TOO - Teekay Offshore Partners L.P.

[heth247]

What is working capital in this business? Is it just payment timings or do they carry inventory?

 

Also, why would they need to increase it? Could it be due to the new vessels delivered in 2018 starting service?

 

Finally, on the “normalised FCF” method, why would you only include maintenance and replacement capex? Growth capex is usually excluded from FCF too. And you have  to have a lot of confidence in the ROIC on growth capex to consider growth capex FCF to equity.

 

Petec, those are good questions to ask but I don't have good answers, hope other people can help.

 

Regarding the "growth capex", if they are only invested in the shuttle tanker business (which is what we have been seeing so far after BBU took over), then I have confidence on the ROIC. But actually, I consider the order of the 6 new shuttle tanker more as "replacement" than "growth" because they have lots of old shuttle tankers. I don't think they need to invest more for "growth",  they just need to increase the existing fleet utilization rate to 100%, e.g. the Tow segment, the unemployed Arendal UMS, and new contract for Varg, Ostras and Piranema are all "growth" opportunities without the need for heavy growth capex.

 

 

[petec]

I consider the order of the 6 new shuttle tanker more as "replacement" than "growth" because they have lots of old shuttle tankers.

 

Fair enough, but then you definitely need to take this off FCF!

[heth247]

I consider the order of the 6 new shuttle tanker more as "replacement" than "growth" because they have lots of old shuttle tankers.

 

Fair enough, but then you definitely need to take this off FCF!

 

there is no doubt the FCF is severely masked for 2019/2020 due to the new build of the 6 new shuttles. The question is, are they going to continue ordering at such a high rate after 2020?

 

If we look at their existing shuttle fleet, they have 6 built in 1998/1999 (if you count two 50% owned into 1), which will retire soon. So I believe the 6 new shuttles will replace them. But they also have another 5 shuttles that were built from 2000~2002 (also counting two 50% owned into 1), which probably will retire after 2022. So I ask myself, are the 6 new shuttles in order enough to replace the loss of income from the 6+5=11 old shuttles? I think they probably can because those old shuttles probably are getting very low rates for now.  Another way to think about it, TOO's shuttle tanker segment ebitda margin is 47%, while KNOP's is close to 76% with newer ships. So I would think newer ships are earning much more than older ones.  Welcome to punch holes in in my thinking.

 

 

 

 

 

 

[Seth Lowry]

No - they were obligated to purchase these vessels under their MSA agreement, it was not a growth capex decision.  They will have replaced 10 of 33 vessels after this so capex will drop.

[Steven B]

So Seth, correct me if I'm wring, you're sating that these newbuilds are maintenance CAPEX? I guess the growth part will be the higher rate they command.

[heth247]

No - they were obligated to purchase these vessels under their MSA agreement, it was not a growth capex decision.  They will have replaced 10 of 33 vessels after this so capex will drop.

 

Thanks, didn't know that. Does the 10 include the 3 that are already delivered in Canada? Basically, in addition to the 6 that are being built right now, do they need to order another 1 or 4 new shuttles?

 

So Seth, correct me if I'm wring, you're sating that these newbuilds are maintenance CAPEX? I guess the growth part will be the higher rate they command.

Steven, my understanding is that they are "replacement" in terms of shuttle ship count, but "growth" in terms of ebitda. Seth, correct?

 

[Seth Lowry]

In the 10 I referenced it includes the 4 from last year.  These vessels should be considered replacement capex not growth capex.  Maintenance capex is just dry docking expense. 

 

However, incremental contribution margins on newbuild replacements flow through the income statement at inflation plus levels.  CEO spoke about this dynamic on the 2Q18 conference call if anyone wants additional details.  So we are replacing vessels at 1:1 but cash flows at slightly >1:1, But in my view, these shouldn't really be viewed as growth, they are just replacing a lot of the older fleet.

 

Replacement capex in this business is lumpy. I was just making the point that refreshing 1/3 of the fleet in 2-2.5 years when they are 20 year assets is a lot.  Replacement spending will drop considerably after these vessels because they won't have any material obligations coming due from their MSA.  TOO is basically spending >2x run-rate replacement cost for shuttle tankers right now.   

 

Run-rate EBITDA is $650M to $725M over the next few years

Cash + Pref Interest $220-$230

Maintenance $30-40

Replacement $200-$250

FCF ~40c to 50c 

[Seth Lowry]

Actually, I just checked, 2 of the 6 vessels are actually on order for the master agreement, the other four are replacement capex for the COA fleet. 

[heth247]

In the 10 I referenced it includes the 4 from last year.  These vessels should be considered replacement capex not growth capex.  Maintenance capex is just dry docking expense. 

 

However, incremental contribution margins on newbuild replacements flow through the income statement at inflation plus levels.  CEO spoke about this dynamic on the 2Q18 conference call if anyone wants additional details.  So we are replacing vessels at 1:1 but cash flows at slightly >1:1, But in my view, these shouldn't really be viewed as growth, they are just replacing a lot of the older fleet.

 

Replacement capex in this business is lumpy. I was just making the point that refreshing 1/3 of the fleet in 2-2.5 years when they are 20 year assets is a lot.  Replacement spending will drop considerably after these vessels because they won't have any material obligations coming due from their MSA.  TOO is basically spending >2x run-rate replacement cost for shuttle tankers right now.   

 

Run-rate EBITDA is $650M to $725M over the next few years

Cash + Pref Interest $220-$230

Maintenance $30-40

Replacement $200-$250

FCF ~40c to 50c

 

Seth, so you are saying new ships only get slightly better cash flow (inflation+) than the old ships? That is a surprise to me. Although new ships need to pay debt while old ships are probably debt-free, I would have thought new ships get much higher rate and they also have less drydock time in the first few years.  How shall we think about the fact that TOO's shuttle tank segment ebitda margin is in the middle 40%, while KNOP's is close to 80%?  That is a big difference. Is KNOP just having better contracted rate, or because of their newer ships?  Thanks.

[Seth Lowry]

Here are some rough modelling assumptions...

TOO_Cash_Flow_Model.docx

[heth247]

Here are some rough modelling assumptions...

 

thanks, Seth, that was very nice of you to share these models! A few questions:

1. You modeled the cash interest of shuttle tanker seg at ~$70M unchanged for the next three years. But aren't they supposed to take more debt when those new builds were delivered?

2. The detailed shuttle tanker analysis are very helpful. I can see that the big chunk of increase of ebitda comes from the COA pricing increase of 15% in 2019 and 10% in 2020. Are these assumptions or fact in the agreement? Sorry I haven't dig deep into their COA agrrement, where can I find the details of the COA?

3. What are the economics of the new ships? E.g. I can see you put there $14.2M in 2020 for the "Net COA ebitda newbuild contribution". Assuming this is for one ship, then we have a ship that cost $150M, but only earn $14.2M in ebitda, the ROA is < 10%. Is this the right way to look at it?

 

Btw, I have managed to get access to the MS analyst report that has $1 target for TOO. His quality of models is no where close to yours!

 

[Seth Lowry]

Just a quick update I forgot to add in one credit facility in that analysis, so cash interest is closer to $200. 

 

For Heth...

On the shuttle tanker interest, they are raising debt, but they are also swapping out an expensive acquisition facility.  But it is fair to add 3-5 million in 2020 to shuttle tanker interest as well. 

 

The COA pricing increases are estimated, although the company said they were confident of sustaining 20%-30% pricing increases a few quarters ago and nothing has really changed from a supply/demand standpoint for this part of the fleet - so those rates are just a discounted estimate. 

 

I would agree with your ROA math, but these vessels typically run a little cheaper than $150M. 

[heth247]

Hi, Seth, thanks for the reply.

 

For Heth...

On the shuttle tanker interest, they are raising debt, but they are also swapping out an expensive acquisition facility.  But it is fair to add 3-5 million in 2020 to shuttle tanker interest as well. 

 

Which expensive acquisition facility were you talking about, is it the 500M revolver at 7.3%, or the 250M public bonds at 7.1%?

 

Maybe I don't quite understand what you mean by "swapping out"... I thought they are going to add at least 540M new mortgages for the 6 new ships ( $150M x6 x 60%=540). Let's call it 500M since maybe those ships cost less than $150M each as you said. Assume 5% interest on this 500M, that is 25M extra interest payment. How could it be only 3~5M extra?  Are you saying this 500M new debt will replace some of existing debt, so the net of new debt is only something like 100M?

 

[heth247]

20-F is out today, just quickly went through it, here are a few things I found out:

 

[*]Seth was right about the cost of new shuttles. They say the total cost for two new shuttles ordered in July 2018 is $270M, so that works out to $135M each, not $150M

[*] 2018's operating cash flow was reduced significantly by a $83M "changing in working capital", of which $70.6M was "Advances to affiliate" (page F-33). This should answer Petec's question earlier.  So I guess this kind of "Advances" can be received back some time in the future, right?

 

Please add if you find anything interesting or worth of attention.

 

[whistlerbumps]

Why aren't interest derivative costs part of "normalized" FCF?  It seems like these swaps are part of their financing plan (ie. if they had just done fixed rate debt it would have been at a higher rate) so shouldn't those be considered normal?  In 2018, they lost 38mln because rates were below 2.9% (p.52 20F).  While these losses will decrease if rates go up, shouldn't that decrease be offset by an increase in the interest rate from their variable rate debt?  Thus, shouldn't we consider some level of interest derivative costs as part of ongoing financing costs?

[heth247]

Why aren't interest derivative costs part of "normalized" FCF?  It seems like these swaps are part of their financing plan (ie. if they had just done fixed rate debt it would have been at a higher rate) so shouldn't those be considered normal?  In 2018, they lost 38mln because rates were below 2.9% (p.52 20F).  While these losses will decrease if rates go up, shouldn't that decrease be offset by an increase in the interest rate from their variable rate debt?  Thus, shouldn't we consider some level of interest derivative costs as part of ongoing financing costs?

 

They also have unrealized gain of 56M on those derivative instruments so the net is about 12.8M gain for 2018. I don't know what should be the proper way to "normalize" this kind of thing. Shall we only look at the "realized" part? For 2018, 2017, 2016, all the realized part are losses (page F-29  of 20F): 39M, 77M, 60M. I guess that is because interest rate kept rising. But how do you model this? Maybe should just model it with a little higher interest expense?

[Seth Lowry]

New deck for reference...

Combined_Deck_3-2019.pdf

[heth247]

New deck for reference...

 

Thanks for sharing this, Seth. I don't understand TOO's IR, they spent time to create this presentation for a conference, but did not want to share it with investors directly on their website....

 

One question is about slide 14, I am not sure how to understand it. It says "Majors going from 900 to 300 kboe/d". Isn't that a bad thing for TOO, since their customers are mostly the "Majors" (Shell, ENI, XOM)?

[RadMan24]

It's a little silly to value TOO using adjusted EBITDA figures. It's much safer to value it on a cash basis, assuming a 10 year deleveraging, and going that route, than simply placing an adjusted "multiple" on it.

 

Adjusted EBITDA does not give a clear picture of TOO's financial flexibility, underlying business costs, and prior, let alone current, capital allocation decisions.

 

TOO looks cheap on an adjusted EBITDA basis no matter how you slice it. Just look at what EBITDA itself excludes: Dry dock expense, amortization of in-process revenue contracts, and interest expense. All recurring cash outflows, to the tune of $240 million (via cash basis) in 18. No sign of dropping for several years either, but it is acknowledged interest will eventually drop if they successfully pay down debt.

 

From that point of view, one just has to determine if the underlying assets provide enough punch to get through the next 5+ without sinking. Last year, they had $80 million in FCF. This includes net capex. It also includes an adverse working capital impact of probably $50-80 million that may reverse in 19 or 20. At $80 million, you have a FCF yield of 15%. Improve FCF by $50 million, you have a 25% yield, even after assuming all warrants exercised.

 

Seems like fair value around $3.25 or so a share. Also seems like a fair margin of safety.

 

[heth247]

It's a little silly to value TOO using adjusted EBITDA figures. It's much safer to value it on a cash basis, assuming a 10 year deleveraging, and going that route, than simply placing an adjusted "multiple" on it.

 

Adjusted EBITDA does not give a clear picture of TOO's financial flexibility, underlying business costs, and prior, let alone current, capital allocation decisions.

 

TOO looks cheap on an adjusted EBITDA basis no matter how you slice it. Just look at what EBITDA itself excludes: Dry dock expense, amortization of in-process revenue contracts, and interest expense. All recurring cash outflows, to the tune of $240 million (via cash basis) in 18. No sign of dropping for several years either, but it is acknowledged interest will eventually drop if they successfully pay down debt.

 

From that point of view, one just has to determine if the underlying assets provide enough punch to get through the next 5+ without sinking. Last year, they had $80 million in FCF. This includes net capex. It also includes an adverse working capital impact of probably $50-80 million that may reverse in 19 or 20. At $80 million, you have a FCF yield of 15%. Improve FCF by $50 million, you have a 25% yield, even after assuming all warrants exercised.

 

Seems like fair value around $3.25 or so a share. Also seems like a fair margin of safety.

 

There is no doubt that it is cheap at price of $1.16, even based on current FCF that has been greatly masked by the abnormally high CAPEX for now. But it is still unclear to me what is Brookfield's plan to get the price to $4+. If you view this as a "mid-stream" business, who would pay 9X EV/EBITDA for a MLP if it does not pay out a distribution? However, "increasing distribution" has never appeared in the business strategy in any presentation from them, since Brookfield's take over. What is their end-game?

 

[Steven B]

More ships a coming....

 

http://www.globenewswire.com/news-release/2019/04/04/1797379/0/en/Teekay-Offshore-Partners-Secures-New-414-Million-Debt-Facility-for-Four-Shuttle-Tanker-Newbuilds.html

 

 

[Bluffy]

From my point of view this reads like financing for 4 of the 6 newbuilds which were already on the book.

[Steven B]

Yeah i think you're right. Price action is crazy though

[heth247]

Yeah i think you're right. Price action is crazy though

 

I am not surprised about the price action at all. TOO shouldn't be trading at the low $1 level. A main reason for the $1 target from the MS analyst back in Feb was that there was a funding gap of $500M, and now that gap has almost been filled, and at the rate of LIBOR + 2.25% = ~5%, it looks good to me.

[tombgrt]

I should have bought more after reading the excellent posts here the last few months. I bought some long dated calls around November/December (so poorly timed indeed!) that I considered a zero already but could yet turn out to be fine.

 

Thanks for the discussion here, everyone! Excellent read.