The MLP Sector Fund Review

by: Arbitrage Trader


What drives this sector.

A detailed look at closed end funds.

How to hedge and boost performance.

A master limited partnership (MLP) is something long term investors buy as an income vehicle, because of its regular cash distributions earned in the oil and gas sector. Last year proved once again that there is no free lunch. The attractive yields MLPs provide are far from risk free.

In this article we will review the MLP sector as it is now and take a close look at open and closed end funds. The energy sector has recovered, but does that necessarily mean MLPs are a buy? This surely was the most interesting sector since oil prices started collapsing.

There are several interesting topics that will be discussed in the article:

  • Does the MLP sector move with oil?
  • What happens to premium/discount of closed end funds in relation to NAV direction?
  • Do MLP open end funds differ from each other?
  • How to use closed end funds deviations and Z-score.

Open end funds

The benchmark for the MLP sector is the ALPS Alerian MLP ETF (NYSEARCA:AMLP). This is easy to trade because of its 1 cent spread through out the trading hours. It is always available to short and has options on it. Some basic info for the fund can be seen here:

The trend in the MLP funds is to create portfolios that are concentrating more than 60% of their assets in the top 10 holdings. This is the case with any MLP closed end fund as well. There is almost no diversification effect because the holdings are highly correlated to each other.

Although we will use AMLP as the benchmark in the MLP sector, statistically speaking it makes almost no difference if you have AMLP or the JPMorgan Alerian MLP Index ETN (NYSEARCA:AMJ):

Source: author's software.

The simple regression model is good enough prove that AMLP and AMJ are the same thing with a different name.

On the bottom left chart you can see the total profit/loss for 100 shares long in AMLP/AMJ. At some point the loss was more than 40%. I am sure there is not a single reader who doubts the price of oil caused the large losses in the MLP sector, but let's look further.

AMLP vs the The United States Oil ETF (NYSEARCA:USO)

Why would the price of an MLP company be related to the price of oil? Most of the MLP companies are not even drillers. In normal times there should not be any real correlation between the company's profitability and the price of oil.

In times of panic the situation is different. People sell because oil is dropping. No one searches for logic when he is losing money.

On the following chart we will compare a portfolio of AMLP vs USO:

For the last 150 days we have a correlation coefficient of 0.83. This was even higher 3 months ago. I remember seeing the 150 days correlation as high as 0.95.

And then suddenly the correlation is gone for the last 80 days. USO starts dropping and AMLP 'does not care'. The regression model is broken. The simple question here is where is the mistake? Does AMLP have logic to follow USO or was it just a moment of panic that made these 2 assets move together?

Let's include some sentiment in the party with the SPDR S&P 500 Trust ETF (NYSEARCA:SPY), and some gas with the United States Natural Gas ETF (NYSEARCA:UNG):

The model is stronger now and the correlation coefficient is relatively stable. My personal (not statistical) conclusion is that currently sentiment is so bullish that people just forget to look at USO. It seems people do not remember the panic in the MLP sector.

On the other hand the whole MLP panic might have been without any real logic behind it and is rightly ignored in today's market. MLPs may have dropped with oil, then recovered with oil, but that doesn't mean the correlation was justified. The recent disconnect is interesting, but it may just be that normal trading patterns have resumed.

Deviations from averages should really be your main concern if you are invested in MLPs. This is where the real opportunity for profit lies. The best place to search for such deviations is....

MLP closed end funds (CEFs)

If you are invested in MLP companies and you don't use closed end funds as a hedging vehicle, you are either an amateur or such a big fish that liquidity in CEFs is not enough for you. I understand that this might sound arrogant, but this is not an opinion or insult, this is just a fact. The next picture illustrates this statement in the best possible way:

There were almost 3 months of panic. In these 3 months the Cohen & Steers MLP Income and Energy Opportunity Fund (NYSE:MIE) was either at extreme discount or at extreme premium. The fund was either not responding to its falling NAV, thus widening the premium, or was falling too much, widening the discount.

A long term holder of MLPs could short MIE on every extreme move of its premium and save a big part of its portfolio value by the profits generated as it returned to its mean.

Don't be fooled by the Z-score. Z-score is a standardized way to look for extreme valuations. There is one big disadvantage; there is no 'standard deviation of the standard deviation' in the formula. A closed end fund might have had a very stable premium/discount that has never deviated. Let's say it is 0.5%. Then it narrows its discount from (-)15% (average) to (-)13%. This gives a Z-score = (-13+15) / 0.5 = 4. This will be probably the most overvalued closed end fund by Z-score, but obviously it is way undervalued.

The other situation is when the standard deviation is very high; if you are only watching Z-scores, you may easily miss an overvaluation. This is currently the case with many MLP closed end funds.

When do premiums widen?

Ask anyone and he will tell you that a bullish sentiment and a raising sector are the number one reason for a fund to narrow its discount or to widen its premium.

In fact this is not the case with MLP CEFs. You may have noticed the average discount for MLP closed end funds has actually widened since March. This is the biggest rally MLP sector has had for a long time and the bullish sentiment has not caused any overvaluations. To make it even more interesting we will take a look at some closed end funds that are trading close to their 'normal' lowest discount levels.

Funds to buy based on a regression model

I constructed a portfolio of 7 closed end funds I currently hold vs AMLP. The position was taken at 2 standard deviations and is currently at a profit of almost 1 standard deviation:

This is a simple example of how to use the deviations for switching between AMLP and a portfolio of CEFs.

I have concluded that z-scores change randomly for closed end funds and they just need some time to go from overvalued to undervalued and then to overvalued again. I am betting that they will soon be overvalued again, but this is a personal trading bet that in no way advice to buy.

What next in the MLP sector?

I am in no way an expert in big market trends or in understanding businesses that I do not manage or even work in.

What I know is that human behavior is naive. The MLP sector was cut in pieces last time oil made new lows. It is just a matter of time for people to panic again.

Currently no long term investor is feeling any pain. The world is such a nice place to live in with cheap gasoline, and equities at all time highs. I do not believe the double digit current yields to be true and stay away from such temptations.

My current position is a hedged one, as always, but I am a little more short in AMLP than I am long in MLP CEFs.


Oil prices and MLPs were highly correlated through most of 2016, but there has been a significant disconnect of late. This may be a concern for investors in the sector, but at least there are opportunities to buy CEFs at discounts to average. This provides a buffer in any downturn, and can boost yield on a rally.

Click 'follow' by my name at the start of the article to receive more updates like this.

Disclosure: I am/we are long CEM, NML, KYE, MIE, KED, CTR, KMF.

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.

Additional disclosure: I am short AMLP