Deutsche Telekom AG (DTE) Earnings Call Transcript & Summary

Good afternoon, and welcome to Deutsche Telekom's Webinar on accounting topics. This conference will be recorded and uploaded to the Internet. The presentation will be available on the Investor Relations website. As you can see with me today are Dr. Robert Hauber, Head of Group Controlling; and Andreas Puy from the Investor Relations team. Robert and Andreas will give an overview of various accounting topics that we hope will be of particular interest and useful to you. After this, we have time for Q&A. Before I hand over to Robert, please pay attention to our usual disclaimer, which you'll find in the presentation. And now it's my great pleasure to hand over to Robert.

Thank you, Hannes. Good afternoon, and to everybody and also a very warm welcome from my side to our webinar on a couple of topics we consider to be of interest to you. We'll first spend some time on the effect of the deconsolidation of the tower business. The major focus today, however, will be on the accounting differences between U.S. GAAP and IFRS, when it comes to KPIs, like EBITDA, net income, free cash flow and net debt. Or in other words, how to get from one figure to the other. We know that this has been a frequent source of misperceptions in the past and we hope that we can shed some light on a few related accounting complexities. In the end, I believe it's actually easier than you might have believed, and I hope we fulfill this kind of expectation. So, as said before, let us start with the deconsolidation of the tower business. As you all know, we agreed to contribute the GD Towers unit into a joint venture with DigitalBridge and Brookfield. We expect the closing of the transaction to happen in the next days or weeks. And what is important, our guidance for the full year 2023 will assume that the GD Towers is not included for the full year. So now let us have a look what happens to the adjusted EBITDA of the segments and the group. Now let me start on the left side, and first of all, we will lose in the GD segment, the EBITDA AL of EUR 0.8 billion. Please also keep in mind that we booked in 2022 EUR 0.1 billion for the held for sale effect. So that is that net EUR 0.7 billion. Now what's happening in the segment of Germany. Now first of all, and as you can see on the slide, there will be a benefit against the situation pre the transaction. So we had, before the transaction, and that's important in the headline EBITDA and OpEx, not a depreciation of EUR 0.7 billion. I will explain to you in a second why this is not a leasing but an OpEx. This is a leasing between the segments. And that is why, as it's deconsolidated on a group level, we decided and opted for the no group internal leasing liability with the introduction of IFRS 16. Now after the closing, we will see a new leasing OpEx of EUR 0.9 billion, which you see here in the slide, that indicates that the costs are going a little bit up, but this is very much driven by the allocation of the interest portion because, in the lease comps -- in the leasing IFRS, you have an interest part, which is a little bit higher at the beginning and goes down later in the years. And that is the reason why this is predominantly a little bit higher at the beginning. Then we have a further factor in this reconciliation, which looks a little bit strange, but which is easy to explain. This is the sale and leaseback gain, which is part of the usual sale and leaseback accounting. Now in simple words, the gain, which comes from the fact that we are sailing that above the book value will not be booked in the other operating income, as you maybe would expected. Instead, this part will reduce our Right-of-Use assets, and thus, consequently, will reduce lower lease-related D&A going forward. We expect that this sale and leaseback effect will be around EUR 0.3 billion, and this will last -- EUR 0.3 billion per year. That's very much important. And then this will last for the next 8 years. So all in all, the EBITDA AL for the German segment will be at the outset, roughly EUR 0.1 billion higher than before the transaction. Or in other words, the pro forma results in Germany, in 2022 will be EUR 0.1 billion higher than the actual result. Please keep in mind, and I will come back later to that, that this is a noncash view as the underlying cash payments are not affected neither by IFRS 16, nor the sale and leaseback accounting effect. I will come back later to that one when I explain the free cash flow effect. Now the effect on Austria, and therefore, the European segment is pretty much neutral so that we finally expect that the impact on the pro forma Deutsche Telekom ex-U.S. EBITDA will be in the magnitude of EUR 0.7 billion as you see it on the right side on the slide, which is the sum of the magenta bars of the segment, GD and the segment Germany. Now let's move on to the impact on the free cash flow and the net debt. Now what is obvious is that we will lose the cash EBITDA in the segment of GD. Then we will save some CapEx in the magnitude of EUR 0.3 billion on the other side. And as mentioned on the previous slide, the EBITDA tailwind of EUR 0.1 billion, we expect from the segment Germany is noncash. Now as we started this reconciliation from EBITDA AL here, we have to neutralize this effect if we want to bridge it to the free cash flow AL. Then there is another tax effect due to the fact that the tower business was profitable. We are paying taxes within the tower business. This will be deconsolidated, therefore, within the future consolidated figures, we will have less tax of Deutsche Telekom. So all in all, we expect a negative effect on the free cash flow of minus EUR 0.2 billion. However, there is one thing which we need to explain. And that is basically that we will get from the operations in the future cash back. This will be in the first year's, probably not be dividend, which would usually shown in the free cash flow statement, but rather of the effect that we get the payback of the shareholder loan. As -- in our definition, this is not part of the free cash flow. You will not see it here. But this kind of cash, which is then coming from the operations and then paid to DT will for sure then reduce the net debt. We are not showing it in the free cash flow definition. But therefore, you could say, if you consider that one, that the impact of the free cash flow is overall rather neutral. The impact on the net debt is very much like we have said in the announcement of the deal. So we expect a cash inflow of around EUR 10.7 billion. At the same time, we expect an increase in the lease liabilities of net EUR 4.5 billion, which is consisting on the -- of the 2 effects, which you see here in the slide, which is, on the one side, EUR 6 billion of new lease liabilities, which are then compensated a little bit by the deconsolidation of EUR 1.5 billion of current ground leases. So the transaction will significantly decrease the net debt, including and excluding leases, and bring a considerable reduction in the leverage ratios. Now turning on to the next page, where you see the full effect on our P&L. And we didn't talk about revenue so far. That's why this chart gives you a full overview. So therefore, let me also explain you the effects on the revenues. So we will lose EUR 0.3 billion of third-party revenues. But, on the other side, we will continue to generate some revenues with the tower company, approximately EUR 50 million, so the net impact will be around EUR 0.2 billion. Then in the middle, you see then the effect on headline and EBITDA AL. The headline EBITDA loss will be only EUR 0.1 billion because we lose the former tower OpEx. On the way to the EBITDA AL adjusted, we slightly benefit from the deconsolidated lease OpEx for the existing ground leases. But of course, we have no lease OpEx from the no external MLA. You will recognize the total effect of the EUR 0.7 billion, which you also have seen on the Page 5. And to make the picture complete, we need to consider the deconsolidation of other than lease D&A and we covered the impact on taxes as well before. Overall, net income, and unfortunately, I cannot show you the full picture, overall, net income will be slightly diluted by 0.3 billion on first sight. However, this does not consider expected interest savings, which, for sure, will be a positive effect, and additional positive contribution from the JV on our financial income. We have not provided you with the exact quantitative impacts here as yet equity result will be dependent on the result of the purchase price allocation, which is not final yet. And interest savings will depend on our final decisions on how we use the proceeds in the next months, but you can be sure there will be some. So in a nutshell, we would only expect a very limited impact on the adjusted net income. With that, let me hand over to Andreas.

Okay. Thanks, Robert, and good afternoon to all of you as well from my side. Let us now move on to the main -- to the second main topic of today's presentation. The difference in the accounting of T-Mobile U.S. in our IFRS books versus the disclosure under U.S. GAAP. The next slide is taken from the appendix of our Q3 presentation and shows some of the variance in current reporting of and for TMUS under the different accounting standards and also communication preferences, which can sometimes be a bit difficult to reconcile, whereas we adjusted core EBITDA, the adjusted EBITDA as per U.S. GAAP and the adjusted EBITDA AL as per IFRS. All these KPIs have their own growth rates, and even the bridging items can vary from year-to-year or quarter-over-quarter. But not everything is that complicated. Let us now first focus on the EBITDA. As you know, the leading KPI for TMUS is a so-called adjusted core EBITDA. A reconciliation between core EBITDA and the EBITDA after leases is simply the handset leasing revenue, which is being disclosed by TMUS. And as you know, TMUS has deemphasized handset lease as means of distribution and so have chosen core EBITDA as their key metric. Core EBITDA, of course, is a much better proxy for cash EBITDA than adjusted EBITDA, including these handset leases. And as you know, our core EBITDA has been growing strongly, our EBITDA has not. We decided to stick with the EBITDA given the temporary nature of this headwind. And this has resulted in a headwind for our reported EBITDA and additional complexity. The page here illustrates the magnitude of the impact from the handset lease unwinds, which also impacted our top line by the same magnitude. It is to be expected that handset lease revenues will further diminish in 2023. And so this headwind and the resulting complexity will become increasingly irrelevant from this year onwards. Next, let us take a closer look at the bridge between IFRS and U.S. GAAP. As we principally use the same handset lease revenue figure in both accounting worlds, the difference between core EBITDA and adjusted EBITDA as per U.S. GAAP and adjusted EBITDA after leases and adjusted core EBITDA as per IFRS is the same. But typically, with adjusted EBITDA after leases of DT for TMUS is a bit lower than the U.S. GAAP result for the adjusted EBITDA. There are many different bits and pieces that are being assessed slightly different between the 2 accounting worlds. However, when looking at the most relevant items, you end up with a relatively small number of factors. First, versus share-based compensation of TMUS, and a magnitude of roughly 0.4 billion in the first 9 months, which is not included in TMUS' U.S. GAAP adjusted EBITDA but included under IFRS. This is simply a question of a definition of the adjusted EBITDA. The second element is the impact of IFRS 16, which is another EUR 0.4 billion negative. What is behind that? Under U.S. GAAP, leasing expenses are straight lined over the time of a contract, while under IFRS 16, there is the so-called front-loading of the expense due to the interest element. While in a multiyear view, the expenses in both worlds are the same, the front-loading effect of IFRS 16 initially leads to higher leasing costs in IFRS 16. And a second, less important element here within the negative 0.4 billion is the so-called T3 transaction. We will cover that in detail a bit later, but the essence here is, in U.S. GAAP, T3 is depreciation, but in IFRS, it is leasing OpEx. The third element is an item, we will also come back to later when talking about net debt of the spectrum lease. The spectrum leases are simply a contractual liability under U.S. GAAP, and as such, part of the OpEx. But under IFRS, a non-leasing financial liability, which means that the expenses are lower below the EBITDA and consequently increase the IFRS 16 EBITDA for TMUS by 0.3 billion. All of these effects mentioned so far are pretty consistent over time. And that is not -- but that is not true for the next and last one. The fourth element of significance is the treatment of the so-called renewable energy purchase agreement or, REPAs. These agreements are embedded derivatives under IFRS, but not under U.S. GAAP and for fluctuation of this derivative is other operating income or expense under IFRS. While the impact in the first 9 months of this year was rather limited, they can be highly volatile from quarter-to-quarter. You may, for instance, recall that in the third quarter, the impact of REPAs was minus 0.1 billion. And in the full year 2021, there was a positive impact of plus 0.3 billion. And that was the main reason why the so-called bridge between U.S. GAAP and IFRS was actually at 0. And unfortunately, these effects are not really predictable. Other elements do exist as well. But as you can see, they play only a minor role in the bigger scheme of things. So in a nutshell, we would expect the bridge to be at roughly USD 0.6 billion annually in a typical year under the assumption that the derivative result is 0, which is our base case for planning purposes. Moving on to the next slide and to the net income, probably one of the biggest mysteries from the outside on first sight, but fortunately, not that difficult to reconcile. All things considered, the relevant effects are principally timing related. Let us begin with reported net income of TMUS. TMUS does not report or publish an adjusted net income, so we have to start at $1.1 billion of net income in the 9 months 2022. The impact of the derivatives of minus $0.3 billion is a big difference between the 2 accounting worlds. They are part of the ongoing operational result under IFRS, but not included in U.S. GAAP. So this leaves you then with reported net income for TMUS under IFRS of USD 0.8 billion. The special factors itself are identical to the special factors shown by TMUS in their quarterly reports, except for 1 item, the stock-based compensation. Because this is kind of a special factor for the adjusted EBITDA purposes of the U.S. GAAP results of TMUS. It is still also being included in the reported net income as per U.S. GAAP. Therefore, we subtract the stock-based compensation from the special factors and only add back the 4.4 billion of merger-related costs, or 1.5 billion wireline write-off, the 0.5 billion of others as per TMUS' disclosure, and then we apply a deferred tax rate of 25% on this. And this resulting 4.8 billion are then added to the 0.8 billion to arrive at an adjusted net income for TMUS of 5.6 billion. Finally, we then have to distribute this result according to the shareholding of minorities. And for DT, we, therefore, arrive at an adjusted net income contribution of 2.7 billion for -- of TMUS in the first 9 months. Let's now move on to the free cash flow. And as promised, this is by far much easier to handle. You can see that under IFRS, we have a slightly higher free cash flow result for U.S., free cash flow than under U.S. GAAP. What is driving this principally and structurally? We've heard that name before. It is the spectrum lease. U.S. GAAP considers this as an ongoing contract with payments recorded within the cash flow generated from operations, whereas under IFRS, this is a financial liability where the repayment of liability is recorded in the cash flow from financing activities. Other elements are not so significant and stem from a variety of different factors. Let us now move on to the net debt. When you look at the net debt outside in, you may be tempted to do the following calculation. You start at a net debt ex leases for DT as we disclose it. It adds EUR 110 billion. You will subtract the 70 billion of U.S. debt as per TMUS U.S. GAAP disclosure; adjust for the currency, which was around plus EUR 2 billion, and then you arrive at a figure of EUR 38 billion for the ex-U.S. business. However, this figure is missing the correct value by around EUR 5 billion. The correct figure for the net debt ex leases for the European business is around 33 billion. Conversely, in our books, the net debt of TMUS is EUR 77 billion and not EUR 71 billion. What is driving this difference? The reason is basically twofold. First, there are differences in the accounting standards; and second, TMUS and DT Group are using a different definition of net debt. First, there is the so-called tower obligation of approximately EUR 2 billion. This is a part of a tower obligation of USD 4 billion disclosed by TMUS. We will cover the other 2 billion a bit later. Also known, this part of the tower obligation also known under the name Global Signal. In 2005, Sprint entered into a lease out and lease-back arrangement of approximately 6,400 sites with Global Signal, that is a third party that was subsequently then acquired by Crown Castle. Due to its specific nature, this transaction is classified as a financing transaction, both under U.S. GAAP and IFRS. However, whereas we chose to present this as part of our net debt, TMUS decided to define its net debt as excluding tower obligation. In both accounting worlds, the assets and liabilities are on the balance sheet and we record interest expenses on it. The second bigger factor is the treatment of spectrum leases. We've talked about that item already regarding the EBITDA and the free cash flow bridges on the previous pages. TMUS leases spectrum from various partners with varying expiration terms, which generally extend through 2050. The total spectrum lease and service commitments, including renewal periods are around USD 6 billion. These leases are a financial liability, not a leasing liability under IFRS, and as such, part of the net debt with a net present value of EUR 2 billion. In U.S. GAAP, these liabilities are not liabilities, but basically a contractual liability or obligation to be executed and is not reflected on the balance sheet, neither as financial liability nor as a leasing liability. And finally, there is the other. This 1 billion of difference is mainly attributable to the different treatment of embedded derivatives and slightly different factoring treatment. Let us now move on to the view of the net debt, including leases, and do the same exercise. And let us begin again with the IFRS 16 figure of 152 billion. We then subtract EUR 71 billion of TMUS as U.S. GAAP net debt. We further subtract USD 34 billion of operating lease liabilities, we all know from TMUS' disclosure, or adjusted for currency, currency EUR 35 billion and only the operating lease liabilities as for financing lease liabilities are already included in the net debt of TMUS as per U.S. GAAP. Then you would arrive at EUR 46 billion of net debt inc leases for the ex-U.S. business and would be off right figure by approximately EUR 8 billion, as correct figure is EUR 38 billion. What is now included in the EUR 8 billion? There are 3 items we already know from the previous page. First, the tower obligation or Global Signal with EUR 2 billion; second, the spectrum lease with another EUR 2 billion; and the #3, others with EUR 1 billion. And then additionally, another tower obligation of EUR 2 billion. As for TMUS' balance sheet, the tower obligation has a value of USD 4 billion, which is made up of Global Signal and the so-called T3 sale and leaseback transaction. What is T3? In 2012, TMUS entered into a sale and leaseback agreement with Crown Castle, involving approximately 6,200 tower sites. However, due to the specific nature of these contracts and subtle differences in the way how this contract was assessed by IFRS and U.S. GAAP, this transaction was a so-called failed sale and leaseback under U.S. GAAP, but a leasing contract under IFRS. Consequently, the assets remained on the balance sheet of TMUS as property, plant and equipment, and the purchase price paid is a financial liability. Whereas they were assessed as a leasing contract under IFRS, and with the introduction of IFRS 16 then recorded as a leasing liability on the balance sheet. So again, in a nutshell, the tower obligation consists of 2 items; Global Signal transaction done by Sprint, and the T3 transaction done by T-Mobile U.S. Both are a financial liability under U.S. GAAP but excluded from the net debt presentation. And in IFRS 16, Global Signal is a financial liability and T3, a leasing liability. So what do we want you to take away from today's webinar? First, on the tower transaction, this will lead to a significant deleveraging of DT Group. Overall impact on free cash flow and EPS is rather limited. But keep in mind, the accounting treatment of the cash flow, of the cash inflow from joint venture in the first 2 years being in cash from financing activities. However, we will make those transparent. Second, on the U.S. The key message, I think, is that the reconciliation of free cash flow, EBITDA and net income between 2 accounting worlds in the end is not that difficult as you may believe on first sight. On the net debt, I think the key message here is that the leverage of the ex-U.S. business is considerably lower than you would think. So this was -- I've come to the end of my prepared remarks, and we're now looking forward to your questions, and I hand it back over to Hannes.

Excellent. And thank you very much, Robert, and thank you very much, Andreas, for these explanations. Now we can start with the Q&A part and give you a bit of time for questions. [Operator Instructions] So it's almost as complicated as the accounting, but I'm sure you can handle it. And so the first question is from Ottavio at SocGen.

Could you hear me?

Yes.

Probably, you don't just see me. Let's say that.

Hear and not yet see, but yes, now we can see. Excellent.

It does clarify a number of points that it's always unclear to follow through. So that's pretty useful. And my question related to the first part of the presentations on the tower. Now it's pretty clear how you do the accounting. But I just want to understand a bit of the sensitivity because it's clear, the cash proceeds. But when you come to the leases, is it correct that you only capitalize the first term of the lease, not the number of years you plan to use the towers. So if you can give us a bit of sensitivity, if you would have capitalized not the 8 or probably it's 8 years from memory, it should be the first term, but 16 or 32, how would have been the impact? And is it also the case that like it happened in the U.S., in the next 8 years, if they asset core remains as it is, we're going to get more liabilities coming to the balance sheet because you have to renew, and therefore, you have to capitalize the lease, again, and therefore, will be added to the balance sheet. And if you can also tell about the WACC you use on the NPV? And the fourth one, it's, this is basically the impact next year. But I believe that the GD Tower is investing a lot to build new towers. If you can give us a bit of sensitivity how the impact will move over the next 6 to 8 years, at least for the first term of the lease?

Okay. Let me start. So I think the accounting treatment is that the first 8 years will be accounted for and that's it. I mean, of course, you can then take the 8 years and then scale forward -- go forward 8 years and assume that there will be a new agreement, and then it's up to you how you want to, let's say, take account of this. But also, you are surely aware that durations differ. And 8 years is probably a pretty fair duration for an average tower liability in the context of our industry. But again, so we're not commenting on the extensions in terms of their nominal value or present value. But as you also know, it's -- there's 3 extensions. So the total duration is around 30 years. On the second question, we are not disclosing the WACC. We are disclosing the liability. And then of course, we have annual increases related to the volume growth of the towers, the -- since we have coverage increase in Germany, but also there will be a capped, let's say, inflation related increase capped at 3%. And then you also have the gross value of the leases, which is the 6 billion. So you could actually almost reverse it out, but we are not providing this number. Maybe Andreas you take over here?

Yes, sure. I mean, look Ottavio, how you should look at this. And I don't really want to give guidance here. But there are a couple of things you know. You know that we want to expand the mobile network in Germany. We have been crystal clear on that on our Capital Markets Day. We have been crystal clear on that one during the entire flow of the presentations we did so far. So what is going to happen from an accounting perspective is, we now have a leasing contract on the number of sites. Now let's assume we're adding additional sites to the network next year, when this basically would constitute a new leasing contract and thus then constitute a new leasing liability. So that is how it's going to work mechanically. Is this going to be, in the end, finally, a big surprise to you? I don't think so, because in the end, this is also not a big surprise to us, because it is a part of our planning. And you know that we've always communicated that we want to further extend our network reach and also the network density in Germany, okay?

Okay. Not sure if this completely answers your questions, but that's as much as we can provide right now. And then, of course, we will have to also talk about these things as we move along. But currently, I'm not seeing further questions. [Operator Instructions] I'm seeing that we have a question from Mathieu. Mathieu, can we please have your question. Mathieu Robilliard.

Can you hear me?

Very well.

Great. Thank you very much for the presentation. That was very helpful. I had a question about how the rating agencies look at all these adjustments that you described with regards to the net debt that TMUS, the net debt that the group, when they define their ratings? I mean, is that something that is well understood and taken into consideration by the rating agencies to your knowledge? Or is there some work to be done there?

So the rating agencies to the best of my -- I mean these are all completely transparent items, right, that are fully understood by the rating agencies. As you know, the rating agencies make their adjustments. They adjust for leases. They will try to have a consistent view of the indebtedness of the company. We are both in investment grade anyway. So I think, it's -- they are well understood. I think that's an interesting discussion how rating agencies look at -- to the previous question, from Ottavio, how they look at the accounting treatment versus the economic realities of tower leases, right? Because in the end, all tower leases are basically perpetual, but the accounting treatment can be very different. And surely, we are not known for short durations. We are known for long durations, but I would say this is all very transparent to the credit world. I think it's less transparent to the equity markets, which is why we provided this, let's say, additional explanations today. Not sure if this is, let's say, I can't speak for the rating agencies in precise detail. Anything you want to add, Robert or Andreas?

No, I think I fully agree with Hannes. We do have our annual conversations with the rating agencies. They are fully aware of our planning and they are also fully aware of our balance sheet, and let's say, a bit complicated issues of coming from one accounting world to the other.