Leveraged ETFs - Why Do We Have Them?

Via Peter Tchir of TF Market Advisors [4],

According to Barron's [5] as much as 91% of “triple” ETF’s might be owned by individual investors. That figure seems shocking, and as the article admits, could be wrong, but it is scary. The activity in TVIX the past few weeks does indicate a strong retail presence – I would like to think professionals didn’t bid something up to an 80% premium to NAV, knowing that the share creation process could be re-instated, virtually assuring that the premium would collapse to 0.

TVIX had its own special issues, in that it was tied to futures contracts and had to deal with futures “rolls”, but all leveraged ETF’s have some problems. Ignoring the costs, which are higher than other investments (leverage has a cost as does the more frequent rebalancing), the biggest problem is that the returns are “path dependent”. If investors hold the leveraged ETF for extended periods (and it seems that this is more common than I previously thought) their returns may not be what they expect. The total return of a leveraged ETF may differ significantly from what a typical investor might expect based on the movement of the underlying asset.

Questions about why an investor is sophisticated enough to make a leveraged bet, doesn’t have a futures account, or at least a margin account where they can do this on their own are for another time. This will just take a quick look at how leveraged ETF’s perform over time relative to an index, and that the results may be surprising. We look at moves in the index, the 2x ETF, the -2x ETF, and what happens if you invested equally in the double long and double short.

No Change in Index with Minimal volatility

For a simple base case, the underlying index barely moves. It has minimal volatility and ends up unchanged. Both the leveraged ETF’s experience losses in this case, small ones, but still both have losses.

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	-2x	Avg Dbl
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	101.0	102.0	98.0	100.0	1	99.0	98.0	102.0	100.0
2	100.0	100.0	99.9	100.0	2	100.0	100.0	99.9	100.0
3	101.0	102.0	97.9	100.0	3	99.0	98.0	101.9	100.0
4	100.0	100.0	99.9	99.9	4	100.0	100.0	99.9	99.9
5	101.0	102.0	97.9	99.9	5	99.0	98.0	101.9	99.9
6	100.0	99.9	99.8	99.9	6	100.0	99.9	99.8	99.9
7	101.0	101.9	97.8	99.9	7	99.0	97.9	101.8	99.9
8	100.0	99.9	99.8	99.8	8	100.0	99.9	99.8	99.8
9	101.0	101.9	97.8	99.8	9	99.0	97.9	101.8	99.8
10	100.0	99.9	99.7	99.8	10	100.0	99.9	99.7	99.8
Return	0.0%	-0.1%	-0.3%	-0.2%	Return	0.0%	-0.1%	-0.3%	-0.2%

So the drag on the leveraged ETF’s is minimal, but this is just the theoretical return in a “frictionless” environment. The reality is that there is a cost to the leverage and some costs as the leveraged ETF’s need to rebalance each and every day to prepare for the next days “double the daily return” trading. I expect the underperformance would be worse if that was accounted for.

No Change in Index with Extreme Volatility

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	-2x	Avg Dbl
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	103.0	106.0	94.0	100.0	1	99.0	98.0	102.0	100.0
2	100.0	99.8	99.5	99.7	2	106.0	111.9	87.6	99.7
3	97.0	93.8	105.4	99.6	3	103.0	105.5	92.5	99.0
4	92.0	84.2	116.3	100.2	4	100.0	99.4	97.9	98.7
5	97.0	93.3	103.7	98.5	5	97.0	93.4	103.8	98.6
6	100.0	99.1	97.3	98.2	6	92.0	83.8	114.5	99.1
7	103.0	105.0	91.4	98.2	7	97.0	92.9	102.1	97.5
8	106.0	111.1	86.1	98.6	8	100.0	98.6	95.7	97.2
9	99.0	96.5	97.5	97.0	9	103.0	104.6	90.0	97.3
10	100.0	98.4	95.5	97.0	10	100.0	98.5	95.2	96.9
Return	0.0%	-1.6%	-4.5%	-3.0%	Return	0.0%	-1.5%	-4.8%	-3.1%

Suddenly the returns are very scary. Yes, the time series example is extremely volatile, but it illustrates the example that both the doubles lose money under scenarios with lots of intra-period volatility that ends up going nowhere. The leveraged short ETF does worse, because as the index decreases in value, each 1 point of move is a greater %, so that greater % return is applied to a higher price on the short ETF. That means the reversal is more punishing than for the leveraged long. When the index is going up in price, each point move has less of a % impact. Strange, but yes, that is the problem with these leveraged ETF’s. They are not buy and hold assets, as they create losses in circumstances where the “index” didn’t move.

Stock Index Grinds Relentlessly

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	-2x	Avg Dbl
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	101.0	102.0	98.0	100.0	1	99.0	98.0	102.0	100.0
2	102.0	104.0	96.1	100.0	2	98.0	96.0	104.1	100.0
3	103.0	106.1	94.2	100.1	3	97.0	94.1	106.2	100.1
4	104.0	108.1	92.3	100.2	4	96.0	92.1	108.4	100.2
5	105.0	110.2	90.6	100.4	5	95.0	90.2	110.6	100.4
6	106.0	112.3	88.8	100.6	6	94.0	88.3	113.0	100.6
7	107.0	114.4	87.2	100.8	7	93.0	86.4	115.4	100.9
8	108.0	116.6	85.5	101.0	8	92.0	84.6	117.8	101.2
9	109.0	118.7	84.0	101.3	9	91.0	82.7	120.4	101.6
10	110.0	120.9	82.4	101.7	10	90.0	80.9	123.1	102.0
Return	10.0%	20.9%	-17.6%	1.7%	Return	-10.0%	-19.1%	23.1%	2.0%

So in this example, the double index earns more than 20% even though the index return is only 10% over the period. Investors can’t complain about that. Somewhat more difficult to believe is that an investment
split between the double long and the double short makes money on a move like this. More on the downside scenario than the upside scenario – because each 1 point index move is more as a % on the downside. This scenario highlights how well the leveraged ETF’s perform when there is minimal volatility.

A Gap Move where next period continues the Move

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	-2x	Avg Dbl
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	100.0	100.0	100.0	100.0	1	100.0	100.0	100.0	100.0
2	100.0	100.0	100.0	100.0	2	100.0	100.0	100.0	100.0
3	100.0	100.0	100.0	100.0	3	100.0	100.0	100.0	100.0
4	100.0	100.0	100.0	100.0	4	100.0	100.0	100.0	100.0
5	100.0	100.0	100.0	100.0	5	100.0	100.0	100.0	100.0
6	100.0	100.0	100.0	100.0	6	100.0	100.0	100.0	100.0
7	100.0	100.0	100.0	100.0	7	100.0	100.0	100.0	100.0
8	100.0	100.0	100.0	100.0	8	100.0	100.0	100.0	100.0
9	108.0	116.0	84.0	100.0	9	92.0	84.0	116.0	100.0
10	110.0	120.3	80.9	100.6	10	90.0	80.3	121.0	100.7
Return	10.0%	20.3%	-19.1%	0.6%	Return	-10.0%	-19.7%	21.0%	0.7%

So in this scenario we see a big gap in price right near the end of the period. The next period adds to the move. The double ETF’s outperform in that the long would be up 20.3% on a 10% move, and the short would be
up 21% on a -10% move. The second day’s move is what drives the divergence. Again, an investment equally weighted in both the doubles would have a positive return.

A Gap Move with Retracement

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	-2x	Avg Dbl
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	100.0	100.0	100.0	100.0	1	100.0	100.0	100.0	100.0
2	100.0	100.0	100.0	100.0	2	100.0	100.0	100.0	100.0
3	100.0	100.0	100.0	100.0	3	100.0	100.0	100.0	100.0
4	100.0	100.0	100.0	100.0	4	100.0	100.0	100.0	100.0
5	100.0	100.0	100.0	100.0	5	100.0	100.0	100.0	100.0
6	100.0	100.0	100.0	100.0	6	100.0	100.0	100.0	100.0
7	100.0	100.0	100.0	100.0	7	100.0	100.0	100.0	100.0
8	100.0	100.0	100.0	100.0	8	100.0	100.0	100.0	100.0
9	112.0	124.0	76.0	100.0	9	88.0	76.0	124.0	100.0
10	110.0	119.6	78.7	99.1	10	90.0	79.5	118.4	98.9
Return	10.0%	19.6%	-21.3%	-0.9%	Return	-10.0%	-20.5%	18.4%	-1.1%

So in this case, the move on the second last period is very big, and then we get a small retracement. Now both the leveraged ETF’s underperform the “expectation” of a 20% return. Unlike the prior example, where an investment in both doubles produced a positive return, here they produce a loss. They do not perform well when there are reversals.

Volatility around a Grind

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	-2x	Avg Dbl
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	103.0	106.0	94.0	100.0	1	97.0	94.0	106.0	100.0
2	102.0	103.9	95.8	99.9	2	98.0	95.9	103.8	99.9
3	105.0	110.1	90.2	100.1	3	95.0	90.1	110.2	100.1
4	104.0	108.0	91.9	99.9	4	96.0	92.0	107.9	99.9
5	107.0	114.2	86.6	100.4	5	93.0	86.2	114.6	100.4
6	106.0	112.1	88.2	100.1	6	94.0	88.1	112.1	100.1
7	109.0	118.4	83.2	100.8	7	91.0	82.4	119.3	100.9
8	108.0	116.2	84.8	100.5	8	92.0	84.3	116.7	100.5
9	111.0	122.7	80.0	101.4	9	89.0	78.8	124.3	101.5
10	110.0	120.5	81.5	101.0	10	90.0	80.5	121.5	101.0
Return	10.0%	20.5%	-18.5%	1.0%	Return	-10.0%	-19.5%	21.5%	1.0%

We have introduced some minimal volatility to the “grind” scenario. The leveraged ETF’s outperform again, but not by as much as the straight grind scenario presented earlier.

High Volatility

Day	Index	2x	-2x	Avg Dbl	Day	Index	2x	Avg Dbl	50 in Each
0	100.0	100.0	100.0	100.0	0	100.0	100.0	100.0	100.0
1	102.0	104.0	96.0	100.0	1	98.0	96.0	104.0	100.0
2	100.0	99.9	99.8	99.8	2	100.0	99.9	99.8	99.8
3	97.0	93.9	105.8	99.8	3	103.0	105.9	93.8	99.8
4	94.0	88.1	112.3	100.2	4	106.0	112.1	88.3	100.2
5	100.0	99.4	98.0	98.7	5	100.0	99.4	98.3	98.8
6	104.0	107.3	90.1	98.7	6	96.0	91.4	106.2	98.8
7	102.0	103.2	93.6	98.4	7	98.0	95.3	101.7	98.5
8	102.0	103.2	93.6	98.4	8	98.0	95.3	101.7	98.5
9	106.0	111.3	86.2	98.8	9	94.0	87.5	110.1	98.8
10	110.0	119.7	79.7	99.7	10	90.0	80.0	119.4	99.7
Return	10.0%	19.7%	-20.3%	-0.3%	Return	-10.0%	-20.0%	19.4%	-0.3%

With increased volatility, neither of the leveraged ETF’s get to their “expected” 20% returns.

What does it all mean?

Professional investors know they have to “rebalance” every day. You do not have double the return over the period, you get double the daily return, which can be very different. But why does a professional investor need these? They have access to futures and to margin accounts, so there is no reason for these to exist just for professional investors. They must exist for retail investors, and I find it hard to believe that retail investors understand the range of potential returns on the leveraged ETF’s for any given holding period return of the underlying index.

I don’t see a need for these products except for small retail investors who can’t get leverage any other way, and I suspect they don’t understand how these things really work, as they are the most likely to buy and hold these things.