Is MORL And Other High Yield 2X Leveraged ETNs Impacted By Re-Balancing Decay?

|
About: UBS ETRACS Monthly Pay 2x Leveraged Mortgage REIT ETN (MORL), Includes: BDCL, CEFL, REML
by: Lance Brofman
Summary

Leveraged ETNs effectively must buy additional shares of the underlying components when the market price of the components rises and sell when the market falls.

Some investors are concerned that selling at the bottom and buying at the top which could impart a downward bias call decay, to the returns of leveraged ETNs.

The existence of such a downward bias or decay would be a violation of the efficient market hypothesis.

There is now a long enough price history for MORL and other High Yield 2X Leveraged ETNs to use market back-testing to see if decay is a cause for concern.

What Is Decay and Should Leveraged ETN Investors be Concerned With It?

I have been following high yielding Leveraged ETNs such as: the UBS ETRACS Monthly Pay 2X Leveraged Mortgage REIT ETN (NYSEARCA: MORL), UBS ETRACS 2X Leveraged Long Wells Fargo Business Development Company ETN (NYSEARCA: BDCL) and UBS ETRACS Monthly Pay 2X Leveraged Closed-End Fund ETN (NYSEARCA: CEFL) for an number of years. These ETNs emulate a portfolio based on a specific index that is 2X leveraged. In order to maintain the 2X leverage, the size of the implicit basket of securities in the portfolio must be periodically rebalanced. This entails increasing the size of the assets that one share of the leveraged ETN represents if the market value of the index components increase and reducing the size of the assets that one share of the leveraged ETN represents if the market value of the index components decrease.

In my November 3, 2014 article: Do Leveraged ETNs Buy High And Sell Low? And Does It Matter? I addressed the issue of possible decay involving these securities. Decay in this context refers to the concept that adjusting the size of a portfolio to maintain a specified degree of leverage could impart a downward bias to the returns on a leveraged ETN. Since MORL is my largest holding, this concerned me.

Most of the alarmist concern over decay involving periodically rebalanced leveraged ETNs begins with the simple observation that if one buys a security that then declines by 50%, in order to get back your original investment it must increase by 100%. This simple math can be applied to projected results of a hypothetical 2X leveraged ETN. Assume a 2X leveraged ETN starts period 1 with 100 shares outstanding and a net asset value of $10. Then the number of shares outstanding in the ETN remains constant over two periods. Thus, it begins with $1000 in equity. In order to be 2X leveraged it must implicitly borrow $1000 and use the proceeds to buy a total of $2000 in the underlying security or basket of securities.

For simplicity sake assume that the underlying portfolio basket consists of 200 shares of a security initially valued at $10. For the purposes of this calculation it does not matter whether the ETN holds 200 shares of a $10 security or a diverse portfolio whose weighted average net asset value is $10. Likewise, this calculation at this point is exactly the same as a margin account where you start out with 200 shares of a $10 stock with a margin loan of $1,000 and thus have an initial equity of $1,000.

Now assume that the underlying shares or portfolio declines by 20%. Thus, the price of the underlying stock goes from $10 at the beginning of the period, to $8 at the end of period 1. The margin account or 2X leveraged ETN would the suffer a 40% decline in net asset value. At the end of period 1 the number of underlying shares would be reduced to bring the degree of leverage by to 2X in the margin account. That would mean selling enough shares to reduce the loan size to $600 from the initial $1,000. Thus, 50 shares would be sold at the $8 price. As is shown below, period 2 would then start with 150 shares of underlying securities and a total net asset value of $600 for the ETN which still has 100 shares outstanding of each with a net asset value of $6.

If the underlying security increases by 20% by the end of period 2 it would bring it to $9.60. This would bring the equity to $840 and the net asset value of the ETN to $8.4. At the beginning of period 3 the loan would be increased with more shares purchased at $9.60. That would not change the amount of equity or the net asset value of the ETN.

This means that the 2X leveraged ETN or margin account underperformed an unleveraged investment when there is a 20% decline then followed by a 20% rebound since an unleveraged investment in the original $10 stock ended up being worth $9.60 after the 20% decline followed by a 20% increase. As compared to the $8.40 for the unleveraged investment in the same security over the same holding period.

Change

period

shares

price

value

loan

equity

NAV

start1

200

$10.00

2000

1000

1000

10

-20%

end1

200

$8.00

1600

1000

600

6

start2

150

$8.00

1200

600

600

6

20%

end2

150

$9.60

1440

600

840

8.4

As discussed earlier to get even after a 50% decline in an investment, it needs to increase by 100%. Thus, it is no surprise that a 20% decline in an investment needs a 25% increase to get even. In all of these examples expenses are not considered. However, we do know that all borrowing will involve some interest expense, which will reduce the return on the leveraged investment relative to one not financed with borrowing.

When we consider a 20% decline followed by a 25% rebound, we still see the 2X leveraged ETN or margin account underperforming an unleveraged investment. As shown below the underlying security starts at $10 and falls to $8 at the end of period 1 (also the start of period 2) as in the first case. However, now in period 2 the underlying security increases 25% back to $10.

Change

period

shares

price

value

loan

equity

NAV

start1

200

$10.00

2000

1000

1000

10

-20%

end1

200

$8.00

1600

1000

600

6

start2

150

$8.00

1200

600

600

6

25%

end2

150

$10.00

1500

600

900

9

Here, even though the underlying security recovers to its original price, the equity of the 2X leveraged ETN or margin account still declines by 10%. Thus, it might be inferred that in this example the 2X leveraged ETN or margin account experienced decay of 10% relative to an unleveraged investment in the underlying security over the same period.

If the order of the decline and increase is reversed, the underperformance of the 2X leveraged ETN or margin account in terms of total return still exists. As can be seen below, an increase of 20% followed by a decrease of 20% has the same ending equity values as the 20% decline followed by a 20% increase. At the start of period 2 the loan is increased by $400 and 33.33 additional shares are added to the portfolio. The 20% decline in period 2 brings the underlying share price to $9.6 while the net asset value of the 2X leveraged ETN falls to the same $8.40 as in the example with a20% decline followed by a 20% increase.

Change

period

shares

price

value

loan

equity

NAV

start1

200

$10.00

2000

1000

1000

10

20%

end1

200

$12.00

2400

1000

1400

14

start2

233.33

$12.00

2800

1400

1400

14

-20%

end2

233.33

$9.60

2240

1400

840

8.40

If we use an example of an increase of 20% followed by a 16.67% decrease, the amount required to bring the underlying security back to the original $10, the 2X leveraged ETN or margin account still shows a negative total return over the entire period, in contrast to the 0% return on the unleveraged portfolio.

Change

period

shares

price

value

loan

equity

NAV

start1

200

$10.00

2000

1000

1000

10

20%

end1

200

$12.00

2400

1000

1400

14

start2

233.33

$12.00

2800

1400

1400

14

-16.67%

end2

233.33

$10.00

2333.33

1400

933

9.33

This is starting to look scary. One might conclude that decay is inevitable for 2X leveraged ETNs. However, this would be what economists refer to as a fallacy of composition. The fallacy is in the assumption that something is true in all cases because it is true in one or some cases.

In my original article mentioned above, I said:

"...The fact that leveraged ETNs effectively have to buy additional shares of the underlying components when the market price of the components rises and sell when the market price of the components falls, concerns some investors. This looks like selling at the bottom and buying at the top. However, buying when prices rise is not the same as buying at the top. That is because no one really knows when a top occurs.

Some investors fear consistently buying when prices rise and selling when they fall could bias downward the total return from leveraged ETNs. If this was the case, it would be a violation of the weak-form of the efficient market hypothesis. The term "weak-form" distinguishes this form of the efficient market hypothesis from the stronger forms. Many more in the field of finance adhere to the weak-form of the efficient market hypothesis which states that market price information is fully reflected in the market, than adhere to the stronger forms, which state that all information is fully reflected in the market.

Another way of stating the weak-form of the efficient market hypothesis is: that technical analysis does not work. This means that you cannot use any system or analysis of the price history of any security to achieve an abnormal return. The stronger forms of the efficient market hypothesis state that neither technical nor fundamental analysis is useful in terms of achieving abnormal returns. There is no way to absolutely prove either form of the efficient market hypothesis. The major evidence supporting the weak-form of the efficient market hypothesis is that no one has ever demonstrated any statistically significant system, which makes use of market price information that does achieve abnormal returns. Critics of the weak-form of the efficient market hypothesis point out that if someone did find a technical system that produced abnormal returns they would likely keep that information to themselves rather than publishing in an academic journal proof that the weak-form of the efficient market hypothesis is wrong.

If buying after the market went up by some amount or made new highs and selling after the market went down by some amount or made new lows did cause a downward bias in the returns on leveraged ETNs or any other security that would be a direct violation of the weak-form of the efficient market hypothesis. This is because doing the exact opposite: buying after the market went down by some amount and selling after the market went up by some amount would then have to produce an upward bias, which is referred to as an abnormal return, which violates the weak-form of the efficient market hypothesis..."

There are no laws that insure that actual securities conform to the efficient market hypothesis. Thus, my explanation that decay resulting from 2X leveraged ETNs should not be a concern because it would be a violation of the weak-form of the efficient market hypothesis, did not assure all readers. Some preferred to see market back test data. At that time (2014) the UBS ETRACS Monthly Pay 2X Leveraged ETNs did not have much price history. However, there is enough historical price data now.

In order to investigate whether decay is inevitable for periodically rebalanced 2X leveraged ETNs, separate from the efficient market hypothesis, we should reexamine the fallacy of composition concept. The first clue that the four examples of "round trip" price movement involving a security increasing then decreasing or vice versa do not prove decay is inevitable, is that in all cases where the underlying security constantly increases the leveraged ETN or portfolio outperforms the unleveraged on (ignoring interest expenses).

What may not be obvious, is that comparing a periodically rebalanced 2X leveraged ETN to a 2X leveraged ETN that is not periodically rebalanced, in all cases where the underlying security either constantly increases or decreases, the periodically rebalanced leveraged ETN or portfolio always outperforms the 2X leveraged ETN that is not periodically rebalanced. Periodically increasing leverage as the market moves higher, means increasing exposure to a rising market. Periodically decreasing leverage as the market moves lower, means decreasing exposure to a falling market. Remember, it is the periodically rebalancing to maintain 2X leverage, by adding to the portfolio after a rise in the price of the underlying security and selling after a decline, which is the alleged cause of decay for 2X leveraged ETNs.

The existence of examples, such as one direction price movement of the underlying security in either direction where periodically rebalancing to maintain leverage always outperforms non-rebalanced portfolios, indicates that just because some paths or set of paths show periodically rebalanced 2X leveraged ETNs underperforming the underlying securities, it is not necessarily true that a bias exists that causes rebalanced 2X leveraged ETNs to underperform on balance.

Before examining the results of market back tests of periodically rebalanced 2X leveraged ETNs as compared to the underlying securities, a few facts should be discussed. In general, when the underlying securities are rising in price, use of leverage increases the total return on any investment. Likewise, leverage amplifies losses when are securities are falling in price. This means that the use of actually observed market data for holding periods when either the underlying securities rose or fell on balance over the entire holding period is not particularly useful.

When we can identify relatively long periods where the starting purchase price and the ending price are the same, we can attempt to isolate the impact of any decay that might have occurred as the periodically rebalanced 2X leveraged ETN adjusted their portfolio size and thus their exposure to market movements was changed many times in both directions before ending up where they started.

For MORL, CEFL and BDCL there are exchange-traded securities based on the same underlying portfolios that do not employ any leverage. For MORL there is Market Vectors Mortgage REIT ETF (NYSEARCA: MORT). For CEFL there is YieldShares High Income ETF (NYSEARCA: YYY). and UBS ETRACS Wells Fargo Business Development Company ETN (NYSEARCA: BDCS), the unleveraged ETN companion of BDCL. There is also another high yield 2X Leveraged Monthly pay ETN based on an index of mREITs similar to those in the portfolios of MORL and MORT. The X-Links™ Monthly Pay 2xLeveraged Mortgage REIT ETN (NYSEARCA: REML) is an exchanged traded note that is based on the FTSE NAREIT All Mortgage Capped Index of mREITs. That is the same index used by the unleveraged iShares Mortgage Real Estate Capped ETF (NYSEARCA: REM). Thus, REML is a another way to get high yields from a leveraged ETN. As I indicated in REML: Another Way To Get 20% Yield, Or A Warning? REML is still very thinly traded, which can be both an advantage and disadvantage at times.

Market Back Testing

The procedure in the empirical analysis of financial market data when trying to gain useful investment insight, is to use as many observations as feasible while also attempting use the most current data. More observations increase statistical significance. The more current the data, the less likelihood that the relationship or phenomenon you are trying to quantify has changed over time. Anyone who has ever tried to estimate the beta of a security is aware of this. MORL and CEFL adjust monthly to maintain 2X leverage. BDCL only adjusts quarterly so it does not provide as much significant information.

When utilizing actual data involving real securities, expenses and fees cannot be ignored. This is especially true when comparing an ETN or ETF that uses leverage, and thus incurs interest expenses, to one that does not. If you are not as enamored with applications of theoretical finance as I am, you might be only, or at least more, interested in whether MORL, CEFL and BDCL did or did not exhibit decay as a result of periodically adjusting to maintain 2X leverage, as opposed to the implications for the efficient market hypothesis.

If decay was significant, the risk-adjusted total return on a periodically rebalanced 2X leveraged ETN should be less than that of a fund or ETN that had an identical portfolio of underlying securities and did not employ leverage. Of course, during a period where the underlying securities ended up at the same market price as they started at, a 2X leveraged ETN or margin account will have a higher total absolute return than the same basket of securities not using leverage. This is especially true in periods of low interest rates and high positive carry from the borrowing. The question is whether on a risk-adjusted basis, the 2X leveraged ETN or margin account outperformed the unleveraged version.

Volatility as measured by standard deviation can be used as a measure of risk. The volatility of 2X leveraged ETNs is twice that of their unleveraged companion versions. Thus, for decay to be observed, the total return over a "round trip" holding period for 2X leveraged ETNs should be significantly less than twice that of their unleveraged companion versions. Differences in fees and expenses or the timing of ex-dates could account for some deviation from what might be expected otherwise. One would expect that interest expense incurred implicitly with a 2X leveraged ETN or explicitly with a leveraged margin account, should cause total returns over a "round trip" holding period for 2X leveraged ETNs to be slightly less than twice that of their unleveraged companion versions.

It should be noted that dividends are a significant factor in the total return over a "round trip" holding period for any basket of high yielding securities. Thus, if decay exists in violation of the efficient market hypothesis, its' effect could be offset or even overwhelmed by another possible violation of the efficient market hypothesis relating to what could be termed a "dollar cost averaging" impact from reinvesting the significant monthly dividends from high yield 2X leveraged ETNs. It also should noted, that in a round-trip holding period when one of the securities ends the period at the same price it started at, the end price of a UBS ETRACS High Yield Monthly Pay 2X Leveraged ETN will be lower than the start relative to unleveraged ETF with the same underlying portfolio. This is because the interest and other expenses are deducted from the net asset or indicative value of the UBS ETRACS High Yield Monthly Pay 2X Leveraged ETN. Expenses for the unleveraged ETFs come out of the dividends. This does not bias the total return comparisons on way or the other.

In order to do a "round trip" holding period total return comparison, there must be two dates, hopefully as far apart as possible, where one of the securities being compared began and ended the period at the same price. For MORL and MORT we can use the period starting on 8/15/2013 and ending on 6/13/2017. The closing price of MORL was $18.44 on both dates. The total return on MORL for that period, assuming reinvestment of all dividends was 124.70%, this is significantly more than twice the 52.65% total return on MORT for that period using the same reinvestment of all dividends basis. There was plenty of volatility in the price of MORL over that period that would have been sufficient to result in decay if ant was present. The range for MORL over that period was a low price of $9.25 and a high of $24.59.

Picking a period where the price of MORL was identical at the start and end, could bias the comparison, since the different way that expenses are treated by MORL could mean that a period where the price of MORL was unchanged could have been a rising market on balance. To check about this, I found a period where MORT was unchanged. MORT closed at $24.72 on both 7/10/2014 and 6/13/2017. Even though MORL declined from $22.00 at the start of that period to $18.44 at the end, the total return on MORL for the period was 60.54% again more than twice than the 25.25% of MORT during the same period. Again using the same reinvestment of all dividends basis.

CEFL closed at $18.15 on both 8/19/2015 and 6/9/2017. The total return on CEFL for that period assuming reinvestment of all dividends was 39.80% slightly less than twice the 21.44% total return on YYY for that period using the same reinvestment of all dividends basis. That is about what would be expected given the interest expense incurred by CEFL and not YYY. Over that period the price of YYY only increase from $19.40 to $19.61. Thus, it was a period over which the underlying closed-end funds that comprise the portfolios of both CEFL and YYY did not significantly rise in value.

I was not able to find any holding period where either BDCL or BDCS started or ended with the exact same price. The closest I was able to find was the period from 7/20/15 to 6/12/2017 where BDCL started at $19.33 and ended at $19.34. The total return on BDCL for that period was 35.42% slightly less than twice the 18.28% total return on BDCS for that period using the same reinvestment of all dividends basis. That is about what would be expected given that interest expense is incurred by BDCL and not BDCS. As was mentioned earlier, BDCL only rebalances leverage quarterly as opposed to the monthly, so data from the comparison of returns from BDCL and BDCS is not that significant. There was not enough price history for REML to gather any meaningful round trip comparisons with the unleveraged version REM.

Summary and Conclusions

The results and analysis suggest that either there is no significant decay associated high yield periodically rebalanced 2X leveraged ETNs, or that whatever decay does occur is offset or overwhelmed by the dollar cost averaging inherent in the reinvestment of the dividends, or some other factor. For most investors it should not matter why decay does not appear to be present for these high yield periodically rebalanced 2X leveraged ETNs, only that it is probably not a factor that should concern them.

It also should be noted that even if you are not convinced by the weak-form of the efficient market hypothesis or the empirical results which show no statistically significant downward bias due to periodic rebalancing to maintain the leverage at 2X, you could also offset any possibility of such downward bias yourself. If you reduce the size of your position in a leveraged ETN anytime the size of the assets per share are increased due to a rise in the market value of the components, and increase your position, also proportionately to the reduction in a 2X leveraged ETN anytime the size of the assets per share is decreased due to a decline in the market value of the components, you will precisely offset the effect of the periodic rebalancing to maintain the leverage at 2X. However, I would suggest it is a waste of time and commissions since the analysis indicate that there is no downward bias because of the periodic rebalancing to maintain the leverage at 2X for these 2X leveraged ETNs.

It also might be noted that MORL is based on an index of mREITs, which are leveraged themselves and are managed by humans who are prone to overreact to declines in the market and generally must reduce the size of their leveraged portfolios of mortgage securities in response to market declines for those mortgage securities. This effect of buying when the market improves and selling when it declines would impact unleveraged MORT as well and as such it would not appear either way in the round trip results. For CEFL and BDCL there is generally not much selling or buying is response to market movements by the managements of the individual component closed-end funds or business development companies that comprise the indices upon which CEFL and BDCL respectively are based.

Disclosure: I am/we are long MORL, CEFL, BDCL, REM, REML. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.