to later.

This stuff is not for the faint of heart, in the short run or the long run. I think people look at the more recent returns of US equity leveraged ETFs and want that performance. I don't need to tell you guys that you can't buy past performance. It's not all bad news, though. Everything I've talked about so far applies to leverage, generally speaking, but there are lots of different ways to get leverage. You can borrow from your broker with a margin loan, you can borrow from a line of credit from a bank, you can use derivatives like futures or options, and there are probably lots of other ways that I'm not thinking of.

Each means of obtaining leverage has its own trade-offs. I'm going to continue to talk about leveraged ETFs, but to be clear, I'm not endorsing them specifically or arguing that they're the best way to get leverage exposure to stocks. Leveraged ETFs apply leverage to an underlying asset for you inside of an ETF wrapper, often using swaps and or futures. They offer convenience, simplicity, and non-recourse, which I'll explain in a second.

You buy an ETF. That's it. Leveraged ETFs are also available inside of account types where margin and other forms of leverage are often restricted. Non-recourse means you can't lose more than you put in, which is not always the case with leverage. The fund can still have problems in extreme markets, but you are not managing the leverage ratio in the fund, dealing with margin calls, or cleaning up the mess in an extreme situation with a bad outcome, which do happen.

You're also getting institutional financing costs, but you are paying a higher fee to get access to that. The big thing that I want to talk about, though, is something that I have been wrong about in the past, the long-term effect of maintaining constant daily leverage in leveraged ETFs, often referred to as volatility decay. I'm not alone in having raised this concern, which I did in a in a video a while ago. US regulators have given similar warnings to investors.

A 2x index ETF aims to give you twice the daily return of the underlying index net of fees and costs. If a fund has 2x exposure to the underlying index today and the index delivers a positive return, the fund's assets will increase in value pushing down its leverage ratio. Since the fund is promising 2x daily leverage, it will need to buy more stock exposure to push its leverage ratio back up. It would similarly need to reset if stocks go down causing the fund to sell stocks after they fall.

This has been used to argue against daily leveraged ETFs for long-term investors, but it's incomplete. Okay, math time. At the risk of getting too nerdy even for this channel, which is a stretch I know, we can see in a mathematical approximation of volatility drag that the cost of volatility to returns is insensitive to rebalancing frequency. In this equation, drag is the return drag between L times the underlying return and the leveraged portfolio's return.

N is the number of rebalancing periods, L is the leverage ratio, and sigma squared is the per period of variance of the underlying asset. If we rebalance daily, N is 252 trading days in a year, and sigma squared is the daily variance of the underlying asset. If we move to a weekly rebalancing just as an example, N decreases by a factor of five, but sigma squared, which is now the weekly variance rather than the daily variance, also increases by a factor of five.

The change cancels out. This is only true if there is no serial correlation in returns, meaning if returns in one period are unrelated to returns in the next period. If there is serial correlation, per period variance will not scale with the number of days in the period. With positive serial correlation, more frequent rebalancing is actually good. If serial correlation is negative, more frequent rebalancing is bad. The main insight is that it depends on the specific return outcome that you get, which is not something we can know in advance, and therefore not something that should make you conclude frequent rebalancing is a bad thing in advance.

It really becomes an empirical question whether daily resets with actual stock returns result in a value destruction. A 2026 working paper focused on single stock leveraged ETFs looks specifically at the claim that daily resets lead to under performance relative to a static leveraged portfolio. The author explains, as I just mentioned, that daily rebalancing can work against you in a specific realized return outcome, but not in all outcomes.

They focus on single stock leveraged ETFs and find empirically that daily resets in this setting, the single stock leveraged ETF setting, do result in a cost over their benchmark due to the specific return paths of individual stocks in their sample. Of course, single stocks are typically far more volatile than stock indexes. The natural question is how diversification changes the result. A 2022 paper asks whether investing in an index fund on margin and rebalancing once per holding period or using a leveraged ETF with daily resets would have produced better outcomes in historical data.

The authors find that leveraged ETF returns were higher during strong bull and bear markets, but lower during moderate bull and bear markets. They also find that leveraged ETFs provide reduced downside risk and are less risky than margin during every bear market since 1927. A 2025 working paper, notably funded by ProShares, which is a company that issues leveraged ETFs, but written by an academic, looks at multi-day leveraged ETF returns and finds that long leveraged index ETFs align closely with portfolios without daily rebalancing at holding periods up to a quarter and reasonably closely out to a year.

When their performance does diverge at the extremes, sizable underperformance is less likely than same-size outperformance. The author concludes that daily leverage resets in leveraged ETFs are not the problem they have historically been made out to be. A 2026 working paper studies five determinants of leveraged ETF performance: volatility drag, fees, financing costs, which products are launched, and market timing. To be clear, volatility drag is a real thing, and leverage increases volatility drag because it increases volatility.

That is distinct from any rebalancing effect, which is a point that the authors of this paper do note. The authors illustrate how volatility drag scales with volatility for 2x leveraged ETFs. You can see how each increase in volatility drives a larger wedge between the leverage return and the underlying asset return. This is related to a concept called the break-even return, or the return an underlying asset must earn in order for a leveraged strategy holding that asset to earn at least the same return as the underlying.

They give the example that for an underlying asset with daily volatility of 5%, a 2x leveraged ETF requires an annual return of nearly of the unlevered asset. That doesn't even include financing costs, which would push the break-even return up even higher. They find that leveraged index ETFs have been beneficial to investors, and while some single stock leveraged ETFs have done well, they face higher borrowing costs, high volatility drag, and poor timing by investors.

Separate research from Morningstar shows single stock leveraged ETF investors have been more likely to lose over 75% of their investment than to outperform the underlying stock outright. And that 19% of single stock ETFs have lost over 75% since inception, while only 18% have outperformed the stock that they track. The authors show that expense ratios for equity index and single stock ETFs are around 1%, which is not surprising.

But they highlight another important data point that's easy to gloss over. They estimate financing costs, how much it costs the fund to get its leveraged exposure to the underlying asset. This is important because the expense ratio does not capture the total costs. The embedded financing rate on a swap contract, for example, won't show up in an expense ratio. The 2x leveraged index ETFs in the sample have financing spreads 0.5% above the risk-free rate, but that funding spread is on top of the risk-free rate.

When you add up the funding spread, the risk-free rate, and the fund's fees, the total cost of owning a 2x leveraged index ETF could easily be above 5% per year at the current risk-free rate. For the life history of many of the popular leveraged index ETFs, this number was a lot lower, further boosting their performance. Single stock leveraged ETFs have market cap weighted financing spreads nearly 10 percentage points above the risk-free rate, which is crazy.

Putting the total cost of a 2x leveraged single stock ETF close to 12%. You can certainly see how the headwinds for single stock leveraged ETFs in particular add up. Another 2026 working paper directly defends leveraged index funds for long-term investors. The authors explain that daily rebalancing produces convex payoffs that differ from, but do not underperform linear benchmarks. And they show empirically that daily leveraged rebalancing is not an inherently value-destroying activity.

They conclude the demonization of leveraged and inverse equity index ETFs, particularly the false accusation that these funds erode value via volatility decay, is a major disservice to investors that has likely discouraged long-term investors from using these products in ways that would help them achieve their financial goals. The other thing is leveraged ETFs are not all or nothing. You could use leveraged ETFs alongside regular index ETFs to target your overall desired leverage level.

Just because 3x leverage exists in an ETF wrapper doesn't mean you need to go all in. For example, mixing 75% regular index funds with 25% 2x leveraged index funds gives you a more moderate 1.25x leverage. Note, in this case, you would need to manually rebalance your leverage back to your 1.25x target over time. Alternatively, there is a newish product here in Canada, which is 1.25x leveraged equity index funds. They use cash leverage and rebalancing bands rather than daily resets.

They have not existed for long, but they have delivered on what they set out to do. There's nothing magic about 1.25x leverage. It's just a product worth mentioning while I'm on this topic. Leverage is a double-edged sword. It amplifies positive and negative returns, increases volatility and volatility drag, and it can come with high costs, both financing costs and fees. In theory, it can make a lot of sense for some investors to use leverage if they can stomach it to increase their exposure to stocks beyond the amount of capital they have to invest.

It has been a long-standing commonly held belief by many, including me, that daily reset leveraged index ETFs are not a good tool for long-term investors. This belief is unfounded. Daily leverage resets do not cause volatility drag. Volatility causes volatility drag. And even so, leverage and leveraged ETFs have been useful for well-diversified long-term investors. Single stock leverage ETFs, on the other hand, which are being issued in huge numbers, are a lot harder to make the case for, at least in my opinion.

I've been calling them slop since earlier this year, and I think new research continues to support my slop thesis. If you think leverage might apply to your situation, but you are not sure whether you can stomach the extreme ups and downs, you can use PWL Capital's free psychometric risk tolerance assessment tool to investigate your risk tolerance.