Deutsche Pfandbriefbank AG

XETRA:PBB

Earnings Call Transcript 2023-Q2

[Andreas Arndt, executive]

Ladies and gentlemen, dear all. Thank you very much for dialing in and a very warm welcome to our analyst call regarding PBB's second quarter and half year results for 2023. I'm happy to walk you through the key developments, and we will have our usual Q&A session thereafter. The environment remained challenging in Q2, especially for the real estate markets in Germany, in Europe and in U.S. and the U.K. At the same time, interest rates continued to rise also in the second quarter. Inflation kept high and some economies are at the edge of recession. Yet despite all that, PBB remains on track. PBT was EUR 81 million for the first half. On this basis, we are happy to confirm our full year guidance of EUR 170 million to EUR 200 million for 2023. As you know, the bank that we were fully booked in Q1, no further effects in the second half of '23. And looking ahead, we expect the second half to benefit from increasing NII dynamics. Also potential risk provisioning should benefit from our solid stock of provisions and management overlay.

Real estate funds portfolio strongly increased by EUR 900 million to EUR 30.2 million with improving margin.

There were almost no prepayments, new business margin stays an elevated level of 200 basis points. The new business volume starts low in reflection of market environment. Therefore, we adjust our full year guidance to EUR 6.5 billion to EUR 8 billion. Still a good figure, we believe in current market environment as we stay selective and rather opt for strong margins than for higher volumes or risk. After all, this is what PBB stands for, being a risk conservative business, delivering returns in all parts of the cycle.

All in all, we expect our real-estate finance portfolio to stay above EUR 30 billion. We make good progress, very good progress on our strategic initiatives in a tightly focused manner. We initiated a cost-cutting program, which aims to reduce cost to the 22 level with medium-term target cost income ratio of below 45% by 2026. Against forecast cost for 2023 or gains planned cost levels for 2026. This amounts to an asset of around EUR 40 million. Which would bring us then eventually back to the 22 levels amounting to EUR 224 million and a cost-income ratio of below 45%.

First, the strategic initiatives as we have already laid it out to you last time, the progress is on all categories, and I'll walk you through that briefly. I'm talking to Page 6. As mentioned, the real estate finance portfolio has grown strongly to now above EUR 30 billion from EUR 29.3 million as of last year, 22 million with improving portfolio margin. Even in times of low new business, we grow our book profitability to a margin improvement and careful asset selection extension management.

Furthermore, we increased our retail deposit base to EUR 5.6 billion, which is EUR 1.2 billion up compared to EUR 4.4 million by the end of last year. And just recently, we entered into a cooperation with [indiscernible] investment platform, and we further broadened our brand and distribution channels, and therefore, our deposit franchise value. We also made further progress with regards to our earnings diversification initiative. The operational setup of our real estate investment management is in progress. The strategy for our first product has been defined. We plan foreign equity as well as [ on ] that product and further details will follow soon expectedly at the extra real fair in Munich end of October this year.

With our expertise in commercial real estate sector and in close cooperation with our partners, Universal investment, which is our custodian and Amundi, which is our sales partner, we are on the way to develop and expand PBB invest into a strong investment manager. It is -- we still believe that this is exactly the right time to set up a solid and expandable infrastructure without legacies to be ready to start when the market returns. Now overall, the markets have not yet recovered, but we believe that opportunities will arise in the coming months to build up attractive portfolios for institutional investors, both on the debt and the equity side.

And the next point is digital credit workplace the credit process has been put to the next level, increasing our process efficiency. And the noncore unit optimization is running public investment finance and value portfolio has been bundled into the new non-core unit. We are thus concentrating on the real estate finance and real estate investment management areas, which are promising for our business model, and we'll continue to invest here both in terms of personnel resources and capital.

The benefits of merging the parts of the noncore unit can be seen already in Q2. As the entire public sector portfolio is [ narrow ] and run down, the previous narrow restrictions on asset sales do not longer apply. This gives us more room in the individual selection of assets that can be solved at a profit. And with a stronger reduction on the asset side, the possibilities for bond buybacks expand in times of rising interest rates, the interest induced spreads are also increasing and enable buybacks at favorable conditions as was to be seen. In this way, we actively are opening up another possibility for the bank to benefit from the current higher interest rates as an alternative to the loss of lost shortflow income and in addition to the substitution advantages already described for cheaper retail deposits.

One of our further key strategic initiatives is the green transformation, where we also made good progress. First, as you know, we target the green share of our real estate finance portfolio at least 30% until 2026. The share is now around 20%. 70% of the portfolio are scored and have been vetted the green share of [indiscernible] portfolio is at around 28%.

With regard to Green Consulting, we found a new estate entity called Eco Estate GmbH, together with our partner, Groß & Partner which aims to advice on green transformation projects and the first joint projects are in preparation or actually in the pipeline. We want to develop this into an overall managed to green approach to be leveraged also in other business fields, notably in our real estate investment management. And that brings me to another key element of our strategic agenda for '26, the cost efficiency, and that's on Page 7. When detailing strategic initiatives, we announced that we would also focus more on our cost. Despite churn cost discipline, we currently see and we'll see in the current year, visibly higher costs from significantly increased regulatory requirements and first and foremostly, from necessary strategic investments. We have, therefore, taken extensive measures to significantly reduce our current costs, i.e., in the order of around EUR 40 million. The cost program as part of our strategic project B 2026 aims to bring the cost-income ratio back to a level of 45% by 26% and thus to the -- basically back to the levels which we did show in the year 2022 at an amount of EUR 224 million.

Cost savings were spread into 40% personnel costs and 60% nonpersonnel. And this corresponds to the gross FTE reduction of approximately 15% for 130 FTEs over the next 3 years to come. In addition, there are approximately 20 FTEs of CAPVERIANT, which we already communicated to you, which are also being made redundant, I come back to that. Main drivers of personnel cost reduction will be increased process efficiency, especially credit workplace, the digital platform for our credit flow in that process reviews and discontinuation of the Public Investment Finance and CAPVERIANT segments.

Of course, we will carry out these reduction measures in a socially responsible manner and take into account company agreement and demographic development structures and natural fluctuations. And we have every intention to start with the exercise before the end of this year. In terms of CAPVERIANT, we have largely completed the measures, as previously announced, by the end of the third quarter, both in terms of personnel as well as in terms of infrastructure. Nonpersonnel cost reductions will predominantly be driven by reduction of [ IT -- ET ] through in-sourcing or a new service contract and consultancy costs as well as specific cost savings over the whole organization as a result of an in-depth cost review.

Program costs are already largely covered either through existing provisions related to CAPVERIANT on the credit workplace or by cost-neutral measures such as natural fluctuations. Additional costs will be financed by countermeasures and have already been accounted for in our business planning and guidance. With that, let me come to Page 9 to the operative highlights for the first half '23. On the business side, real estate finance new business stays low at EUR 2.5 billion versus EUR 4.3 billion last year in reflection of the overall market sentiment and market environment with some good news in that. The volume shows the increase in dynamics with EUR 1.5 billion in Q2 after EUR 1 billion in Q1. And the average gross interest margin stays at an elevated level of 200 basis points, i.e. 30 basis points up versus last year. And C, real estate finance portfolio strongly increased by EUR 900 million to EUR 30.2 billion with improving margin on the stock. On the funding side, we made strong progress in further building up our retail deposits and substitution for senior unsecured market funding. Retail deposits are up by 27% year-to-date with currently further strong inflow.

Now if you follow me, please, to Page 10, a brief overview on markets and I'm sure there will be further questions coming up later on. All overall, there's no secret around the commercial real estate. Transaction volumes did see a further strong decline in Q2, both in Europe and the United States, as you can see from the chart and the determining factors for commercial in the estate remain difficult and uncertain. Further increased interest rates with the remaining uncertainty on further development, although we assume that there will be some flattish development over the next months to come. persisting high inflation and in addition, structural trends, such as working from home and ESG, which caused further uncertainty and pressure on price. Property prices remain under pressure with strong differentiating dynamics between prime versus nonprime, core versus noncore, inner city versus periphery, green versus non-green and so on. While we expect markets to come to converging the use on the relevant pricing components, such as the interest rates. All in all, the prices on real transactions are not expected to bottom out before 2024.

Yet, I think, especially in such a situation, PBB business model proves its value. PBB is rather positioned on the conservative side, i.e. predominantly prime and core with conservative risk parameters and focus on strong structures. PBB's stock of risk provision provides a solid buffer parts of further expected property price movements already processed and reflected in Stage 1 and 2 model-based provisions.

And finally, with our strategic initiatives, we now lay a solid ground for increasing profitability and value creation.

Now Page 11 on the new business and the portfolio view. As I already pointed out, expectedly and in a reflection of the market environment, new business volume remained at a rather low level with EUR 2.4 billion. But with an increasing dynamic and with elevated margins from 200 versus 170 before. In this challenging market environment, we remain true to ourselves staying selective and balancing risk and return rather going for margin than for volume or risk. This naturally comes with a higher share of extension, which is around 41% for the first half '23.

Therefore, as mentioned, we adjust our new business guidance for the full year to between EUR 6.5 billion to EUR 8 billion, and we're further increasing new business dynamics and new business volumes in the second half of '23. Despite lower new business volume, we see positive development in the strategic graph portfolio, volume increased significantly to 30.2%, and also the portfolio margin is moving up. Now please allow me to skip Slide Slide 11. Keep that for your own reading that we have been very selective on the United States and U.K. new business should not come as a surprise. And on the Nordic side, as you can see, the larger-than-usual share is owed to a fairly large [ case ] development structure, which we did finance with the long-term client right in the middle of Stockholm. That's not a strategic shift per se.

On Page 12 -- oh page 13 it is. Real estate finance new business and portfolio view. That brings me to the topic of portfolio quality. Given the current market environment, it is not of a surprise that we see some movement in risk parameters. However, all overall, PBB remains well positioned with an average LTV of 51% and average [ IC ] of 300%. And in that context to that slide, please allow me for clarifying comments the LTV cluster show on the chart, which are shown on the chart are not sliced, i.e., the respective clusters are showing the total loan amount, not only the loan part applying to them.

As we -- in the recent days and weeks received several requests on valuations, some clarifications on that. Some additional explanations might be given at this point.

First, it is important to understand that in the context of valuations, PBB focuses on prime properties in core inner city locations with conservative risk parameters. Property values in these segments are more stable than nonprime, noncore B locations and B properties. And second point, NPLs do not necessarily cause risk provision. PBB Follows an approach with rather conservative risk parameters and conservative valuation. While for instance, the transaction might be classified as unlikely to pay. Overall, actual and most up-to-date valuation may return a collateral value that completely covers the outstanding loan, which we term as an NPL with 0 LLP. We continuously and intensively monitor our portfolio. It is not only about regular reviews, but it's a matter of constant monitoring. And as per the end of Q2, we have scanned our total portfolio by real estate prices with a particular focus on U.S. and office.

Expected valuation adjustments for PBB's portfolio in the second half of '23 and '24 are taken into account into our model parameters when we talk about Stage 1 and 2 risk provisioning and that is for the U.S. office portfolio and a further valuation adjustment of around 15% to 20% European office between 5 million to 10 million and total office portfolio 8% to 12%. And just to be clear on that point, these figures are based on the total portfolio scan and reflect PBB's individual portfolio structure and positioning, i.e. predominantly prime in core. And very important to note, they're built on already adjusted valuations from preceding quarters, i.e., on the adjustments, which we made already second half last year.

Now on Page 14, NPL portfolio. The NPL portfolio development reflects stressed market environment and thus as a dynamic, which should not surprise especially with regard to U.S. business and to the office segment. However, the NPLs which we took on, and that's most important.

We're coming in at EUR 17 million risk provisioning needs for new NPLs therefore, remain on the moderate level.

Now again, there are two more pages on portfolio quality, which is Page 15 and 16, which is an update to what we did show to you last time. I would skip that here. But if you have questions, please refer to me thereafter.

Now that brings me into the funding topics on Page 17. Funding activities reflect the optimization of our refinancing and liquidity structure in two directions: A, alignment with lower new business volumes and B, focus on retail deposits and substitution of senior and secured capital markets funding. Against this background, we launched a public buyback tender for three senior preferred and one public sector fund brief in Q2 with a final volume of around EUR 400 million. I'll get back to that point when we come to the P&L section. Furthermore, as of end of Q2, we repaid EUR 1.8 billion on the remaining TLTRO. So which leaves us with a remaining volume of EUR 900 million, which is planned to be repaid in '24.

Our liquidity position remains comfortable with an LCR of 163 and an NSFR of 114. And as you know, PBB manages its liquidity on a 6-months basis on a stress basis for 6 months i.e., the liquidity buffer must withstand a 6-month stress test, while on a regulatory note, only 1-month stress is required.

Now Page 18 on retail deposits. As I mentioned, retail deposits are up by 27% to 5.6%. The number of clients increased by 80% since [ 12 21 ] from 41,000 to more than 75,000. We continue to build out our franchise and brand marketing. In this context, we entered into a cooperation with [indiscernible] the operator [indiscernible] investment platform. 82% of the money of the inflow is term money with average term of 3 years. Potential unexpected outflows of core money are sufficiently covered by cash and cash equivalents, i.e. there's no liquidity risk from our point of view from any core money, which we have on our books. Retail deposits are nearly 100% guaranteed either by government deposit insurance scheme or by private bank scheme in Germany.

Let me go directly to Page 20. As usual, the overview on the financials. And as usual, a few comments on selected line items.

The first one to mention is the extraordinary strong net income from realizations, which amounted to EUR 42 million in the first half of '23. which mainly benefits from the sale of financial assets from noncore unit and the liquidity by -- sorry, the liability buybacks. At this point, the consequences of the strong increase in market interests are also noticeable for us with a time delay and more modestly, where other market participants benefit from rising interest rates from high site deposits. PBB participates by monetizing interest induced changes in values for asset sales and bond buybacks. And prepayments, which was a source of realization income in earlier times, prepayment stayed low due to the market situation. As usual, Bank levy, I mentioned that is being booked upfront in Q1 will not be a topic for the second half of the year.

Now with that, we turn to net interest income, which is Page 22. As mentioned in several calls before, compared to last year, the NII is being influenced by the fact that there is no further benefit from [indiscernible] and from flow income as interest have moved up significantly.

However, at the same time, the average real estate financing volume has increased strongly from 28% to 29.6% or 6%. And as an increased portfolio margin as a reflection of increasingly materializing effects from higher new business margins and retail deposits. On this basis, we expect the positive NII trend to continue and even show an increasing dynamic in the second half of the year with the NII increasing by approximately 10% in the second half.

Now page 23, on this provisioning is only up EUR 2 million against last year, i.e., EUR 21 million. Net additions in Stage I and II of EUR 10 million in Q2, mainly reflect increased PDs for individual financings and updated valuation parameters in the light of the current market environment. partially compensated by release of management overlay. As anticipated, uncertainty factors materialized in Stage 1 or 2 additions Foodstore.

The remaining management overlay of EUR 28 million covers assumptions about further deterioration of market values in office markets going forward. Net additions in stage 3 of minus 9 are driven by two U.S. office loans, partially compensated by a release from 4 deals, which we put back into the solid part of the business. The stock of risk provisions remains on a stable level, providing a cushion for potential future defaults to find that on Page 24. And you also see that we are fairly stable and flat on the basis points covering the portfolio at 130 bps and 34 bps.

Now on costs, which is Page 25, '23 remains a year of investment. Investments in strategic projects continue in line with our strategic agenda. The cost increase, which we see first half of '23, which will stretch into '23 2nd half as well, are related, first of all, to specific measures to strategic initiatives i.e., the digital credit Workplace, which now sort of comes into full production set up.

The second one is the setup of real estate investment management. Green consulting and an important thing because directly supporting our business is the marketing cost for retail deposits, just to name a few illustrative things. A further driver is the insourcing of functions, i.e. functions, which so far have been mostly project-driven being performed by external service providers or consultants which can be and shall be covered more efficiently by ourselves. And the first step, this comes with cost, but going forward, will result in a cost relief. And what we do against that, what we do in terms of cost-cutting program, I have just already reflected and presented to you.

Now turning to Page 26 on PBT, or what I've described returns of PBT of EUR 81 million, which underpins our full year guidance of EUR 172 million, EUR 200 million based on the following expectations for the second half. As mentioned, the positive NII trend is expected to continue, and I expect it to increase by 10%, benefiting from stable real estate financing volume or portfolio and further increasing portfolio margin.

Realization income expected to stay at low levels. So no further strong impact from second half of the year. As I said, bank levy has already fully been booked. So that's not a matter for the current quarter or the second half of '23 and the conservative in lasting loss allowances buffer risk provisioning going forward.

Now finally, a word about capital, that's on Page 27. The CET1 ratio is somewhat down to 16.0% from 16.6% for the second quarter '23 based on a slightly increased risk-weighted asset level in reflection of current market environment and somewhat lower capital position due to technical effects and the fact that we paid out a dividend.

Now to summarize my presentation and to put it briefly, we have stormy times and there's a lot of things to do, but we remain on track as we laid it out to you now at various times. We confirm our full year PBT guidance of EUR 170 million to EUR 200 million, which we expect to be supported by further increasing NRI dynamic.

We expect our portfolio to stay above EUR 30 billion while we expect the portfolio margin to further increase. We adjust our new business guidance to EUR 6.5 billion to EUR 8 billion in reflection of challenging market environment. However, solid pipeline supports increasing new business dynamics in the second half. Our conservative and lasting loss allowance buffer risk provisioning going forward. And most importantly, we push forward our strategic initiatives in a tightly focused manner as part of this, we have initiated a cost-cutting program with medium-term cost/income ratio target of below 45% by 2026. Thank you for your attention. I'm happy to take your questions now.

[Operator]

[Operator Instruction]. The first question comes from Borja Ramirez Segura from Citi.

[Borja Ramirez Segura, analyst]

I have two quick questions, please. The first is on the assumptions for real estate prices going forward. If I understood well, as per the Slide #13 of your presentation, you expect a decline in the office portfolio values of 8% to 12% in the real estate values. I think in the previous call, you mentioned a higher number than that, but just to, if I understand this correctly.

And then my second question would be on the increase in the nonperforming loans in the quarter, mainly in the U.S. if you could please provide some details if this is due to the covenant bridge of loan-to-value? Or is it a subjective NPL? And Also, how do you see the risk provisioning going forward on this NPL?

[Andreas Arndt, executive]

Thank you, Mr Ramirez. Now on the real estate prices, that's sort of easy to explain. When we talked about the changes in valuations last time, we were basically reflecting to the view on last year and first quarter. The way you have to read this on Page 13 is basically we did account for when you look at the way we set up the models and the way we also went through the portfolio. We did account for approximately 10% decrease in valuations for the office portfolio in second half '22. And we add to that taking the European office portfolio another 5% to 10% for '23 to come. So that all in all, basically gives since mid of year, combined devaluation effect throughout the portfolio which we also considered into our Stage 1 and 2 models of around -- well, depending on which figure you take between 15% to 20%. So that should answer your first question.

On the NPL. We look at the total exposure for the U.S. business of EUR 5.5 billion, of which presently 9 exposures are in default with the single loan loss provision Stage 3 of EUR 17 million in total. So that's where we are.

And of the 9 defaults, which we have, 3 are with an LLP, six are with 0 LLP. Now as I did describe to you, there might be various reasons. There might be short falls in cash flow, there might be other sorts of covenant breaches. The majority of cases where we have a default, but no LLP is the case where we have a valuation. Well, we have a valuation well in excess of what we need to cover the outstanding loan amount. And therefore, we may see covenant breaches, we may see unlikely to pay situations, but the value of collateral is sufficient and ample to cover the loan and the cost of unwinding the engagement and therefore, comes with this 0 LLP. I hope that sort of covers your points.

[Borja Ramirez Segura, analyst]

Yes. One quick follow-up, if I may. Regarding the provisioning calendar or provisioning backstop from the ECB. I would like to ask your views with regard to the NPLs with no coverage.

[Andreas Arndt, executive]

I'm not sure that I understood the first half of your question.

[Borja Ramirez Segura, analyst]

Sorry. What I meant to ask, so the provisioning calendar from the ECB So...

[Andreas Arndt, executive]

Ramirez, your line is very bad. I can't understand you really.

[Borja Ramirez Segura, analyst]

Sorry, I meant the provision in calendar for NPLs from the ECB?

[Andreas Arndt, executive]

The provisioning calendar?

[Borja Ramirez Segura, analyst]

Or backstop is also known as the provision in backstop?

[Andreas Arndt, executive]

The backstop issue. Okay.

[Borja Ramirez Segura, analyst]

So how do you have -- this is related to the NPLs with no provisioning.

[Andreas Arndt, executive]

No, sorry, the -- what comes into backstop goes against the capital not against provisioning and has always fully been accounted for as reflected in the capital. So we have no outstanding issues there.

[Operator]

The next question comes from Johannes Thormann, HSBC.

[Johannes Thormann, analyst]

Two questions from my side. First off, on your U.S. office NPLs. Could you give more details, what regional exposure is this all New York or other locations and also the one which was removed. What has driven this removal? Did you -- were you able to restructure it and so on? And last but not least, on the U.S. NPLs as you said, six do not need any provision because the collateral is sufficient. Which LTVs do we have to think about to get to such comfortable levels?

And the second thing is I'm struggling a bit with your own loss of being risk conservative, but then you contradict yourself basically by doing development loans in Stockholm, which is currently probably at least in my list, one of the three most problematic markets in Europe. So I'm struggling a bit with that philosophy.

[Andreas Arndt, executive]

Okay. I mean, on the last one, I can't and won't be too specific because we don't discuss clients on that side. All I can say is one of the best locations you can have in the city, and it is with a partner we know for long. And we do that against an LTV, which is way below 50%. So that's the rare opportunity, which we see there and not a reason for being concerned.

The first part of your question is on the U.S. office NPL. Now across all categories, including office, the vast majority of the exposure is on the East Coast. We have -- 75% is being done in New York and Boston , Washington. 12% of the exposure goes to Chicago. And the 12% is on the West Coast.

[Johannes Thormann, analyst]

Yes, I understood this. But on the 9 NPLs, is this the same or is it only -- where is it?

[Andreas Arndt, executive]

Yes and no. It's basically on all three regional categories. It's the same number of defaulted loans between each case. But as there's a higher degree of being affected by the defaults on the West Coast, obviously. And on Chicago, which show the three defaults each. But the 9 defaults, which I mentioned, you should also take note that three of them already being sort of delisted and put back into Q period due to agreements with its sponsor, with the investor partially by equity injections, partially by other farther reasons and 4 of them are put up for sale. And with an outcome which we can't really foresee but where we have no LLPs on as we go forward. Does that cover your questions ?

[Johannes Thormann, analyst]

Yes. Thank you. Oh, probably one follow-up on the net income from realizations. We had a whopping amount this quarter, as you said, because you're more able to sell noncore business or noncore portfolio. Is this an indication for the next quarters? Or was this a one-off amount?

[Andreas Arndt, executive]

Well, neither. No. It's not a recurrent item. We really can't say that, but it is something which is also not strictly one-off in the usual sense of the word. It is the result of the strategic decision to put the entire public sector on run down, which accounting-wise opens more possibilities to actively manage the asset side. And when opportunities arise, we will do that. And the same goes for bond buybacks. Again, that's not something which is sort of needed to be collected every month or every quarter, but is in the interest rate of environment, which we have and the spread environment, which we have where we see opportunities to crystallize gains, which are driven by higher interest or higher spreads. We will continue to do that. That's partially our way of sort of collecting the advantages of increasing interest on our part as others do it in different forms and ways.

[Operator]

Next question comes from Tobias Lukesch. Kepler Cheuvreux.

[Tobias Lukesch, analyst]

Few questions from my side as well, please.

Firstly, I would like to touch on this asset devaluation risk that you use for your model. As discussed before, there were actually much higher value issues with Q1 results of '20 to '23 basis go 35% in a net risk base. Now you're saying you have 15 to 20 in your model. So I was just wondering like how this feeds through the model and what you think of RWAs with that regard potentially and maybe also how you used the management overlay in this regard, as I understand there were some factors you anticipated, which for it was used. Then secondly, on the conservativeness of risk provisions, maybe you can highlight a bit again, I mean, how you differentiate yourself from some peers. I mean at the end of the day, you are focusing on very low LTVs, usually.

However, in the U.S., now we have seen a lot of movements, also some provisioning with the NPL changes, I was wondering how this works compared to the average Street? And lastly, on the costs, you highlighted basically in your presentation that you now target this kind of $40 million, 40% coming from personnel, 60% from non-personnel, 15% [ FP ] reduction. So what is new about that? Maybe you can give a bit more of some cost trajectory? How do you expect costs to evolve over the next years until '26. Very lastly, I mean, CAPVERIANT, maybe you can confirm roughly EUR 5 million impact on that EUR 40 million. So how is this waterfall basically working over the next years?

[Andreas Arndt, executive]

Okay. That's a full strength. Thank you, Lukesch. Now I'll start with the sort of conservative risk profile. And I want to exemplify that with one example. It makes a difference whether you have, for instance, a loan, or b loan or mezzanine structure. So we tend to be on the a loan, the senior structure, whereas others occasionally have engaged into higher LTVs, higher up the risk ladder and therefore, see further -- how shall I say, further impacted by risk provisioning these states. So it might be the same property. It might be the same risk class of that, but it makes a difference whether you covered the first 50% of the property or whether you sit behind the 50% with 30% to come. And that's a constellation, which we see not infrequently, especially in cases where we syndicate with some of our partner banks. That is perhaps one reason why you see high provisioning needs, in some cases, as regards to or as opposed to what we show.

Now on the valuation model, I think there might be some sort of misunderstanding we do not decrease the valuation parameters of the valuation model. It simply -- if some of the devaluation, which we expect has already taken place last year, we don't need to double up for that amount. So you have to see that as a composite figure going through. When we had, for instance, a property, which was valued at 100 on the first of July 2022 and was devalued by 10%. And there's another 10% coming on top of that, you basically go through to a 20% devaluation by now.

So that's to be seen in a compounded way. Now as your question to the management overlay, I think one of the driving forces was that interest were going up. And as that's being reflected in the risk cost, we could remove the management overlay to that extent, i.e., EUR 14 million, which was attributed to that specific factor. As the risk materializes in the model, the reason to provide for management overlay actually becomes obsolete and needs to be removed. And therefore, it sort of flattens out on that basis.

Now on costs. How do costs evolve? How do we see that coming? We said we'd like to go back to the cost income ratio of less than 45% and the cost level in absolute terms, which we've seen in 2022 which is a EUR 274 million for personnel cost and nonpersonnel costs.

What we see this year is a fairly significant amount of investments, which we take through the P&L straight away which basically, as a forecast figure for the entire year brings us to an amount of EUR 248 million to EUR 250 million in total cost.

Coming from strategic investments coming from the credit workspace -- workplace, which in terms of being put together now is in the full swing of investment. It is related to marketing expenses, which are needed in order to build out the retail deposit base and so on. Now that EUR 25 million increase, plus what we expect in terms of further investments and further wage drift up until '26, which makes up for another EUR 50 million is the EUR 40 million, which we expect to come down to in 2026 as the amount which we need to save in order to go back to the levels which I just have described. In that, there is a cost element to it, which you rightfully quoted on CAPVERIANT, which amounts to not EUR 5 million, that's too much on the high side. This is somewhat below that, but ballpark figure is about right. So does that give you a perspective on the cost side?

[Tobias Lukesch, analyst]

Yes. Maybe it would be really interesting. I mean it's below 235 million guidance out [Indiscernible]. Now this was not revised, so...

[Andreas Arndt, executive]

Lukesch, we can't understand you. It's a bad line, sorry. Lukesch, are you still there?

[Tobias Lukesch, analyst]

Better now? So you have this EUR 235 million cost guidance for below EUR 235 million cost guidance for '23. And I was just wondering how this evolves '24 '25. How much of cost to achieve you have when the benefit really comes through, that would be good if you could shed a bit more light basically on the path towards that EUR 224 million in '26?

[Andreas Arndt, executive]

Now the -- I'm not quite sure whether I fully understood what you mean, but the EUR 224 million is basically the target for 2026. Basically, what we say, were the measures with the investments which we have with the cost drift, which we have -- all that taken into account, would like to flatten that out up until '26, including further investments to come and including wage drift to come for the next 3 years to come back to the basis of EUR 224 million in 3 years' time in '26. So, basically, we do that in order to enable cost -- sorry, to enable the investments which we have on our list and to flatten out the cost increase, which we did see in '23.

[Tobias Lukesch, analyst]

Okay. So that means the whole thing is very much big and loaded.

[Andreas Arndt, executive]

No, the cost savings need to set in '24 already. But 2023 is the year of investments. And we can't sort of accomplish significant cost cuts on that side for '23. As a significant part comes from FTE reduction anyway. And we are just in the process of negotiating with the workers council and putting that into action is a matter of the third, fourth quarter, you won't see any impact from that side emerging in '23. You cannot. So it is a matter for '24, '25, '26 to come as we go forward. But also very clearly, it's not backloaded. We tried to accomplish a major portion of that already '24, '25.

[Operator]

[Operator Instruction]. There's one question coming from Philipp Häßler. Pareto Securities.

[Philipp Häßler, analyst]

Philipp Häßler from Pareto. I have two quick questions. Firstly, please, on the net income from realizations in Q2, the EUR 28 million, how much of this came from liability buybacks?

And then on your U.S. office portfolio, do we understand your strategy correctly that 100% of your portfolio is in A locations? Or is it too optimistic from my side?

[Andreas Arndt, executive]

No. On the first one, realization income, that's 100% -- sorry, the EUR 28 million for the second quarter compares against EUR 24 million, which we took out of the liability side. So it's about 80%, which comes from that. And on the office portfolio, and the A locations. I would say we did a very cautious and very careful selection of business over the last 3, 4 years in the United States. So in principle, I would answer with a yes. The problem is that sometimes you discover only 3 or 4 years later, that prime location, which you assumed to be a prime location did not turn out to be a prime location. That still may happen. But from today's perspective, I would say we had very strict and very clear selection parameters and Therefore, we are confident to say we have invested in prime locations and prime offices.

[Philipp Häßler, analyst]

Okay. Just wanted to follow up on your answer. So do you see the risk that, for example, in San Francisco or L.A., that former A locations become B locations. So this risk is clearly existent.

[Andreas Arndt, executive]

That is probably fair to say because of the overall political circumstances, which you see prevailing in Los Angeles and San Francisco and partially also in Seattle. The combination of how the inner cities are being populated just now and the fact that people make excessive use of home office makes it very difficult to decide what is going to happen to inner cities, especially in these 2 or 3 cases. But very clearly, if you would have asked me 1 or 2 years ago, whether the places where we invested in San Francisco or Los Angeles would be prime, I would say, prime prime. And we see slow by the movement also from the political side to do something about that. But it is very clear also that from a from a segmentation point of view, New York is very clearly positioned in terms of what is A and what is B location. And still will be.

[Operator]

There is one follow-up question coming from Borja Ramirez.

[Borja Ramirez Segura, analyst]

I have a quick follow-up, if I may. In Slide 16 on the office portfolio, if I understood well, it is mentioned that the loan-to-value is expected to increase. So it is currently 53% for the office portfolio. I would like to ask if you could please give some indications on the potential level of loan-to-value that you expect.

[Andreas Arndt, executive]

Now with the present business outlook, I mean, first of all, I tend to be cautious in making such predictions where we will land. And usually, that's not been part of our guidance, predicting average LTV figures. We have been seen being very sort of consistent and stable on the evolvement of average LTV in the respective asset classes. And due to that, I would speak.

[Operator]

As there are no further questions. At this point, I'd like to hand back to you, Mr. Arndt for some closing remarks.

[Andreas Arndt, executive]

No, I just have to say thank you for dialing in. Thank you for your interest and looking very much forward to also meet you in person at the next opportunity. Have a good day, and thank you for coming in and all the best to you. Thank you.