Author Topic: ATCO - Atlas Corp  (Read 230325 times)

petec

  • Hero Member
  • *****
  • Posts: 2636
Re: ATCO - Atlas Corp
« Reply #700 on: December 16, 2020, 02:00:56 PM »
Agreed. And it wonít just be the Washingtons. Thereís another big shareholder that needs cash flow, as discussed extensively on this board.

But also: while the hypothetical ROEs look great, I did point out that we donít know the real IRR because we donít know the lease terms (nor, indeed, the rates the ships will earn when the leases end). Also, these assets have *relatively* short-lives, and an x% ROE on a short lived asset is much less valuable than the same ROE on a long-lived one. Treat my maths as a paper exercise only.

I think the bolded part is key. Wouldn't returns be significantly lower once you account for required debt payments? The company has been trying to reduce the scheduled principal payments (which is fine by me and makes sense...to an extent), but these ships ultimately have a ~25y useful life.

So if you assume the required debt payments on these vessels are amortized over say 15y, IRRs on these recent deals look weak.

When Sokol joined Seaspan as Chairman, he berated the previous management team for levering up to build new ships, and then when the downturn hit, their balance sheet didn't allow them to buy (almost) new ships at less than 50% of cost (ie, the Sam Zell strategy - your competitive advantage is your opportunistic buying which gives you an asset cost base that is >50% lower than everyone else).

Newbuild ship costs haven't dropped by that much, and with charter rates not looking that different from ~2015 (I think) it seems like they're just doing what the previous management did. How is this point of view incorrect?

Depends on the timing of debt repayments but assuming theyíre back end loaded IRRs could be quite high.

Whatís your source on new build costs? Not that it matters much since the vast majority of capital deployed has been in acquisitions.

FFH MSFT BRK BAM ATCO LNG TFG


Nelg

  • Newbie
  • *
  • Posts: 24
Re: ATCO - Atlas Corp
« Reply #701 on: December 16, 2020, 04:43:53 PM »
Agreed. And it wonít just be the Washingtons. Thereís another big shareholder that needs cash flow, as discussed extensively on this board.

But also: while the hypothetical ROEs look great, I did point out that we donít know the real IRR because we donít know the lease terms (nor, indeed, the rates the ships will earn when the leases end). Also, these assets have *relatively* short-lives, and an x% ROE on a short lived asset is much less valuable than the same ROE on a long-lived one. Treat my maths as a paper exercise only.

I think the bolded part is key. Wouldn't returns be significantly lower once you account for required debt payments? The company has been trying to reduce the scheduled principal payments (which is fine by me and makes sense...to an extent), but these ships ultimately have a ~25y useful life.

So if you assume the required debt payments on these vessels are amortized over say 15y, IRRs on these recent deals look weak.

When Sokol joined Seaspan as Chairman, he berated the previous management team for levering up to build new ships, and then when the downturn hit, their balance sheet didn't allow them to buy (almost) new ships at less than 50% of cost (ie, the Sam Zell strategy - your competitive advantage is your opportunistic buying which gives you an asset cost base that is >50% lower than everyone else).

Newbuild ship costs haven't dropped by that much, and with charter rates not looking that different from ~2015 (I think) it seems like they're just doing what the previous management did. How is this point of view incorrect?

Depends on the timing of debt repayments but assuming theyíre back end loaded IRRs could be quite high.

Whatís your source on new build costs? Not that it matters much since the vast majority of capital deployed has been in acquisitions.

I agree the economics look good if debt repayments are back-end loaded but then the risk just gets transferred to lenders, and I believe most of the debt is recourse. I understand the "but this time the industry is an oligopoly" argument though, which might/should result in vessels not being scrapped before their useful lives (ie, like what happened during the last downturn).

Maybe I'm just a grumpy old credit guy, but creditors should want their principal back well before the useful life for the above reason, plus most customer credit ratings aren't that strong either. IMO the prices Atlas is paying for these vessels does not provide much margin of safety if there's a repeat of a ~2016-type downturn.

Newbuild cost is from a Jefferies report which sources from Clarksons, which is the same industry publication usually quoted in Seaspan's presentations. I don't know how to post pictures on here, but here's the data (I'm eyeballing it from graphs - I don't have the exact numbers):
- 8.5-9.1k TEU 5-year ranges:
NB costs: ~$82-90m, vs currently ~$87m
5y old vessels: $25-$70m, vs currently ~$50m.

- 13-14k TEU 5-year ranges:
NB costs: ~$105-115m, vs currently ~$105m.
5y old vessels (data only for 13k TEU): ~$85-$110m, vs currently ~$90m.

They seem to be buying stuff mostly in the ~11-12k TEU range (the last 2 acquired 12k TEU vessels were 2y old and $88m each), and I'm guessing the 5 12.2k TEU newbuilds they just ordered cost ~$500m, so [assuming similar charter rates] it seems like the economics between their acquisitions and newbuilds is similar.

Btw I own Atlas shares.
« Last Edit: December 16, 2020, 04:46:12 PM by Nelg »

ourkid8

  • Lifetime Member
  • Hero Member
  • *****
  • Posts: 1081
Re: ATCO - Atlas Corp
« Reply #702 on: December 16, 2020, 04:47:43 PM »
Are charter rates really not looking that different than 2015? I do not think rates have been this high since ~2008! Please see the below link.  When purchasing 5 new builds that are already backed by 18 charters at the current elevated rates...Maybe they are not following the same path as the previous management. 

https://harpex.harperpetersen.com/harpexRH.csv

Newbuild ship costs haven't dropped by that much, and with charter rates not looking that different from ~2015 (I think) it seems like they're just doing what the previous management did. How is this point of view incorrect?
« Last Edit: December 16, 2020, 04:52:24 PM by ourkid8 »

Nelg

  • Newbie
  • *
  • Posts: 24
Re: ATCO - Atlas Corp
« Reply #703 on: December 16, 2020, 07:16:00 PM »
Yeah sorry, I meant the contracted rates of the stuff they've been buying (not spot rates which as you noted are extremely strong right now). They said the ships they're building/recently acquired are generating EBITDA of ~$10m/year, so assuming ~65% EBITDA margins, the revenues seem to be similar to the 15 vessels delivered to Yang Ming in 2015-16 (which are generating revenues of ~$16.5m/year).

But anyway you're right, the 18y newbuild contracts at those rates are far better vs say the YM ships which were only on 10y contracts.

Edit: Should note the YM vessels were 14k TEU, so the recently-announced newbuilds likely have slightly lower newbuild costs than the YM ones. So yeah the economics may very well be good on those, assuming no MSC (or whoever the liner was) credit issues.
« Last Edit: December 16, 2020, 07:20:00 PM by Nelg »

gfp

  • Lifetime Member
  • Hero Member
  • *****
  • Posts: 2150
Re: ATCO - Atlas Corp
« Reply #704 on: December 17, 2020, 06:25:06 AM »
An update on how these notes priced, conversion prices, etc  -

https://www.sec.gov/Archives/edgar/data/1794846/000119312520319538/d63418dex991.htm

NBL0303

  • Full Member
  • ***
  • Posts: 222
Re: ATCO - Atlas Corp
« Reply #705 on: December 17, 2020, 06:42:11 AM »
An update on how these notes priced, conversion prices, etc  -

https://www.sec.gov/Archives/edgar/data/1794846/000119312520319538/d63418dex991.htm

I must confess I'm not sure I understand the net effect of the capped call transactions, i.e. why they are choosing this structure. Anyone have any insight on that?

gfp

  • Lifetime Member
  • Hero Member
  • *****
  • Posts: 2150
Re: ATCO - Atlas Corp
« Reply #706 on: December 17, 2020, 06:55:09 AM »
Not sure if this helps but this is how I understood it -


NBL0303

  • Full Member
  • ***
  • Posts: 222
Re: ATCO - Atlas Corp
« Reply #707 on: December 17, 2020, 07:55:41 AM »
Thank you gfp - I guess my question is more like, why did Atlas choose this structure for this offering? What's in it for them now versus other types offerings?

petec

  • Hero Member
  • *****
  • Posts: 2636
Re: ATCO - Atlas Corp
« Reply #708 on: December 18, 2020, 12:01:16 AM »
Sokol grew Mid-American at a 23% ROE for about 20 years.  His strength is capital-intensive businesses and making them very efficient.  Bing is no slouch either!  Cheers!

Yes. Iíve never really trusted that number because my source is Prem and I think he can be a bit sloppy with figures. But Sokol is definitely one of my big reasons for owning Atlas.

Actually, all you have to do is look at what Mid-American was earning when Berkshire acquired it in 1999 ($104M) and what it was making in 2010, shortly before Sokol left...$1,131M.  That's about 26% annualized over 10 years in earnings growth.  He essentially did the same thing 10 years earlier growing earnings from $10M to $104M.  Cheers!

Well, no. You also have to look at whether Berkshire contributed capital during that time. Iím sure you and others know the answer, but I donít, which is why I take the CAGR as unproven!

Yes, BRK invested about $5b of additional equity capital after the acquisition (and another $1b of equity capital after Sokol left to fund the NV Energy acquisition).

If anyone's wondering, this is the history on CalEnergy/MidAmerican (I have gone through all their 10-Ks, though not in great detail - so I think the below is mostly correct:
- Sokol took over in Feb 1991.
- FY 1991 results:
NI $26.6m (vs 1990 NI $12m)
FFO $47.6m (don't have the 1990 number)
FD shares 36.5m (don't have 1990 FD shares outstanding)
= EPS $0.73/share & FFO $1.31/share

I don't have 1991 balance sheet numbers, but the below are 1992 numbers:
Recourse Net Debt/ FFO: -0.3x (ie, net positive unrestricted cash)
Total Net Debt/ FFO: 2.8x

- BRK acquisition of MidAmerican closed sometime in 2000.
- FY 2000 results:
NI $133m
FFO $597m
FD shares 43.8m
= EPS $3/share & FFO $14/share

Recourse Net Debt/ FFO: 3.6x
Total Net Debt/ FFO: 9.8x

Leverage was clearly higher, but when BRK acquired MidAmerican, they owned 2 regulated utilities in the UK and US and had a larger independent power plant portfolio...ie, the business quality was superior vs the small geothermal portfolio they owned when Sokol took over.

- IMO the "Sokol era" ended around 2009. He sold most of his MidAmerican shares (ie, to BRK) in 2009 for a cool $123m, and started the NetJets restructuring around this time.
- FY 2009 results:
NI $1.16b
FFO 3.3b
FD shares 75m
= EPS $15/share & FFO $44/share

Recourse Net Debt/ FFO: 1.6x
Total Net Debt/ FFO: 5.2x

There were huge one-time gains from their "failed" acquisition attempt of Constellation Energy Group in 2008 - they made a ton of money from termination fees and their financial restructuring there. And of course, Sokol was credited for the BYD investment which was made in 2008 and I believe is still held within the MidAmerican/BH Energy entity.

Thanks, this is useful. So 15/0.73=20x eps in 18 years for a CAGR of 18%, partly driven by an increase in leverage? That seem fair?
FFH MSFT BRK BAM ATCO LNG TFG

petec

  • Hero Member
  • *****
  • Posts: 2636
Re: ATCO - Atlas Corp
« Reply #709 on: December 18, 2020, 12:30:46 AM »
Agreed. And it wonít just be the Washingtons. Thereís another big shareholder that needs cash flow, as discussed extensively on this board.

But also: while the hypothetical ROEs look great, I did point out that we donít know the real IRR because we donít know the lease terms (nor, indeed, the rates the ships will earn when the leases end). Also, these assets have *relatively* short-lives, and an x% ROE on a short lived asset is much less valuable than the same ROE on a long-lived one. Treat my maths as a paper exercise only.

I think the bolded part is key. Wouldn't returns be significantly lower once you account for required debt payments? The company has been trying to reduce the scheduled principal payments (which is fine by me and makes sense...to an extent), but these ships ultimately have a ~25y useful life.

So if you assume the required debt payments on these vessels are amortized over say 15y, IRRs on these recent deals look weak.

When Sokol joined Seaspan as Chairman, he berated the previous management team for levering up to build new ships, and then when the downturn hit, their balance sheet didn't allow them to buy (almost) new ships at less than 50% of cost (ie, the Sam Zell strategy - your competitive advantage is your opportunistic buying which gives you an asset cost base that is >50% lower than everyone else).

Newbuild ship costs haven't dropped by that much, and with charter rates not looking that different from ~2015 (I think) it seems like they're just doing what the previous management did. How is this point of view incorrect?

Depends on the timing of debt repayments but assuming theyíre back end loaded IRRs could be quite high.

Whatís your source on new build costs? Not that it matters much since the vast majority of capital deployed has been in acquisitions.

I agree the economics look good if debt repayments are back-end loaded but then the risk just gets transferred to lenders, and I believe most of the debt is recourse. I understand the "but this time the industry is an oligopoly" argument though, which might/should result in vessels not being scrapped before their useful lives (ie, like what happened during the last downturn).

Maybe I'm just a grumpy old credit guy, but creditors should want their principal back well before the useful life for the above reason, plus most customer credit ratings aren't that strong either. IMO the prices Atlas is paying for these vessels does not provide much margin of safety if there's a repeat of a ~2016-type downturn.

Newbuild cost is from a Jefferies report which sources from Clarksons, which is the same industry publication usually quoted in Seaspan's presentations. I don't know how to post pictures on here, but here's the data (I'm eyeballing it from graphs - I don't have the exact numbers):
- 8.5-9.1k TEU 5-year ranges:
NB costs: ~$82-90m, vs currently ~$87m
5y old vessels: $25-$70m, vs currently ~$50m.

- 13-14k TEU 5-year ranges:
NB costs: ~$105-115m, vs currently ~$105m.
5y old vessels (data only for 13k TEU): ~$85-$110m, vs currently ~$90m.

They seem to be buying stuff mostly in the ~11-12k TEU range (the last 2 acquired 12k TEU vessels were 2y old and $88m each), and I'm guessing the 5 12.2k TEU newbuilds they just ordered cost ~$500m, so [assuming similar charter rates] it seems like the economics between their acquisitions and newbuilds is similar.

Btw I own Atlas shares.

I actually think the opposite argument might apply regarding cash flow timing and IRRs. Seaspan have been moving away from asset-level debt, most notably with their big secured lending facility. I don't think there are any principal repayments associated with that facility - from memory it has a bullet maturity, and is effectively a huge revolver. When investments are funded out of this facility (or indeed any other corporate-level bullet debt), 100% of early-year cash flows accrue to equity and the debt can be paid off with later cash flows. Obviously this comes with a risk, which is that they might have trouble extending the revolver, but if you are prepared to take that risk the IRRs could be very attractive.

BTW I disagree that the cost difference of $88m vs $100m suggests similar economics. Assuming similar cash flows and therefore similar debt, a 12% difference in total cost represents anything up to about a 50% difference in equity requirement. It's potentially transformative, and my guess is that the IRRs on the acquisitions are significantly higher than the IRRs on the newbuilds. For the newbuilds, Seaspan is effectively just a capital provider clipping a spread until the lessor buys the ships in 18 years. This reduces the risk (not to zero, but it reduces it) so it is reasonable to assume the IRRs are lower or the leverage to achieve the same IRR is higher.

Re scrapping of vessels, I had a very interesting chat with a shipper recently who pointed out that the widening of the Panama canal had a significant one-off effect on obsolescence. Essentially a lot of ships built for the old canal (Panamax) became obsolete overnight as larger, more efficient ships came into play. Today, the limits on ship size are imposed by port capacity and journey length (you can't use huge ships on short journeys because the dwell time in ports is too long). To achieve another step-change would require hundreds of little projects - dredging ports, extending docks, etc., across dozens of jurisdictions - and these will not happen in a coordinated fashion. This means the last 15 years have seen a one-off supply shock in the shipping industry. That is part of what caused owners to scrap vessels before their useful lives ended.

Add the fact that the industry has been working off the 2008 newbuild excess for over a decade, and there is very little spare capacity and virtually no order book, and I don't worry about early scrapping. The only thing that concerns me on the supply side is the belt-and-road initiatives which are increasing the capacity to move containers across Asia by rail.
« Last Edit: December 18, 2020, 12:34:06 AM by petec »
FFH MSFT BRK BAM ATCO LNG TFG