Midstream Distributions Are Not Safe: Don't Be Fooled By DCF

Summary

• Midstream Companies heavily promote Distributable Cash Flow (DCF) as a metric to judge distributions against.
• DCF vastly overstates a company's true cash generation ability.
• Don't expect high coverage ratios to protect distributions (or allow for companies to raise them).
• Article examines FCF and DCF coverage for large midstream names including ENB, EPD, KMI, ET, OKE, WMB, MPLX, MMP.

Midstream companies often tout their cash distributions as a key underpinning in their investment case. As proof that the yields won't be cut, and in fact may be raised, companies point to Distributable Cash Flow ("DCF") coverage. The figures below show DCF coverage (DCF / Distribution) ratios for the largest publicly traded midstream companies for the most recent few years.

Name	2016	2017	2018	2019
Enbridge Inc. (ENB)	1.92	1.53	1.65	1.55
Enterprise Products Partners LP (EPD)	1.22	1.24	1.58	1.63
Kinder Morgan Inc. (KMI)	4.03	4.00	2.68	2.21
Energy Transfer LP (ET)	0.98	1.17	2.38	1.96
ONEOK Inc. (OKE)	2.53	1.62	1.80	1.74
Williams Cos Inc. (WMB)	1.44	1.61	1.69	1.79
MPLX LP (MPLX)	1.23	1.28	1.49	1.51
Magellan Midstream Partners LP (MMP)	1.25	1.25	1.25	1.40

Click to enlarge

Source: Company press releases and filings. ET is pro forma values provided by ET for consolidated company.

Based on the data in this chart, things look pretty good for midstream companies. Even for the worst capitalized amongst this group (MMP in 2019), DCF could decline by 28% and MMP could still cover its dividend.

However there is a reason why strong DCF coverage doesn't result in dividend increases - because the companies can't afford it. DCF is a non-GAAP metric which for most midstream companies is computed as cash flow from operations (without changes in working capital) minus "maintenance capital expenditures". Some companies use the term "sustaining" instead of "maintenance".

Maintenance capex presents an issue - DCF sounds good, like many an adjusted metric promoted by public companies, but digging deeper it doesn't make much sense. The maintenance capex figure is supposed to be the amount needed to keep current assets operational, e.g. if the company stopped growing and just maintained what it had, maintenance capex is how much it would cost.

However, when you compare maintenance capex with depreciation and amortization (D&A) expenses, things start looking shady. D&A is an estimate guided by accounting rules on how much the value of an asset declines each year driven by wear and tear. The D&A expense takes into account the estimated life of the assets, including the estimated life of an oil and gas field that is attached to an asset (most relevant for G&P assets). In other words, D&A is not just some made up number - it is an estimate based on the actual prospects and lives for actual assets, for how much the assets value declines each year. While it won't be exact, maintenance capital expenditures should be pretty close to D&A.

However, maintenance capital for the midstream group is across the board far below D&A.

Maintenance Capex / D&A

Name	2016	2017	2018	2019
Enbridge Inc.	30%	40%	35%	32%
Enterprise Products Partners LP	16%	15%	18%	17%
Kinder Morgan Inc.	24%	26%	30%	27%
Energy Transfer LP	21%	19%	18%	21%
ONEOK Inc.	34%	36%	44%	41%
Williams Cos Inc.	28%	27%	31%	27%
MPLX LP	14%	15%	20%	21%
Magellan Midstream Partners LP	55%	43%	33%	40%

Click to enlarge

Source: Company filings, press releases.

The huge gap between maintenance capex and D&A suggests that the true amount required to keep the assets in a steady state is actually quite a bit higher. But why would a midstream company distort the actual costs of operating its business, you ask? Because allows management to tell a better story to investors and to pay out a higher distribution. A high distribution 1) keeps investors interested, and 2) makes the stocks very expensive for short sellers to target.

Looking at total capex vs. D&A, we get a more complete picture. I have included the same metric for Union Pacific (UNP), the largest US railroad. While UNP is of course not the same as Midstream, they are similar in that they are transporting goods over long distances, on fixed networks that require regular maintenance.

Total Capex / D&A

Name	2016	2017	2018	2019
Enbridge Inc.	229%	263%	210%	162%
Enterprise Products Partners LP	195%	191%	236%	239%
Kinder Morgan Inc.	146%	141%	128%	97%
Energy Transfer LP		331%	259%	189%
ONEOK Inc.	186%	140%	500%	808%
Williams Cos Inc.	116%	138%	189%	123%
MPLX LP	222%	207%	221%	230%
Magellan Midstream Partners LP	462%	330%	282%	480%
Union Pacific	172%	154%	157%	156%

Click to enlarge

The above suggests that some midstream companies are investing more heavily than others. However, the maintenance capex figures certainly understate the true maintenance needs of midstream companies.

Free Cash Flow ("FCF")

Free cash flow is a much better proxy for ability to pay dividends. FCF is how virtually all industries are measured by the market, and is widely used as a proxy for the economic profitability of a company. FCF, defined as cash flow from operations less total capex, is the amount of cash left over after a company has paid all of its expenses and capex - this cash is in theory available for stockholders. Management can allocate the FCF how it best sees fit - to M&A, to dividends, to share buybacks, to pay down debt, or just keep on its balance sheet for future opportunities.

What would these companies look like if we looked at free cash flow as a proxy for paying dividends?

Free Cash Flow / Dividends

Name	2016	2017	2018	2019
Enbridge Inc.	0.04	(0.44)	0.80	0.65
Enterprise Products Partners LP	0.31	0.42	0.50	0.48
Kinder Morgan Inc.	1.41	1.26	0.62	1.06
Energy Transfer LP	-	(0.61)	0.04	0.63
ONEOK Inc.	1.41	0.87	0.03	(1.29)
Williams Cos Inc.	1.67	0.43	0.02	0.86
MPLX LP	0.20	0.41	0.46	0.46
Magellan Midstream Partners LP	0.12	0.53	0.66	0.18

Click to enlarge

Source: company filings, presentations.

As seen above, the midstream companies largely have not covered their dividends with free cash flow. No wonder massive dividend raises aren't happening!

What about the Future?

With any stock investment, the future is more important than the past. So let's take a look at how this will change in the future. Midstream companies' guidance today largely calls for lower capital expenditures than in the past. Using sell-side consensus figures for the next two years of free cash flow (which are in my opinion very rosy - given what is happening in the industry, with overcapacity of pipelines in many places and US oil and gas production growth plateauing and likely declining, these estimates seem too high). However, we will use what is available.

Distributions / Free Cash Flow

Name	2017	2018	2019	2020E	2021E
Enbridge Inc.	(0.44)	0.80	0.65	0.82	1.02
Enterprise Products Partners LP	0.42	0.50	0.48	0.88	1.04
Kinder Morgan Inc.	1.26	0.62	1.06	1.21	0.99
Energy Transfer LP	(0.61)	0.04	0.63	1.20	1.77
ONEOK Inc.	0.87	0.03	(1.29)	0.34	1.07
Williams Cos Inc.	0.43	0.02	0.86	0.99	1.08
MPLX LP	0.41	0.46	0.46	0.97	1.16
Magellan Midstream Partners LP	0.53	0.66	0.18	0.99	0.93

Click to enlarge

Source: Company filings, company estimates, Bloomberg consensus for 2020E and 2021E. Estimates are for both free cash flow and distributions - e.g. assumes that KMI will continue to grow its distribution at a high rate, EPD at 2.5%, etc.

The data above suggests that the midstream companies, in the rosy scenario painted by sell-side analysts (which to be fair are informed by the management teams' own rosy picture), most of the companies will barely be able to cover their distributions with free cash flow in two years. In a scenario where free cash flow grows more slowly, companies may have to curtail their distribution growth or even cut distributions to keep afloat. If US oil and gas production does not grow as much as these companies expect (EPD CEO said on the 4Q19 call that he expects Permian growth to continue in 2025), the companies are going to find it difficult to grow their cash generation.

But what about valuation? Don't high yields compensate for this reality?

Not really. Since most of the companies are paying out more than the cash they are generating today, and the underlying assets over the long term are very tied to volatile oil and gas cycles, the yields should be high. There is always a debate as to how high - one which I will not wade into right now since it involves a discussion of what the US oil and gas industry will look like in 10 years. However, I think it is safe to say that the US oil and gas industry will not be bigger in 10 years than it is today. If oil and gas production in the Permian basin falls even half as fast as it rose (and this is possible, given wells in the Permian decline 60-70% in the first year of production), the cash flows that come from moving that volume should have a lot of risk priced in.

Again, regardless of what you look at, the free cash flow of the companies simply does not support the distributions.

2020E Yields

Name	DCF Yield	Dist. Yield	FCF Yield	Net Debt/ EBITDA
Enbridge Inc.	15.3%	8.5%	7.1%	4.8
Enterprise Products Partners LP	14.4%	8.1%	7.2%	3.2
Kinder Morgan Inc.	13.8%	6.4%	7.9%	4.4
Energy Transfer LP	23.1%	12.1%	14.0%	5.7
ONEOK Inc.	11.5%	6.3%	2.2%	3.9
Williams Cos Inc.	17.0%	9.2%	9.1%	5.0
MPLX LP	22.3%	15.4%	13.4%	4.1
Magellan Midstream Partners LP	11.5%	8.0%	8.0%	3.2

Click to enlarge

Source: Bloomberg, company filings.

Oddly enough, Energy Transfer, appears best positioned here. ET is widely acknowledged amongst large midstream companies to have a highly problematic management team, with a track record of poor returns on projects and M&A, causing a discount. ET also has the highest leverage (and some of the riskiest assets, 45% of the business tied to natural gas) on the list.

This brings me to the conclusion of my piece, which is to say that most midstream companies do not in fact have room to raise their distributions - in fact they may need to cut. Given what is currently happening in global oil markets and OPEC/ Russia's latest attempt to hurt US shale production, FCF estimates are almost certainly too high. Current valuations somewhat reflect this, however given the investor base in midstream companies (largely income oriented), distribution cuts will hurt the valuations. Before investing in these companies, it is important to understand that their distributions are not as safe as the companies lead investors to believe.

Analyst’s Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. 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.

Comments

[beginner75]

With no accounting background, I'm totally confused by this article. Having seen so many CPA disagreed with the author, is there any chart here which is at least useful as a comparison matrix between the various MPLs? I'm invested in EPD which is most frequently mentioned in comments but it doesn't appear to score well in many of the charts such as "Distributions / Free Cash Flow". KMI does better than EPD, if I've not read the metrics wrongly.

[HaroldL]

@beginner75

Yes, you are reading things correctly. KMI, a C-Corp, pays a lower dividend than EPD's distribution. MLP's usually have higher distributions than corporations.

There are other factors to consider, such as debt to EBITDA. And then there is actually understanding the businesses of the entities. KMI is natural gas pipelines, while EPD is NGL's.

[beginner75]

Thanks for clarifying. If this were the case, wouldn't KMI's dividend be safer than EPD, assuming that the future market dynamics for piping NGL and NG are similar and KMI doesn't do another slash like in 2015.

KMI has similar debt to equity ratio as EPD but EPD has higher current and quick ratio as compared to KMI, according to Seekingalpha - seekingalpha.com/...

Since a lot of the metrics are conflicting for the MPLs.

[HaroldL]

@beginner75

Current and quick ratios are not going to matter for these large investment grade companies (KMI and EPD) with steady revenue streams. The debt ratio that matters for pipelines is debt to EBITDA. Both KMI and EPD score very well at that metric.

KMI does not have a mountainous growth capex backlog to fund to position itself for the shale revolution like it did in 2015. KMI would have no reason to cut its dividend.

[Qniform]

Thanks for the article. I'm less concerned about the D&A matching maintenance expenditures than DCF assumptions about terminal value. As you alluded in a comment, these basins are not going to produce economically forever. What will pipelines be worth as stranded assets? Only major trunk lines connected to import/export facilities seem to avoid this fate.

[Justin Wiedeman]

So you think that the Operators are simply building billion dollar investments on a whim? Odd reasoning. But in that case, you should not be invested in the sector. Perhaps you sold calls that you don't want to execute?

Can't make sense out of this comment.

[JamesSutton]

What about export?

[sharpsicle]

In the Total Capex / D&A calculation probably a lot of those capex spends are for expansion rather than maintaining existing infrastructure.

I bought ENBL midstream because it's yield allows for a dividend cut of 50% and would still have an 18% dividend yield at $3.50. Yes the dividends will get cut if this low oil/gas environment keeps up. but also when oil and gas recover (which they eventually will) the share prices will move up in a big way.

So much short term thinking in the market creates very good opportunities if you're willing to hold for 3-5 years

[Robin Heiderscheit]

I own ENBL too (at much higher levels, $7ish).

I remodeled it this a m. If you haircut the selling prices by 5%, reduce volumes by 15%, and take one turn off the current EBITDA multiple (take it down from 6x to 5x) the equity is a doughnut.

But who knows, I'll ride it to zero rather than sell at these prices.

[R59]

I own BPMP and CNXM. Both have been steadily increasing distributions, although that may stop with these depressed O&G prices.

[clrodrick]

It is always darkest just before dawn, or so they say. Right now it is pretty darn dark!!!! Likely will get darker as 2020 progresses, but with MLP's collapsing by 50%+ I think it is okay to nibble at various positions. I have a long time frame and set it all to DRIP.

[clrodrick]

@clrodrick - down another 26% today. Ouch. No more buying for me until I see some stabalization

[Robin Heiderscheit]

can they go below zero? . . . asking for a friend

[PipelineDancer]

If a MLP goes bankrupt, you need to google CODI right quick and get out. You can owe taxes greater than your investment. Lots of seeking alpha articles written on it. But I don't see a lot of MLPs that I think could go bankrupt anytime soon, except maybe SMLP. Just be careful. I'm a big buyer of MLPs here.

[Edward Strover]

You've written a very helpful article. My question: isn't D&A what you are allowed to offset against taxable income to account for the depreciation of the asset - and maintenance capex what you actually have to spend to maintain it? Surely, if the taxman allows you to charge off more than you have to spend this is a good thing? However, if your argument is that they are under-spending compared to the what's required to maintain the asset in a serviceable condition then that's a very different thing. What do you think?

[Intelligent Walker]

Thanks for your question. Yes, D&A is a non-cash expense and maintenance capex is a cash cost. The point of my piece is that companies' "maintenance capex" figures understate the amount to actually maintain the assets, and so DCF is a flawed metric that overstates the safety of midstream distributions. It is very hard to know what true maintenance capex is, but I used D&A as a proxy to try and make a point. Clearly a number of readers in the comments have gotten very hung up on D&A, but the piece really isn't about D&A. It is about these companies cash spending and actual ability to maintain their assets and pay / raise their distribution.

[ChuckXX]

Wish you would have included HESM in your analysis.

[lee2253]

Of course nothing is certain. I'd rather be here than in a lot of other places!

[DpSht]

Charlie Munger would agree with you. EBITDA is bull****. The yield pigs are getting slaughtered.

[LongT888]

I would prefer to look at the long term safety of the distributions by looking at the type of assets the midstream is investing and invested in and the balance sheets ability to support them over a short term downturn. For example, moving product from a refinery to a plastics company is not the same as moving LGNs to other countries which is totally different than moving oil from a well to a refinery. The ones I am invested in have proven records of over 20 years of being smart about where to invest and they maintain strong balance sheets and credit ratings.

[Charles1w]

LongT888, Which one's are you invested in now?

[Silly Mee]

EPD has take or pay contracts with strong companies. Yes, this is so extreme even the best balance sheets in the industry are in question.

BUt it is backed by an extremely high coverage ratio. Also a low debt ratio and solid credit rating.

Low oil prices could result in the rich natural gas basins going back into production. As the associated oil natural gas and ngls dries up.

Nor are low oil prices a reason to stop the massive investment in petrochemical assets due to deeply discounted ngls. Taking market share form other naptha base competition in communist China.

[LongT888]

EPD and A little in Kinder Morgan

[gutcheck]

distr cut seems obvious. cant expect 20% yields.

[CapVandal]

Roger that. Or, it is possible the units increase.

[Kinderer]

I'm hoping they dont increase distribution levels. Reinvesting at 11% or even 6% is just fine.

[badbernanke]

The accounting for depreciation and amortization actually is not linked accurately to asset life. It is an often arbitrary accounting attempt to allocate capital costs to some period of time to "match revenue and expense." Maintenance capital expenditures -- expenditures to keep capital assets in good working condition -- and depreciation have no necessary connection , other than that maintenance capital can be expected to keep an asset running beyond its depreciable accounting lifetime. The analysis appears to me to be based upon a false premise.

What DCF tells one is how much cash is available for a variety of undertakings -- capital investment, dividends, stock buybacks, etc. When a company tells investors up front that the DCF is not sufficient for additional capital investment and that such an investment must be met by either borrowing or debt in addition to DCF remaining after distributions, it is basically another way of describing FCF as, potentially, a negative number. Using either metric really should get you to virtually the same analytical space.

One can argue about how much cash flow from assets should go to debt reduction -- if depreciation were the same as the time in which an asset wears out and becomes uneconomic or unusable, perhaps we should see entities (corporations and MLPs) aligning debt repayment for each capital asset with its corresponding depreciable life. But that does not appear to be done by any class of entity, probably because actual useful life and the depreciable life set by accounting rules are the same only in the rarest of circumstances.

[John Patrick]

Pipelines that have high utilization do not depreciate when you actually go to sell them...they appreciate...bottom line. Do you think the Transco pipeline has gone down in value? Is the colonial pipeline worth less today? Did Kinder Morgan take a loss on the sale of their Cochin pipeline or their Transmountain pipeline. Will DAPL be worth less 10 years from now?

There are pipelines that serve oil and gas basins with declining utilization rates. These companies will take impairment charges on these assets as market dynamics change. These companies will dramatically slash their future growth capex. Ultra low interest rates, dirt cheap stock prices, and mid double digit cash flows will be a nice set up for high distributions, share buy backs, and no more projects.

[Power Hedge]

This article shows a very misinformed idea of how midstream works.

Depreciation & Amortization is not a cash expense. It's a tax writeoff. Your car loses 1/3-2/3 of its value the second you drive it off the lot. Does that mean you're paying somebody $10-$20k to maintain it in the first year or two? Basically, D&A is just a tax reduction that the IRS says that you can take to reduce your taxes but it doesn't mean that money is leaving your bank account and its substantially more than it actually costs to maintain these assets.

Free cash flow is cash flow from operations minus capital expenditures (excluding companies like TSLA that make up their own definitions). CapEx for these companies is NOT just the money needed to maintain their existing assets. Every company that you listed is investing heavily to construct new pipelines and other infrastructure because the customers have demanded it and have already signed contracts to pay for it.

None of the companies that you mentioned is at any risk of bankruptcy or a distribution (midstream partnerships do not pay dividends) cuts barring massive bankruptcies upstream. This is a possibility, and that's why we need to look at the financial strength of their counterparties. This article does not do that.

[CapVandal]

OK.

But what about current PE's? These are GAAP.numbers, using GAAP D&A. They are mid single digit for several of your examples, and lower than 10 for most. So the firms are cheap on that basis, including full depreciation. Arguing about yesterdays management metrics is hardly news.

Everyone now knows the market is demanding FCF break even, heavily punishing asset investment. There is blood in the streets here. No one wanting a boring SWAN should bother. But if someone needs this article to tell them there is risk with anything involving O&G....they should be somewhwew else.

[billmatt2]

OMG, another poorly informed author that knows not of what he speaks. Depreciation is NOT a cash cost, particularly with long lived assets. How do these people get published?

[DpSht]

Depreciation is a very real cost that these guys are ignoring in order to pay higher distributions. You are the one who is misinformed. Please google Charlie Munger and his opinion of EBITDA. Hint: He calls it “bull****” earnings.

[creese]

A total waste of time, nothing but a bunch of gibberish!

[Hanger]

Just a reminder about what the bulls said about upstream MLPs a few years ago...

[georgios]

Wow, what an ignorance! You cannot understand even the meaning of depreciation or its practical impact on cash (see $0) or the difference between D&A and maintenance capex --and you are a contributor on SA! Wow