Ladies and gentlemen, welcome to the Q&A analyst conference call annual results 2019. I am Shire, the chorus call operator. I would like to remind you that all participants will be in listen-only mode, and the conference is being recorded. The presentation will be followed by a Q&A session. You can register for questions at any time by pressing star and one on your telephone. For operator assistance, please press star and zero. The conference must not be recorded for publication or broadcast. At this time, it's my pleasure to hand over to Mr. Richard Burden, Head of Investor Relations and Rating Agencies. Please go ahead, sir.
Good morning, good afternoon, everybody. Welcome to Zurich Insurance Group's full year 2019 Q&A call. On the call today is our Group CEO, Mario Greco, and our Group CFO, George Quinn. As usual, for the Q&A, we kindly ask you to keep to a maximum of two questions, and if we have time, we will come back to further questions later in the call. Before we start with the Q&A, Mario will make a few introductory remarks to the results. Mario, over to you.
Thank you, Richard. Good morning, good afternoon to all of you, and thank you for joining us. Before we get into the questions, let me just give you a few remarks from my side. As you know, in 2016, we set ambitious targets, and we launched a bold new strategy, and we have executed fully on them. BOP is up 16% in the past year, and the BOP at ROE of 14.2% is well above the target, as are the cost savings and the net cash remittances. In addition to the financial delivery, Zurich is now a simpler, more efficient business with stronger operations worldwide, with more engaged employees and higher levels of customer satisfaction. The performance of our property and casualty business has been particularly pleasing, with the business showing stronger growth in premiums, as well as an improved underwriting performance and reduced volatility.
The improvement in the accident year combined ratio before natural catastrophes shows that the actions that we have taken to change the mix of the business and improve the quality of the portfolio were the right ones. These have also positioned us well relative to the industry in terms of the current inflationary pressures. This is especially the case in our commercial business, where discipline and focus have driven significant improvement in profitability, in contrast to many of our peers. Looking forward, we see the pricing continuing to improve and exceeding loss cost inflation, which will support both further growth in premiums as well as further improvement in underwriting performance. Our life business continues to perform well, with further underlying growth, with headline results only held back by the strengthening of the U.S. dollar.
Against a backdrop of ongoing low yields, we remain well-positioned for further growth as a result of our decision to focus on protection and capital-efficient savings product already over a decade ago. Both the Zurich own Farmers businesses and the policyholder own Farmers Exchanges continue to grow successfully in the first half of the year. In particular, the Exchanges continue to successfully execute against the objectives aimed at enhancing the business as set out at the Investor Day in 2017. Our balance sheet remains very strong, providing us with significant flexibility to further develop our business while allowing us to continue to reward shareholders through a further increase in the group dividend to CHF 20 a share. Over the past three years, we have laid strong foundations.
Together with our customer-focused strategy and the further strengthening of both our product and distribution capabilities, this gives me great confidence in our ability to meet the even more ambitious targets for the next three years that we presented to you in November last year. Thank you very much for listening. Now George and I are ready to take your Q&A.
We will now begin the question and answer session. Anyone who wishes to ask a question may press star and one on their touch-tone telephone. You'll hear a tone to confirm that you've entered the queue. If you wish to remove yourself from the question queue, you may press star and two. Participants are requested to use only handset while asking a question. Anyone has a question may press star and one at this time. The first question comes from the line of Jon Hocking, Morgan Stanley. Please go ahead.
Hi there. Good afternoon, everybody. I've got two questions, please. Firstly, on workers' comp in the U.S. It seems that the sort of market commentary is that that line is seeing sort of softer pricing. Can you comment on what you're experiencing on that book and whether you're confident that the reserve release pattern we've seen in recent years can continue there? That's the first question. Secondly, on the life business, looking at the investment income, the reinvestment yields for the discrete second half look pretty low over 3%. I just wonder whether there's something distorting those numbers, or is that a new runway we should think about using going forward, and particularly given the dip in yields year to date? Thanks.
Jon, it's George.
On the workers' comp topic. Of all the lines of business, workers' comp in the U.S. is the only one that really exhibits any kind of weakness. I guess we'll come on to the others at some point later in the call. We still see it being 1%-2% reduction with an inflation, both experience and outlook, that continues to be very benign. I think I've said before on one of these calls that, actually at this stage, we would anticipate inflation may be a slightly negative tail around workers' comp. In terms of the fundamental trends that we've talked about on prior calls, I guess over the last 12-18 months, no change on the workers' comp topic. On reserve releases, this has been another strong year for workers' comp for us.
It's one of the drivers, but not the sole driver of the group's overall positive reserve development. I could not promise you could extrapolate all of these positives into the future. Clearly the more recent years, because of the pricing trends, are going to be a bit more competitive, and we need to see how the claims patterns develop. We've had strong releases both last year, last year being 2018 and 2019. We've reinvested. We've recycled most of that to strengthen reserves elsewhere. Our perception of our current workers' comp position at the end of 2019 is that the reserve position continues to be extremely strong. On the investment income topic, what happened? I guess Germany, the interest rate thing happened in general. It has distorted it a bit because you did see a bit of a bounce back towards the end of the year.
Given the book has a bit of a European bias to it, those kind of interest rate moves can have an effect, although maybe typically relate more to what happens in the policyholder side of things than they do for the shareholder.
Great. Thank you very much.
Next question comes from the line of Andrew Ritchie, Autonomous. Please go ahead.
Oh, hi there. First question, I think is the same question I asked at the half year. When I look at the footnotes in the financial statement, I can see a very low contribution on a pro forma basis from OnePath again, especially in the second half. Could you just clarify? I appreciate there might be some restructuring in there, whether some reserve adjustments in there. Maybe just update us on the status of OnePath, actions you've taken and what that means for future profitability. The second question, on the commercial business, the clean combined ratio ex -cat, ex -PYD, was running about 98.6%, I think, in the second half. It still improved year-on-year from the second half 2018, albeit about 80 bps.
Was there anything in the second half commercial, was it unusually high large losses, or did you take any opportunity to do maybe a bit more current year true up on some of the more pressured lines by way of sort of additional conservatism? Thanks.
On the first one, I think I'm going to give you a variant of the answer I gave you when you asked me that question at the half year. On OnePath, we took over control of it at the end of May. We're in the process of integrating. It will require more restructuring than I think we anticipated, certainly when we announced the transaction back in December of 2017. At that point, we'd anticipated some deterioration in DI. The market has seen more. That requires more activity by the team locally to bring it back to the profitability levels that we anticipated. There is a restructuring cost in it. There isn't any impact of reserve strengthening because the business is so new to the extent that we've seen any of that's in the opening balance sheets.
In terms of forward guidance, no change to what I told you before. We still expect to bring the business back to the path that we'd indicated at the time that we did the deal back in December of 2017. It requires us to take certain actions, it will be a bit second half loaded again in 2020, as I mentioned last year. The overall expectation that we have for it is the same. Even though it's a more challenging environment than we anticipated, there are Excuse me, we have a bit of noise in the room. There are some other environmental factors that I think are actually quite positive. You'll be aware what APRA has been doing, and the pressure that exerts around the whole DI topic.
In our view, anything that encourages the market to separate DI from lump sum and price each appropriately is a significant positive step. We think that actually the environment is conducive to the kind of changes that are required. Guidance on our expectation for OnePath for 2020 remains unchanged. On the commercial topic, so there is no real current year true-up. We didn't go back and do the kind of things that we used to do three or four years ago, which is to adjust the entire year in the last quarter of the year. The things that cause challenges in the second half are actually mainly property topics. The commercial business in more than one market has been impacted by property events. Of course, that can happen. I think commercial continues to make progress.
You commented on the fact that there is an improvement over the prior year. We expect it, given the price and loss cost outlook that we have, to continue to improve into 2020. I'm sure we're going to come back to that question very soon.
Sorry, George, I think you've always been cautious of talking about large losses now. Would you say the large loss is above a sort of normalized level, or you don't want to go that far?
No, I'm not sure I'd go that far. I definitely don't want to talk about large. We just have losses. In the second half of the year, we see more property coming through than we had. I think quite a bit more than we saw in the first half. That's the real driver of what you've seen from commercial. I guess the key point to reiterate is that we haven't done a current year true up on the initial picks.
Okay, thanks.
On the other side, Andrew, this is what is sustaining the further price increases in property. Property hasn't yet rebalanced, that's why prices continue to grow and keep moving up.
Cool. Great. Thank you.
Next question comes from the line of Peter Eliot, Kepler Cheuvreux. Please go ahead.
Thank you very much. The first one is on the Z-ECM ratio. I guess your framework still says that at between 120% and 140%, you should consider increased risk taking. You seem to have done the opposite this quarter. I was just wondering if you could sort of square that and say how we should think of your risk appetite from here at this level? Second one was on the crop business. When you think about 2019, should we just sort of put that down to bad luck and move on, or are there any sort of pricing or other implications for that business going forward?
Thanks, Peter. You're absolutely right about what the Z-ECM framework says, but it doesn't say that you have to do that. We have a choice. As you can imagine, in the environment that we've been in, if anything, we'd like to reduce some of the interest rate sensitivity that we have. That's why you see some of that reduced risk taking that you referred to earlier. The capital level is clearly very strong. Again, we obviously actively look at the portfolio that we have. We look at the risks that we currently run. We look at the trends and the expectations that we have for the future. We feel comfortable with where we are at the moment, just given the external environment.
It gives us the ability, not only to give additional assurance around dividends, but if opportunities arise, it also gives us that capital flexibility that can be a huge benefit. On the crop business, I think having had three really good years, it'd be a bit cheap of me saying it was bad luck last year, because at some point we actually have to pay people claims on this business. We've had combination of events last year. We had the prevented planting topic that we talked about at the first half, but it wasn't really in the first half results. It came in the second half. Very late in the year, we had the freeze on the sugar beet. I think one of the really interesting things for me was the efforts that Farmers made, especially around the sugar beet topic, to try and mitigate losses.
I don't think we see it as bad luck. It doesn't change our view of the line of business. We like it. We think that we can manage it well. We're happy to have it as part of the portfolio.
Thanks very much.
Next question comes from the line of Jonny Urwin, UBS. Please go ahead.
Hi, guys. Thanks. Just two, please. On P&C rate versus claims inflation, basically. Rates are up 4% across the P&C book, 9.6% in North America. Just wondering, can you give us an indication of where loss trend is running currently across the whole book and for North America? Secondly, what's your pricing versus loss trend expectation for 2020? I know you flagged positive margin draws, but I'm just trying to gauge the quantum. Thank you.
Thank you for the question. I think on Europe and the U.S. being the two major components, no real change around where we see Europe. Relatively small margin expansion in Europe. If you look at the full year, they're about 2.2%. It was stronger towards the end of the year. We're anticipating that it maintains that strength into 2020. We were more at just over three level at the very end of the year. Loss cost inflation probably running at something like about 2/3 of that level in Europe. The U.S., obviously the U.S. is where all the interest lies currently. From a price perspective, I'll come back to loss cost topics in a second. Generally, sequentially, each quarter was improved last year. Q4 is the strongest.
You've seen in the slides I raised today that we're close to 10% overall in the U.S. book. On the three major lines of business, property, liability, and motor, they're all in double digits. I think when you then look at trends, as soon as we think about trend coming into 2020, I think from a price perspective, we don't see this slowing down. Which is a change to where we were at the half year. I think we were a bit more cautious. I think today we expect this to continue through 2020. Loss cost trends across those different lines of business, our focus on liability and motor, because those are the two that are most heavily affected. On liability, you really need to look at primary distinct from excess. Excess is where most of the action seems to be. We all have loss cost trend picks.
Probably high single, just into double-digit level. Significant margin expansion, but obviously not as much as you would expect given the pure price topic. On motor, which again is in double digits, based on the studies that we've done, we expect to add about a point to the loss cost trend pick for 2020, taking us to somewhere between 5% and 6% overall. On both of these, the margin expansion is attractive. They come from slightly different places in terms of current profitability. They both offer an attractive opportunity. Having said that, I think we said at the Investor Day that we're not chasing share. Part of the reason for the loss cost trend choices that we're making is, I guess, to contain appetite around these topics.
I think we're quite happy to see the growth that we anticipate for 2020 driven by rate rather than exposure. That's broadly how we see it. I hope that's helpful.
Thank you very much. Very helpful.
Next question comes from the line of James Shuck from Citi. Please go ahead.
Hi. Good morning, afternoon, everybody. Two questions from me. On the EPS growth target, the capital market stage greater than 5% 2019 to 2022. You've delivered a strong set of results in these numbers, there was a very high level of capital gains in those numbers, around $ 845 million. I'm presuming that those capital gains will trend down over the next couple of years or so, closer to the $ 400 million level. Really my question is where is the rest of the growth coming from? You're fighting against lower investment income on the P&C side. I'm presuming it's all coming from the combined ratio improvement and a little bit on life, that's going to be low single digit. Perhaps you could just square that implied underlying growth that now looks a little bit stronger given the higher base delivered in 2019.
Second question, around capital position, the changes to the model that you've made and the reduction in the investment risk. Could you just clarify, is there more to come on that side of things? I can see that you haven't published updated Z-ECM sensitivities to credit. 100 basis points increase last reported was negative 17 points, which is obviously a very big number, and I appreciate you don't benefit from any of the buffers under the long-term guarantee package. Do you have a target level for reducing that sensitivity for credit, and can we expect further developments on the Z-ECM from management action? Thank you.
Thanks, James. As you started that question, I was trying to anticipate where you were going. You took it to a different place from the one I was expecting. On EPS growth, I'm going to answer the question I wanted to answer, and I'm going to answer yours at the same time. Just for everyone's benefit, the starting point, of course, is the as published number. As you point out, there is quite a high level of gains. There's also a reasonably high level of realized losses in there, which are mainly related to some of the disposals that we did last year. Either the transaction in Venezuela or the sale later in the year, which is yet to complete but will complete in the first half of the retail wealth platform in the U.K.
In terms of where does the growth come from, I think as we approached the Investor Day, we had a pretty clear sense of what we anticipated in terms of the outcome around net income for the year. That's already baked into the positions that you see there. If you look at that again, I know you're familiar with it, the main driver is going to be the P&C business. We do expect life to contribute. I think your number is right over the period. Probably life growth into 2020 will be slightly stronger than that low or that low mid-single digit type level. It will be a bit higher than that, I think, especially as the OnePath business comes on stream fully this year. Those are the two key drivers. There are other things. Farmers will continue to grow.
Of course, you're aware what the growth rate we anticipate there is. We've got some expense action that we'll undertake. All of these in combination, in roughly the same proportions that you saw at the Investor Day, are anticipated to be the driver of growth. It doesn't assume the same level of gains, just to be clear. On the capital model, it's a really good question. The challenge with it is that the Peter Giger and I, he's our Chief Risk Officer, we've had several conversations since the moves that we saw in Q3, Q4. On the one hand, there's just a pragmatic real-world perspective to this whole thing that in the asset risk continuum, while as you all know, we're not risk free, we're equally not the riskiest either.
For us to do substantial de-risking around fixed income involves taking a pretty significant bet in the other direction. I think we would rather be consistent around this area, even if it still brings volatility to the reported number. I think the approach that Peter and I have agreed is that we will look through some of the temporary volatility. We will look at what the market does elsewhere around UFR, but we want to maintain the model that we have because we think it gives a pretty clear picture of what's really going on. If you want to take the view that the interest rate risk can be paid down over a long period as the UFR tends to do, we can take an active decision to do that at that point in time.
Reducing risk taking around credit is not high on our list of priorities. Other than the comments we made earlier this year around capping credit exposure.
Just on the point around are there further benefits to the Z-ECM model that you would expect through 2020 from either changes to the risk allocation or indeed model changes?
At this point, there's nothing material planned. I know that PRC only has an agenda to take a look at the whole thing, and just make sure that it has all of the most modern thinking in it. I don't think that will cause material changes, either positive or negative. The model has been pretty stable over a longer period. I think you could assume that will continue through this year.
Yeah. Thank you very much.
Next question comes from the line of Michael Huttner, Berenberg Bank. Please go ahead.
Fantastic. Thank you so much. It's really one question. It's a bit complicated, congratulations on achieving all your targets. You're now a life company, if I look at the capital allocation, 54%. On that lovely slide at the beginning where you show the split of gross written premiums in life, most regions are overwhelmingly in your preferred segments except EMEA, which I guess is due to Germany. Also, in the risk report, you show a figure for the interest rate sensitivity jumping from $2 billion for the life segment to $3 billion for - 100 bps, so negative. These are big numbers. I just wondered what your thinking is here, whether you would be planning maybe to sell off your German Life unit, or what would the obstacles be to maybe running it off or transferring the portfolio. Thank you.
Thanks, Michael. I was going to say welcome back.
Thank you so much.
I think you correctly analyze the challenge that we face. The change in capital that you see, that is actually nothing to do with something that changed in the business. We didn't go out and suddenly become more risky. Of course, the financial markets moved quite a bit and the asset intensive elements of the portfolio, which are all in life, passively consume substantially more capital. I think we talked about it a bit before around how do we best manage this? I think if we saw ways that were in our interest and the interest of all of our partners and our clients to improve it, I think you could assume that we would do that. The German situation is a bit tricky. For reasons I think you understand and are aware of.
We do look to ways in which we can improve the capital consumption and the returns on capital that we achieve. In fact, I think I'd mentioned already, it may have been at the Investor Day, maybe it was at the half year, that if you look at the German business actually on a local capital basis, the funded basis, the returns are not so bad. Doesn't sound like a glowing recommendation, but the returns are not bad. The challenge is when we overlay that model that James referred to a second ago, and I know that you understand, that's where the challenge comes from. We continue to look at ways in which we can improve it.
In Germany, currently, we actually have, believe it or not, a few things that are probably higher priority, that are more about where the business is headed and trying to make sure that we can take advantage of the market opportunities. That continues to be a theme both for us and for the leaders in Germany to try to find a way to address this. I don't have a further update beyond that today.
Thank you so much. Again, congratulations. Amazing results.
Next question comes from the line of Farooq Hanif, Credit Suisse. Please go ahead.
Hi, everybody. Slightly simplistic question first. If pricing is accelerating and going to double digits in some of the areas that you are growing in, and there's this widening jaws or quite wide jaws on claims inflation, what is stopping your underlying loss ratios plummeting down? Is that the mix change? Can you talk about the dynamics around the mix change that you want to do and how much further to go? Secondly, on cash flow. I can see the explanation of why it's lower, primarily because of life, lack of capital release. I'm just thinking going forward, what you see as a growth rate in that life cash flow. For example, can we look directly at the OnePath earnings growth in life and directly assume that that will contribute to higher cash flow? Thank you.
Thanks, Farooq. On the first thing, it would be fabulous if the combined ratio would simply plummet. I think the thing to remember, it's my fault rather than yours. In the comments that I made earlier, you need to subdivide the portfolio. If we've got a near 15% rate increase in liability in Q4, that's only on one particular part of the portfolio. If the loss cost trend around the excess component of that particular book is say, 9%, 10%, again, it's on an even smaller part of the portfolio. We are seeing the, as you describe it, I guess the jaws open. I'm trying to work out if that's a good analogy or a bad one. We are seeing that positive margin develop, but it's in particular parts of the portfolio.
I think if you look at the U.S., last year, just the U.S. part of our business, if you're prepared to put crop to one side, you do see a very significant move in the loss ratio. Of course, that's partly driven by this trend. I think we are seeing it turn into improved performance. I don't think it's going to plummet, I guess is the point I'm trying to make.
I guess a comment and a question around that very quickly. Obviously, you're shifting mix as well. Not a question, really. How much further have you got to go on the sort of 52% in specialty and shorter tail lines?
If you think of how we manage mix from here, if you look at it on a written basis, I think the portfolio that we have, ideally, we'd want to have a bit more specialty in the portfolio in the long run. For reasons that we've talked about recently, doing that in the short run would be completely counterproductive. From a short-term perspective, I don't expect major shifts in the portfolio from a written perspective. Now obviously, it's earning through the liability changes it's been earning through for the last couple of years. From an earn perspective, this will be the first year in which we see the full impact of that. I think mix will have some effect on it, but I don't expect mix to have a significant negative offset to what we're seeing on rate and loss cost.
The benefits that we talked about earlier, allowing for some normal randomness around the claim incidents, we expect to see in 2020. On the cash flow topic and growth rates, you referred to last year, so I won't rehash 2019. Obviously, there are a number of things that will drive it as we come into this year. An absence of some of the interest rate volatility will not be unhelpful to the life business as we begin 2020. You point out the OPL impact, and just a reminder for people who may have forgotten what we said in December of 2017, we actually have a higher expectation for cash than earnings because, of course, there's quite a large in-force component to the portfolio. Life will come up a bit from where it was last year.
I've talked about in the past that we're expecting to see something north of $1 billion for life and OPL will have an impact there. The remainder will be driven by earnings growth. We do expect, as you remember from the Investor Day, and as I discussed earlier with James, to see 5% compound annual growth. That will feed into what should be a higher base starting point for 2020.
Okay. Thank you very much.
Next question comes from the line of Nick Holmes, Societe Generale. Please go ahead.
Oh, hi there. Thank you very much. Couple of questions. The first is, there's been quite a wide range of experience with social inflation among your U.S. peers, and I wondered what your take on that is, and how should we read across to Zurich? Then secondly, Z-ECM sensitivity to interest rates is still high, I see. I just wondered, are you doing anything to reduce that? Thank you.
Yeah, thanks, Nick. On the social inflation topic, that's partly reflected in the comments I made earlier, around this topic of the loss picks for next year. It's a bit hard for me to make any relative comments on Zurich versus the rest of the market. I think you guys know that we have done a number of things already. From a mitigation perspective, number one is the price change that we talked about earlier. We've obviously shifted the mix of business significantly. If you look at liability in our portfolio, I think we're down by something like 5 points, 6 points compared to where we were four years ago.
U.S. firm says to us, which is around excess, which is, it's not the only place that has the issue, but it seems to have a bit more of it maybe than some other areas. If you compare what's happening on limits to attachment points, and I'm looking at a graph as I speak, I can see the limits coming down year-on-year from 2015, and I can see the attachment points rising year-on-year from 2017. Obviously as that gap closes, it can't close entirely, otherwise there's no business to be done. Obviously that combination means that you're a bit less exposed to that topic. I think some of the things that we did earlier in the strategic cycle position us well for that.
We do see that social inflation issue, I think maybe not quite as much as some others for the combination of reasons I've just given you.
On the interest rate topic, and I think I referred to it briefly earlier, I think in response to James's question. At the very long end, which is not entirely but mainly a German topic, there are some issuers out there who are now issuing some very long-dated bonds. We were discussing it earlier this week. ALM is the focus for us, but when we're buying bonds that are going to mature in the year 2120, I think we need to be a wee bit careful about just getting too carried away with the yield topic. We are looking for ways to reduce interest rate sensitivity. Again, there'll be a limit to how far we can go, just because our model has none of those features that soften the impact.
I think we would like to bring it down, but try and do it in a relatively pragmatic way.
Great. Thank you. That's very interesting. Can I just come back very briefly on social inflation? Without sounding too arrogant, you kind of think that you anticipated this trend perhaps earlier than some others.
Yeah.
You're pretty comfortable with the position.
I got my boss sitting next to me shaking his head. I would love to tell you that we knew this was going to happen, and we did all of this with this particular scenario in mind, but that would be a lie. We were trying to fix issues that we knew we faced. The positive side effect of that, by doing that hard work at that time, means that we're in a reasonable place today. I can't claim that we had foreseen this particular eventuality.
Great. No, that makes sense.
I think we told all of you in 2016 that we will do this, we ended up, by following the strategic choice, we ended up in the right part of the market. It's not because we saw it's just because we wanted to change the nature of our books. The timing was lucky.
Makes sense. Thank you very much.
As a reminder, if you wish to register for a question, please press star and one on your telephone. The next question is from the line of Vinit Malhotra, Mediobanca. Please go ahead.
Yes, good afternoon. Thank you very much. My question, I have one question only, please, on the retail and other segment, where obviously crop has confused the picture a bit on the underlying combined ratios. Somewhere in the notes or the comments, there is a statement that even excluding crop, the underlying would be better. Could you help us understand that a bit more? Could you provide some more color on how much or what's happening between mid-market or other retail segments or something to help us more? Thank you very much.
Thanks, Vinit. If you put it in context, I'll pick a retail other. We are 93.6% for the entire year, which is, I would give it my highest accolade of not bad. Obviously, the second half is a bit weaker than the first, so we are 94.6% in the second half. crop makes up a reasonably significant slug of the premium in the second half of the year. In retail, we've got about $1.5 billion of earned premium for the year for crop, and the bulk of that appears in the second half. Crop has an adverse impact on the group overall of 0.6%. retail is a bit more than half the book, but to keep the math simple, I'm going to assume it's half. You can double it for retail, and you can double it again for the second half of the year.
I think that's why you get this comment that if you adjust that 94.6% down for something like 4x the crop impact, you still see an improvement.
All right. Thank you, George.
Next question comes from the line of Niccolo Dalla Palma, Exane BNP Paribas. Please go ahead.
Just a couple of last questions for me. On the central cost, you still guide to $ 750 million-$ 800 million. You did much better in 2019. What would explain the deterioration from there? I think you pointed out the lower headquarter cost. The second question is on the reinsurance protections that you have today. Nothing changed in the excess protection. Is there any significant change worth flagging on the quota share? Thank you.
Great. Thanks, Niccolo, for the first question. You're absolutely right. We saw a further significant reduction in the central costs last year. We haven't yet been able to pass all of that on to the businesses for various reasons. It is my intention to do so in 2020, which is why we've guided you back up to that slightly higher level of $ 750 million or maybe slightly higher. We'll pass on the benefits, but with a wee bit of a lag. That's why you see that combination of topics. On the reinsurance protection, you're absolutely right. We haven't changed the attachment points on the catastrophe aggregate. Not really looking to make any significant changes across the programs. We've had a number of renewals on January 1.
The larger cat program actually renewed last year. There's a multi-year contract, so it won't come up for renewal this year. You won't see, obviously for that reason, any significant change there. Not much to add on reinsurance, actually.
Thanks.
The last question is a follow-up from James Shuck from Citi. Please go ahead.
Well, thanks. A couple of follow-ups, please. Just on Farmers. GWP growth is around 2%, but the policy in force keep declining, we're down 2% continuing just since H1. The Net Promoter Scores keep going up, though, Mario, and I know that you've been a big proponent of the link between Net Promoter Scores and increased retention. It doesn't seem to be working at Farmers, at least in terms of the policies in force and also given the rollout to East Coast. Could you just comment a little bit about what isn't working at Farmers, at least on the P&C side? Secondly, on the expense base. The other underwriting expenditure ratio improved to 13.5 points at full year. I think you've intimated that the goal is to get to better than 13 points in time.
I'm not really sure what you mean by in time, what kind of timeframe you've got for that, and are you expecting the absolute level of expenses, $9.2 billion on the controllable cost base, are you expecting those to decline in absolute terms, please?
Should I do the second one first, then we come back to Farmers?
Yeah.
On the expense base topic, I guess it probably depends on whether you ask me and the CEO or you ask someone else. At Investor Day, what we indicated, a bit of both is what we're looking for. We still believe we have pockets of inefficiency in the group, and together with the support of the Chief Operating Officer, we're going after that currently. You will see reduction there. We're also anticipating that because we expect to see continued growth in retail, and for the reasons I gave earlier around the commercial book, actually stronger growth in commercial this year, which would be rate driven. We expect that also to contribute to expense efficiency. Where will that leave us overall? I think you might see big expenses at roughly the same level, maybe slightly reduced compared to prior periods.
It won't be what you've seen in the course of the last three years. Maybe we're looking for something in the kind of the $400 million range rather than the $1.5 billion that you saw before. In timeline over the three year period. It's not that things outside the three year period don't matter, but we're not going to talk about them until we start doing them. This expense commitment is for this three year period.
Yeah. On Farmers, two things have been happening, and they were absolutely planned, and in a sense, expected. One is that we're restructuring the agency force. I think we have been showing the characteristic of the new agents that we're hiring, but we're also closing or merging a number of old agencies. When you do that, there is of course an attrition. The reason we're doing this, it is by having bigger agencies, more structured, we can much better use the power of the data and the capacity to sell products through the forces of these agencies. The second fact is that, not only for the reason of the wildfires, but following the catastrophes, prices have been still growing. When you grow prices, it's difficult to go and gain more individual customers.
That's the law of the market that not even retention or a Net Promoter Score growing can overcome. If you increase the prices, then you will acquire few new customers or no new customers, and the game becomes the one of maintaining your existing customers. We think that especially in the second half of the year, the numbers will turn into positive customer growth, and we're looking forward to see that happening.
I think that concludes Sorry, James, carry on. Do you have a follow up?
That's all right. Listen, thank you. That's fine. Thanks.
Okay. Thank you, James.
Okay. Well, thank you very much, everybody, for dialing in today. If you do obviously have further questions, then the IR team is available for your calls and questions. With that, we'll close the call. Thank you and goodbye.
Ladies and gentlemen, this concludes today's Q&A session. Thank you for participating, and wish you a pleasant rest of the day.