Leveraged ETFs

A lot of nonsense is written about leveraged ETFs (ETFs that usually provide 200% long or short exposure to a given index) on the internet. These funds perform exactly like a margined portfolio of the unleveraged ETF would be expected to perform. In other words (on the long side) two times the index minus the interest cost of leverage minus the expense ratio (which is typically higher than for an unlevered ETF). This means that in the short term they tend to track two times the index very closely but in the long-term drift away from the index due to the interest cost and expense ratio generating a negative alpha. The Proshares annual report gives interesting information on how these portfolios are constructed.

The short funds exposure is achieved through a mix of swaps (similar to CFDs) and futures contracts. The long funds though also hold actual shares. For example, QLD's (I have 200 shares) exposure is achieved 85% through actual shares, 14% through futures contracts, and 100% through swaps. The money not invested in shares is invested in repurchase agreements with investment banks - in other words they lend their cash to these investment banks in return for US Treasury and Agency securities collateral. These provided margin for the futures contracts. The equities were partly used as collateral for the swap agreements.

Looking at the income statement, QLD received $3.6 million in dividends and $4.4 million in interest. Total expenses were $6.9 million. The interest cost of the swaps is apparently capitalized into the swap values. My only puzzle is why not construct the equity market exposure 100% out of swaps and/or futures? I'm guessing it's for tax reasons. Capital gains on stocks removed from the index can be distributed at the long-term rate while distributions from dividends earned may also be taxed at the lower US qualified dividend rate rather than the regular rate that would apply to interest. I could be wrong though. It seems that net gains on futures do not have to be distributed.