Inverse and Leveraged ETFs: Use with Care

There has been an explosion of leveraged and inverse leveraged ETFs on the market in recent years. Basically these Exchange Traded Funds are designed to give a daily return equal to a multiple of a specific index. Typical multiples are 2x or 3x (leveraged) and -1x, -2x or -3x (inverse).

For example, TBT, a -2x fund I have mentioned previously, seeks to give a daily return that is twice the inverse of an index of 20+ year Treasury Bonds. If the bond index falls by 2% on a given day, then TBT should go up 4% on that day. The problem is that this is designed to work per day. It doesn’t necessarily work out that way over the long term.

In a recent paper from Barclays Global Investors, the authors claim that these funds are not suitable as long-term investments. Of course, Barclays Global is a manager of ETFs and a competitor of companies that specialize in leveraged ETFs, so they are not completely objective. Nevertheless the article makes some important points. For those with a high tolerance for mathematical formulas, I recommend reading it in full.

The question that interests me is how these funds should be used by individual investors. Today’s markets are volatile enough for the average investor, so buying a leveraged fund just to increase potential profits is taking on unnecessary risk and maybe just being greedy. I am more interested in the inverse funds than the long leveraged funds, because the inverse funds provide an easy way to take a position against an index or to hedge an existing position, for which an investor would otherwise need to use derivatives or short the index fund.

Why would these funds not do what we expect them to do? For one thing because that’s not what they claim to do – they say that they only track daily movements and not long term movements. Nevertheless, investors expect them to work over time also, and their long term results should be the product of their daily results, so what are the reasons this might not work over the long term?

1. Expenses – these funds have higher expense ratios than traditional ETFs. However, the expense ratios are still usually less than 1%, so even over time this is not a major factor.
2. Daily re-balancing to re-align the fund with the multiple of the index the fund is tracking. Leveraged and inverse funds need to rebalance their hedges at the end of each trading day because their positions move with the index, not with the multiple of the index they are trying to track. Trading fees and Bid/Ask spreads increase the cost of this re-balancing. Remember that even a -1x inverse fund is leveraged, and a -2x inverse fund needs as much leverage as a 3x long fund.

A simple example will help explain this. Consider a 2x leveraged fund trying to track an index with a value of 100. It will start with a NAV (net asset value) of 100 and a leveraged position with a nominal value of 200. If the index falls by 10% to 90, then the fund’s position will fall by 10% to 180 for a nominal loss of 20 – exactly twice the loss of the fund, so far so good. Now, the fund has a NAV of 80 (because it lost 20, twice the daily fall) and an opened position of 180, but the fund’s NAV is 80, so a position of twice its NAV is only 160, not 180. The fund thus needs to reduce its hedge by 20 to get its exposure to the index back to twice its NAV. The opposite also applies – if the index goes up, then the fund’s NAV will increase at twice the speed of the index and the fund will need to increase its hedge.

The result is that the larger the move of the underlying index in either direction, the more rebalancing the fund needs to do at the end of the day.

3. Compounding – the effect of each day’s changes applied to the previous day’s changes gives a different result than expected. This is best explained by an example. Consider an index valued at 100 which falls by 10% the first day and then rises by 10% the next day. We have 2x, 3x, -1x and -2x funds on this index. The table below shows the value of the index and of each fund after each of the two days:

	Index	2x fund	3x fund	-1x fund	-2x fund
Starting NAV	100	100	100	100	100
Day 1 change	-10	-20	-30	+10	+20
NAV after 1 day	90	80	70	110	120
Day 2 change	+9	+ 16	+21	-11	-24
NAV after 2 days	99	96	91	99	96
Expected NAV	99	98	97	101	102
Tracking error	0	-2	-6	-2	-6

The result after two days of trading is that the all the funds have lost money because of compounding. The index is down by 1%, but the 2x fund instead of being down by the expected 2% is actually down by 4% and the 3x fund is down by 9%!

The inverse funds also didn’t do what we would expect, i.e. rising by 1% and 2%, but fell by 1% and 4% respectively. Notice that compounding costs the 2x and -1x funds 2 points, but costs the 3x and -2x funds 6 points.

The larger the daily moves the greater the compounding effect.

The examples above give a simple explanation of some very subtle and complex factors, but the conclusion is clear: the greater the daily moves in the underlying index, the larger the potential tracking errors. This means that for indexes with high daily volatility, these funds will not track the selected multiple of the index over time.

That’s fine in theory, but how have these funds actually performed if an investor would have held them for a few months?

First I will compare the YTD returns of TLT and TBT (an index fund of 20+ year treasuries and its double inverse fund). From Dec 31, 2008 until today, TLT fell from $119.35 to $101.60, a loss of 14.9%. In the same period TBT rose from $37.72 to $46.27, a gain of 22.7%.

If TBT was a perfect double inverse of TLT it would have gained 29.8% as opposed to 22.7%. So, while it didn’t work perfectly, it still gave gains of substantially more than the fall in the underlying index.

In the Barclays Global paper, the authors give two examples: IYF vs. SKF (the Dow financials and its double inverse) and DIG vs. DUG (a double long and double short of the Dow US Oil & Gas index). They make the comparison over a longer period of time (November 08 – March 09 for IYF vs. SKF and September 08 – March 09 for DIG vs. DUG). In both cases an investor would have lost money on both sides of the trade with SKF -25.10%, IYF -36.35%, DUG -17.39% and DIG -74.73%.

For comparison with my example, the YTD values are SKF -37.7%, IYF -17.1%, DUG -3.5% and DIG -21.6%. Not quite as bad, but still a loss either way. However, SKF closed on March 6^th 2009 at a price of $250.07 for an impressive YTD gain of +142.7%, which just goes to show that timing is everything in leveraged investments.

What can we learn from all of this? An investor seeking to hold these funds for anything more than a few days needs to be aware of the inevitable tracking errors. The more volatile the underlying index and the longer the fund is held, the greater the expected loss of value. Holding a leveraged fund on a volatile index over a period of a year is clearly a losing proposition. Of course attempting to get the same effect using call or put options would also be expensive.

On the other hand, if the investor is expecting a strong move in the index over the short to medium term, then these funds are a useful tool - maybe a bit blunt and imprecise, but still useful. The key is timing, when to get in and when to get out.

Disclosure: Long TBT, no position in the other securities mentioned.

Comments

[Think-About-It:]

Thanks Josh!! Very well explained.

Apr 23 01:11 PM

[donzelion:]

SKF strikes me more as a sort of short-term insurance rather than an investment. Sort of like buying short-term insurance on a rental car - this is a vehicle that I intend to ride when the markets seem to be overheating. I buy and sell based on limit orders, and accept the fact that these are not long term commitments (then again, who would ever want to make a long-term commitment to the collapse of a sector of the economy?).

Apr 23 03:34 PM

[henarl:]

Great article. As the author points out,TBT is only good for the short term due to the daily revaluing. By buying TBT and selling a one or two month covered call with a srike at or near the purchase price and repeating the process if the call expires worthless, or if you're exercised out, one can use TBT as a profitable short term trading position while waiting for the bottom to fall out of treasuries.

Apr 24 10:30 AM

[Arkad:]

It really boils down to portfolio rebalancing strategies. The three basic formulaic portfolio rebalancing methods are:

Buy and hold,
Constant mix,
Constant proportion of portfolio insurance.

A buy and hold strategy would be where you take 50% margin at the outset and don't rebalance through to the end of your investment horizon. As your risky asset falls in value your daily margin percentage will increase and as your risky asset rises in value your margin percentage will decrease. Except for interest costs would obtain double your returns over your investment horizon.

In the case of leveraged ETF's they promise (and deliver) daily leverage of 2x. This is the constant mix strategy. This means that when stocks fall (and leverage increases) they have to deleverage, when stocks rise (and leverage decreases) they have to releverage. In other words: buy high, sell low. The constant mix strategy, although it smacks of illogic because you are buying high and selling low can be preferable in trending market (where the market is always moving the same direction between portfolio rebalancing) and can be implemented profitably if you rebalance your portfolio every 6 months or every year (because you are buying more of the security that will continue to rise or selling more of the security that will continue to fall). However, in the case of leverage ETFs where the portfolio is rebalanced on a daily basis you will only find brief periods where constant mix is preferable to buy and hold. Over the long-term you will get less with constant mix than you will with a buy and hold position.

Apr 24 11:19 AM