Leveraged ETFs Aren't the Cause of Increased Market Volatility

Since August, stock market volatility as measured by the VIX has maintained its highest average since 2009. At the same time, the number of leveraged and inverse ETFs has exploded to 249 products.

There are good reasons to suspect that the rise in volatility might be related to these types of ETFs. First of all, leveraged ETFs use derivatives, those notorious weapons of financial destruction. Secondly, they are much more volatile than the underlying indexes that they seek to track. Both leveraged long and short ETFs need to rebalance their portfolios daily by trading in the same direction as the market, potentially exacerbating market movements. Additionally, many investors were caught off guard by the impact of the daily reset on compound returns. Although it is difficult to disprove, we can point to several factors that suggest leveraged and inverse ETFs are not to blame and can offer a few plausible explanations for the increase in volatility based on economic fundamentals.

As my colleague Paul Justice outlined back in 2009, leveraged ETFs can destroy the portfolio of the buy-and-hold investor. We won't rehash how these funds work here, but suffice it to say that the volatility of leveraged funds is greater than that of the underlying index they seek to track. For example,  ProShares Ultra S&P 500 (SSO), which attempts to provide twice the daily return of the S&P 500, had a volatility of return twice that of the S&P 500.

However, leveraged and inverse ETFs have just $32 billion in assets, which is less than the $83 billion in the  SPDR S&P 500 (SPY) ETF alone and just 3.2% of all U.S. ETF assets. But ETFs aren't the only securities in the market; there is another $300 billion in S&P 500 futures and $10 trillion in mutual funds. It is hard to imagine that such a small segment of the market could impact market volatility. Admittedly, leveraged and inverse ETFs do trade more than the average ETF, and at 14%, they account for a disproportionate share of ETF volume. But just as with stocks, most of that volume occurs between individual investors--so-called secondary market transactions--without as much impact on price as a primary market transaction.

If leveraged and inverse ETFs were amplifying market volatility, we would expect that assets would ebb and flow, causing peaks in volatility and but then crashing back to a trough. However, assets in leveraged and inverse funds have been flat as a board, consistently hovering between $30 billion and $36 billion since March 2009. This is despite the fact that the number of leveraged and inverse offerings nearly doubled in that time frame. During a particularly volatile time between Nov. 15 and Nov. 25 2011, the S&P 500 was down seven days in a row for a cumulative loss of 8%, but total assets in leveraged and inverse products stayed precisely at $34 billion. This stability flies in the face of the volatility argument.

Leveraged funds need to rebalance daily, forcing them to trade in pro-cyclical fashion in the same direction as the market. So if the market is up during the day, both a long and short fund will need to buy, potentially forcing the market up even more. While same-way trading is correct in theory, in practice the pattern is much more benign. The trading of individual investors mitigates the impact of same-way trading. For example, on Aug. 8, the S&P 500 fell by 6.6%, and SSO, which attempts to deliver twice the S&P 500 return, was down by 13.3%. A move of this magnitude would require a sizable rebalance for a single ETF, forcing SSO to sell securities in a down market. But the shares outstanding in the fund actually increased, reducing the impact of the rebalance. While investors bought SSO, they sold its leveraged inverse counterpart,  ProShares UltraShort S&P 500 (SDS), mitigating the size of the pro-cyclical rebalance of that fund.

There is another hole in the argument that procyclical trading causes volatility. If it was, we should be able to predict the direction of the market in the final minutes of trading based on how the market has moved earlier in the day. Looking at return data since August, the return prior to the close had little to no predictive power for the final 15 minutes of trading. So if the market was up, during the final minutes of trading it was just as likely to go up as it was to go down with an R-squared of only 1.6%. This indicates virtually no predictability and thus no causation.

Conspiracy theorists have often suggested that derivatives or short selling causes market volatility, but academics who have researched the issue have found scant evidence. In fact, some even suggest their existence reduces volatility. If activity in the ETFs were causing volatility, it should be able to predict changes in VIX, a measure of expected volatility. We ran a model where we attempted to forecast tomorrow's level of VIX based on today's level, and trading volume in both the S&P 500 (proxied by SPY) and SDS. What we found is that higher volumes in SDS actually resulted in lower levels of VIX, although the results was not statistically significant.

Although leveraged and inverse ETFs are not responsible for increased market volatility, there is some evidence that fundamentals are to blame. Because the stock market is forward looking, I used earnings growth one year out to predict VIX. This regression produced an R-squared of 31%, suggesting that earnings drive volatility. Previous spikes in the VIX, such as in 2001, were associated with earnings volatility, while lulls in the VIX during the mid-1990s and mid-2000s were associated with stable earnings. 

Not all ETFs are simple, buy-and-hold type index funds, and Morningstar is not a blind industry cheerleader, but an advocate for greater transparency and education with regard to opaque issues such as synthetics and securities lending. While traders and hedgers might find a use for leveraged and inverse ETFs, we recommend that investors avoid them. If you have a bullish view on a sector, rather than using a levered product, consider shifting your asset allocation. Despite our recommendation to avoid leveraged and inverse ETFs, we do not believe that they are causing an increase in market volatility.