talking point | the blog at

Leveraged ETF Arbitrage?

Posted in ETFs by teslik on November 5, 2009

In response to yesterday’s post (which, to my excitement, got picked up by FT Alphaville), a friend sent me the following chart. It shows FAS and FAZ, Direxion’s leveraged ETFs tracking, respectively, the daily performance of the Russell 1000 financial services index and the inverse of the same index. Both ETFs are 3x leveraged, and as you can see, they have both fallen significantly thus far this year — FAS by around 30%, and FAZ by nearly 100%.


My friend pointed out that there would be an easy arbitrage play here if you could simply short both of these ETFs at the same time. Unfortunately, while you can short an ETF, it would be difficult to short sell a leveraged ETF in practice, because you’d have to find somebody to lend it to you — and most brokerages serving retail investors won’t do this, presumably because they understand that the deal would likely net them a loss in the long-run.

One caveat:  even if you found somebody willing to lend you shares of both FAS and FAZ, so that you could sell them short, the success of this arbitrage strategy would still depend on the volatility of the market, as this post on SeekingAlpha points out. The author, Larry MacDonald, encourages anyone trying this strategy to seek out leveraged ETFs in the markets most likely to be volatile over time.


More Caution on Leveraged ETFs

Posted in ETFs by teslik on November 4, 2009

I’ve written about this before, but I feel it bears¬† more coverage, and with one notable exception (Izabella Kaminska at the FT) nobody seems to be paying much attention. The ETF market strikes me as incredibly ripe right now for abuse (or confusion, at very least), given public interest in investments that seek leveraged or inverse returns, or both. There is a flood of new synthetic ETFs that might strike retail investors as logical vehicles through which to do this, but they are not what they seem at first blush.

Here’s an example. The following chart from Google Finance shows the S&P 500 index over the past year (the red line), and next to it the ProShares Short S&P500 index over the same period (blue).


An investor buying such a product might logically assume that it would track the inverse of the S&P. You’ll notice in the chart that the S&P is up a little under 10% for these 12 months, so an investor might assume that the ProShares index mapping its inverse would be down roughly 10%. In fact, the ETF is down roughly 30%.

The discrepancy in returns comes, in large part, from how these sorts of ETFs work. Unlike buying a stock, which has a set value correlated to the value of a sliver of a company (i.e. it’s pegged to something that represents actual value, even if what that value is can be disputed), the NAV prices of ETFs are somewhat arbitrary. They “reset” after each trading day, and only seek to track their benchmarks for that day, not over the long-term.

Here’s what this means in practice. Say you buy one share of an ETF for $100/share. The ETF seeks to track the inverse of Stock Index A, which is trading at a NAV value of 100. On the first day, Stock Index A rises 10%, to 110. The ETF, assuming it tracks its daily benchmark perfectly, falls 10%, to $90/share.

Then, on day 2, Stock Index A gains another 10% (or 11 points — 10% of 110). The index closes the day at 121. The ETF, meanwhile, loses 10% of $90, or 9 points, and falls to $81/share.

Now, on the third day of trading, Stock Index A falls to 100. So it lost roughly 17.4% of its value. The ETF tracking the inverse gains the same percentage, or ~17.4%. But 17.4% of $81 is $14.06. So the ETF only rises to $95.06.

Result: The index winds up even after 3 turbulent days of trading. The ETF, however, has lost almost 5% over the same period.

The longer you hold a leveraged or inverse ETF, the more magnified this discrepancy becomes, even if the ETF is executing its daily objectives perfectly.

The key point here is that investors should be aware that these tools are not intended to map the inverse or leveraged returns of their benchmarks over long periods of time. This fact is easily lost on novice investors, particularly given that details about how ETFs actually work do not often play a prominent role in the advertisement of these products.

ETF Benchmarks

Posted in Charts, ETFs by teslik on August 19, 2009

Yesterday Izabella Kaminska wrote an excellent blog post for FT Alphaville on some of the not-so-visible risks associated with ETFs. Many ETFs, and particularly those trying to track commodity indexes or leveraged or inverse investments, have drawn heat of late for failing to effectively track their respective benchmarks. I had actually been looking at related charts a few days ago. I’m posting one below:


The red line is the S&P 500 index. The blue line is the ProShares Short S&P 500 ETF, which seeks to replicate the inverse of the S&P (before fees and expenses). The ETF has an expense ratio of 0.95%, so you can reasonably expect that it will lag the exact inverse of the S&P by about 1% per year, or perhaps a little more once you factor in the firm’s administrative expenses as it deals with dividend payments, taxes, etc. In fact, over the past year, the ETF has underperformed its benchmark by 36.1% (the S&P has fallen 23.26%, so its inverse has risen 23.26 percent; the ETF, meanwhile, has fallen 12.84%). If I bought an ETF seeking to track the inverse of the S&P, and the S&P lost over 23% in a year, I would be pretty annoyed if I also lost money (and significant money at that).

Chart via Google Finance