RELX PLC (RELX) 2017 Q4 法說會逐字稿

So good morning, everyone, and welcome to RELX's 2017 results presentation. Thank you very much for coming in. And for those of you on our webcast, thank you for dialing in.

You will have seen we've had a pretty busy few months, so there is quite a lot today to talk through.

Firstly, and most important, I'm pleased to be able to report that we've had another good year with trading continuing to be on track across all of our 4 businesses. Overall, we had good growth in underlying revenue and adjusted operating profit. In sterling, adjusted earnings per share was up 12%, and in euros, up 5%. And we're proposing a dividend increase of 10% in sterling and 6% in euros.

Secondly, we announced today that we're proposing a set of measures that will further simplify our corporate structure into a single parent company, RELX PLC, after 25 years as a dual-listed corporation. We believe that our dual listing no longer serves any purpose, and that this is a natural next step for RELX Group, removing unnecessary complexity and increasing transparency. For the avoidance of doubt, there will be no changes to the locations of RELX Group or its 4 business areas, and there'll be no impact on the economic interests of any shareholders in PLC or NV. And there will be no change in our corporate strategy.

Thirdly, you will, I'm sure, also all have noticed that since the year-end, we've announced the intention to acquire ThreatMetrix for GBP 580 million. I believe this acquisition is a great fit with our very successful fast-growing risk business, and we're hoping to complete that acquisition next week.

In the meantime, we also announced the GBP 700 million buyback program for 2018, and we continue to refresh our boards. In September, we welcomed Suzanne Wood as a nonexecutive director. Suzanne is a U.S. citizen, Group Finance Director of Ashtead Group and brings extensive board experience from both sides of the Atlantic. She is proving to be a strong addition to our boards.

So we've got lots happening, and I will now hand over to Erik and Nick, who will take you through the results presentation and the proposed simplification of our corporate structure.

Well, thank you, Anthony. Good morning, everybody. Thank you for coming and for taking the time to be here today. As you've probably seen from our press release this morning, our positive financial performance continued throughout 2017 with underlying revenue growth across all 4 business areas and with underlying profit growth ahead of underlying revenue growth. We continued to make good progress on our #1 strategic priority, the organic development of increasingly sophisticated analytics and decision tools. And this morning, we also announced the further simplification of our corporate structure.

In 2017, our underlying revenue growth was 4%, underlying adjusted operating profit growth was 6% and earnings per share growth at constant currencies was 7%.

Our overall financial performance was consistent with recent trends across our 4 key financial metrics: underlying revenue growth, underlying adjusted operating profit growth, adjusted earnings per share and return on invested capital. All 4 business areas again delivered underlying revenue growth as well as underlying operating profit growth. So let's look at the results for each business area.

Our STM business grew 2%, in line with prior year, with key business trends remaining positive. In STM primary research, which represents just over half of the division's revenues when you include both subscription and transactional revenues, we continued to enhance customer value by building out broader content sets, more sophisticated analytics and evolving our technology platforms. So far in 2018, subscription renewals are in line with recent years. The percentage of renewals that have been completed as of today is the same as the average percentage by the middle of February over the past 5 years, and the built-in growth rates in the renewed contracts are also at the same level. New sets of research subscriptions have started the year well ahead of mid-February 2017.

In total, we now have well over 5,000 institutional customers globally that have contracts for a broad collection of subscription journals, a number that continued to increase by a few hundred institutions during 2017, in line with historical trends.

The number of articles that we published under a fee-for-article open access model continued to grow double digits, and our market share under this model has now doubled over the past 4 years.

Databases and tools, which represent just over 20% of Elsevier revenues, continued to drive growth in the mid to high single digits across market segments. Scopists added a few hundred new institutional customers last year and is now also approaching 5,000 total customers.

Electronic reference, which represents just under 10%, had a mixed year. Print books, which now represent around 10% of Elsevier's revenues, saw sales declines and return rates back at historical levels in 2017.

And print pharma promotion, which represent less than 5%, declined in line with historical trends after stronger 2016.

Going forward, our customer environment remains largely unchanged. Overall, we expect another year of modest underlying revenue growth, with underlying profit growth continuing to exceed underlying revenue growth.

Risk & Business Analytics grew 8% and revenue growth remains strong across all key segments.

In Insurance, which represents about 40% of the division's revenues, growth was driven by enhanced analytics, extension of data sets and by further expansion in adjacent verticals. The market environment returned to historical trends in the fourth quarter after having been not quite as favorable earlier in the year. So far in 2018, the Insurance market environment has remained in line with historical trends.

In Business Services, which represents around 30% of the division's revenues, growth was driven by further development of analytics across the financial and corporate sectors in a positive market environment. So far in 2018, the market environment for Business Services is in line with historical trends, and we have recently announced the acquisition of ThreatMetrix, a leader in the digital identity space. When combined with our own strengths in physical identity attributes, this acquisition will enable us to offer a unique and comprehensive approach to fraud and identity risk management.

While this acquisition is larger than average for us, it's a good illustration of our approach to acquisitions. We are acquiring a company that we already know through partnership, and we have already learned how the integration of our respective products adds value to our customers.

The government and health care segment, which represents less than 10% of division, continued to drive strong growth by developing more sophisticated analytics.

And Data Services, which represents just over 20% of divisional revenues, continues to drive strong growth across all industries through organic development.

During 2017, we disposed of additional magazines and services assets, reducing print to less than 3% of the division's revenues. Going forward, the fundamental growth drivers of Risk & Business Analytics remain strong. We continue to expect underlying profit growth to broadly match underlying revenue growth.

Legal revenue growth of 2% was in line with the prior year. Underlying profit growth was strong. The margin increase reflects organic process improvements in decommissioning on systems, largely offset by portfolio effects and final exit from a joint venture. Overall, continued growth in online revenues was partially offset by further print declines.

U.S. and European markets remain stable, while other international markets continued to grow well. The rollout of new platform releases continued across our U.S. and international markets with broader data sets and a continued expansion of early-stage legal analytics. In early 2018, the migration of U.S. Legal customers' usage to Lexis Advance is now substantially complete.

Going forward, trends in our major customer markets are unchanged, continuing to limit the scope for underlying revenue growth. We expect underlying profit growth to remain strong.

Exhibitions' underlying revenue growth accelerated to 6%, up from 5% in the prior year. The slight acceleration in revenue growth reflects our continued pursuit of organic growth opportunities, including launching 36 new events. We also completed 5 small acquisitions and piloted several data analytics opportunities.

Europe, which represents around 40% of revenues, continued to see good revenue growth. The U.S., which represents around 20%, continued to see differentiated growth rates by industry sector. Japan and China grew strongly. And while Brazil remained weak, most other markets continued to grow strongly.

Going forward, we expect underlying revenue growth trends to continue, and we expect cycling-in effects to increase the reported revenue growth rate by around 4 to 5 percentage points this year.

Our strategic direction is unchanged. It is still to be a company that delivers improved outcomes to professional and business customers across industries, with our #1 priority being the organic development of increasingly sophisticated information-based analytics and decision tools that drive higher value for our customers.

In addition to our organic development-driven growth strategy, we continue to reshape our portfolio through selective acquisitions and selective disposals. 2017 was an unusually light acquisition year, with only 8 transactions completed for GBP 123 million. But in January, we announced the agreement to acquire ThreatMetrix for GBP 580 million. 2017 was a more active disposal year with 17 transactions completed.

We announced this morning that we're proposing a set of measures that will further simplify our corporate structure, moving from the current dual parent holding company structure to a single parent.

Our corporate structure has evolved over time, and this simplification follows the significant measures that were complete in 2015. We believe that this is the natural next step for RELX, removing complexity that no longer serves any purpose and increasing transparency. There'll be no change to the locations of activities or staffing levels of RELX Group or its 4 business areas as a result of these measures. There is no change to our strategy, and we expect the simplification to be cost and profit neutral for RELX Group, both before and after tax.

And I will now hand over to Nicholas, our CFO, who will talk you through our results and our corporate structure simplification in more detail. I'll be back afterwards for a quick wrap-up and our usual Q&A.

Thank you, Erik. Good morning, everyone. Let me start by expanding on the financial highlights. As Erik said, underlying revenue growth of 4% was maintained in 2017. Growth in underlying adjusted operating profit was ahead of revenue growth at 6%, resulting in a margin improvement of 40 basis points.

Growth in earnings per share at constant currencies was 7%, ahead of profit growth, reflecting a lower interest charge, a slightly lower effective tax rate and the impact of the share buyback program, all partly offset by a drag from disposals.

Return on invested capital improved to 13.1%. Cash conversion was again strong at 96%. Leverage was 2.2x EBITDA adjusted for pension and leases, in line with the prior year. The equalized full year dividends, up 10% for the PLC in sterling and up 6% for the NV in euros. 2017 share buybacks totaled GBP 700 million.

Turning to the income statement. 4% underlying revenue growth gave us 2% constant currency growth, the result of disposals and acquisitions are cycling out, more than offsetting the contribution from acquisitions. Sterling was weaker on average in 2017 compared with 2016, the effect of which was to benefit sterling reported revenue growth by 5%, giving the total growth of 7%. Adjusted operating profit increased by 6% on an underlying basis. After portfolio effects, constant currency profit growth was 3%. The sterling figure was then boosted by currency again by 5%, resulting an 8% overall growth to get total operating profit of GBP 2.3 billion and delivering the improved margin of 31.1%.

The interest charge reduced, a result of lower average interest rates. The average rate paid in gross debt was 3.2%, 0.6 percentage points lower than the prior year.

The tax rate on adjusted profit declined slightly to 22.5%, resulting in adjusted net profit just over GBP 1.6 billion, up 10%.

Reported net profit was up significantly, which is under GBP 1.7 billion, a result of an exceptional noncash tax credit of GBP 346 million, arising from U.S. tax reforms.

Going forward, we assess the impact of tax change in U.S. and other relevant jurisdictions to be a small net positive for RELX Group, although not significant to the adjusted effective tax rate or to cash taxes paid.

The average share count was down by 2% due to the share buyback, converting 5% constant currency net profit growth to 7% constant currency earnings per share growth. The year-on-year weakness of the pound bended adjusted earnings per share, and sterling was up 12% to 81p. Conversely, growth of the euro strength resulted in a slightly lower euro denominated EPS growth of 5% to EUR 0.923. Reported earnings per share were inflated by the exceptional tax credit I just talked about.

In terms of the business areas, you can see how all 4 areas contributed to the overall revenue performance, with the Risk & Business Analytics and Exhibitions continuing to deliver strong underlying growth. Disposals were a drag on overall growth, in particular for Risk, where we continue to divest assets to do the favorable strategy. Cycling decreased Exhibitions' revenue growth by 6 percentage points.

With the pound weaker on average against both the dollar and the euro, currency movements boosted sterling-reported revenues in all 4 business areas by around 5%.

The growth in underlying adjusted operating profit was driven by strong performances from Risks, up 8%, and from Legal, up 11%. Legal profit growth was significantly ahead of revenue growth, reflecting efficiencies within the business and the decommissioning of legacy infrastructure.

STM wholesale had underlying profit growth above revenue growth.

Exhibitions profit growth reflected the cycling out of the [pioneer] events.

With more disposal activity in 2017, portfolio effects were a drag on profit growth, most notably in Legal. In sterling, currency movements were positive for all 4 business areas having 5% to 6% of operating profit in each case.

The group continued to make progress in improving margins, adding 40 basis points to reach 31.1% overall. STM's margin was flat on a reported basis with underlying improvements offset by portfolio effects. Risk margins are up 60 basis points, in that case, helped by portfolio changes.

Despite a significant drag from disposals, Legal improved margins by 40 basis points, a result of underlying profit growth being well ahead of underlying revenue growth. Exhibitions margins were flat in a cycling-out year.

I've highlighted the different currency effects as we've gone go through the results. Our principal currencies are mainly U.S. dollar and the euro, with hedging delaying and smoothing the impact of exchange movements. FX rates will, of course, move from today's levels, but sterling is currently stronger against the dollar by around 7% and broadly unchanged against the euro compared to its average level in 2017. If current rates were sustained, the currency impact on adjusted EPS growth relative to constant currencies will be negative by a mid-single-digit percentage in both sterling and euros.

Most of our invested capital is in U.S. dollars, and capital employed the average FX rates increased accordingly with the currency impact being similar to the impact on profits.

Underlying profit growth did drive higher returns, but that was partly offset by portfolio changes, giving a net 10 basis point improvement in return on invested capital to 13.1%.

Turning to cash flow. Capex was GBP 354 million, equivalent to 5% of revenue. Capital conversion was again strong at 96%. Cash interest was down slightly while cash taxes increased, driven by increased profitability and exchange rate movements. Total free cash flow was GBP 1.5 billion.

Here's how we used that free cash flow. Spending on acquisitions was GBP 123 million, dividend payment is GBP 762 million and the share buyback was GBP 700 million.

From year-end to year-end, currency translation effects were broadly neutral, leaving net debt almost unchanged at GBP 4.7 billion. Leverage, which we calculate in U.S. dollar was 1.9x or 2.2x when you adjust for pensions and leases. The adjusted ratio was in line with the prior year despite the prior year having been flattened by currency movements.

You're familiar with how we think about priorities with the use of cash, and they haven't changed. And you know that the acquisition spend is one of the variables in there. Spend was low last year, as Erik mentioned, GBP 123 million, but we will have the ThreatMetrix deal this year. So the averages hardly move much if you look over a longer period.

So continuing with the aim to maintain leverage in or around the 2.1 to 2.5x range that we've been in, we continue to have capacity for share buybacks. We have announced a further GBP 700 million of buybacks for 2018, with GBP 100 million of this having already been deployed.

Next, I want to briefly highlight the impact of changes in accounting status that will apply for 2018. There are 3 of these standards we're already adopting this year, albeit the effect of one of them, IFRS 9, will be trivial. The 2 that will have an impact for IFRS 15, which all companies will be adopting this year; and IFRS 16, which we are adopting a year earlier than mandated.

IFRS 15 relates to revenue recognition. Whilst there's some cumulative balance sheet effect, there's only a small impact on the full year income statement.

IFRS 16 requires all leases every 12 months to be capitalized on the balance sheet. Our leases are almost all property leases. And based on 2017 year-end numbers, we'll be adding GBP 325 million of net lease liabilities into our net debt. Of course, our headline net debt-to-EBITDA ratio of 2.2x already takes account of operating leases in the calculation, so there is no impact on our quoted leverage.

Taking all the changes together, there is no impact on revenue or profit growth rates. Full year operating profit is unchanged, but there is a small impact on interest, which flows through to adjusted earnings per share. In the appendices, you'll find tables setting out expected restated H1 and full year 2017 numbers, which will become the comparators when we start reporting on 2018 numbers in July.

Finally, I'd like to give you a little more detail on the corporate structure simplification Erik talked about earlier. The simplification will be implemented through a cross-border merger between the 2 parent companies, RELX PLC and RELX NV, with RELX PLC being the continuing entity whereas NV shareholders will receive 1 new RELX PLC share in exchange for each RELX NV share that they hold. Slide shows the expected time line for the simplification. From announcement today, we expect the merger documentation to be sent to shareholders during Q2. The shareholder meeting is to vote on the simplification towards the end of that quarter. The simplification is subject to certain conditions, including approval by both RELX PLC shareholders with a 75% majority and RELX NV shareholders with a 50% majority. Following shareholder approval, U.K. court approval will be sought in Q3, with the merger becoming effective around September.

The simplification will not create any changes on RELX PLC shareholders and there will be continuity for our NV shareholders. The table sets out the current position for PLC and NV shares and how that will change following the simplification.

Existing RELX PLC shareholders will continue to see their shares listed in London with dividends paid in sterling. The ADRs will continue to be listed in New York. Existing RELX NV shareholders will see their NV shares replaced by PLC shares, but those PLC shares will be listed in Amsterdam with a default position seeing dividends being paid in euros. Of course, all shareholders will have the ability to switch their PLC shares between markets, trading in London or Amsterdam. RELX PLC shares will continue to be included in the FTSE 100 Index with a share count that will roughly double. We expect RELX PLC to replace RELX NV in the AEX Amsterdam Index and RELX PLC shares to continue to be included in the STOXX Europe 600 index and other relevant pan-European indices.

With that, I will hand you back to Erik.

Thank you, Nick. So just to summarize what we have covered this morning. During 2017, our positive financial performance continued, and we made further strategic and operational progress. Going forward, key business trends in the early part of 2018 are consistent with 2017, and we're confident that by continuing to execute on our strategy, we will deliver another year of underlying growth in revenue and adjusted operating profit, together with growth in adjusted earnings per share on a constant currency basis.

And with that, I think we're ready to go to questions.

Okay. Maybe today we'll start over here. Start over there.

This is Nick Dempsey from Barclays. Three questions, please. So first one, let's get straight into Germany. Can you talk about specifically how you plan to account for revenues from German institutions in the project deal consortium this year? So in other words, will you make assumptions through the year about what you might receive once the deal is done? Or will you assume no revenues from those guys all year until you get a deal? So how are you going to account for it? Second question, on STM on margins, you're talking about underlying profit growth being ahead of underlying revenue growth. If we assume the FX continues where it is, would you see any margin improvement in that division on a reported basis? And lastly, at Legal, with the drag from selling Martindale-Hubbell Associate, probably be less in 2018, could we expect more margin improvement year-on-year from Legal in '18 than '17?

Okay. Well, I think that I am going to ask Nick actually to address all of those 3. Are you ready?

I thought you might. So revenue recognition, just to be clear, and we won't talk about any individual customer across, but just to be clear how it works, apart from some December, January cost of judgment that we've always had to make, all revenue were recognized in STM in the full year comes from signed contracts with customers. Just to be absolutely clear on that. There is some judgment we have to make during the year. For example, if a contract is commercially agreed but not formally signed during the year, then you have to exercise some judgment as you go through half year, in third quarter and so on. But for the full year, apart from those kind of issues, all the revenue recognized is from signed customer contracts. We will, of course, in 2018 be adopting IFRS 15, which is even more clear cut. And even during the year, you only recognize -- in IFRS 15, you only recognize revenue from signed contracts. I hope that's clear. The second question is on STM margins and the FX impact. As ever, there are lots of moving parts with currencies. You will have seen in this analysis in the appendices that shows the currency impact in 2017. Actually, there's net nil. There was no currency either from the hedging year-over-year or in the FX rates overall. Clearly, the hedging -- the way the pound has moved and the euro has moved, the hedging has been a little bit of a drag the last couple of years that eases off we roll through those contracts. But again, we said, of course, the dollar is now getting a little weaker again as you all have seen. So there are lots of moving parts, and I -- as ever, I will not give a forecast as to what the net FX impact is going to be when we get through the year. The third question was on Legal margins and drag from Martindale-Hubbell. You're absolutely right. Of course, the Martindale-Hubbell joint venture has now come to an end. There was some contribution in '17, they're relatively modest mostly in the first half, which will, of course, go away. So that will still be there in the year-on-year drag. If you look at the numbers for Legal though in 2017, you'll see it's not the only drag from disposal, we have disposal another -- we mentioned it in the release, some other print assets and some service businesses that's come out of there. So there will still be a bit of a drag from disposals in 2018 in Legal.

Okay. Let's continue over here.

Katherine Tait from Goldman Sachs. A couple of questions for me. Just coming back on the STM margins, I think you noted that the underlying business actually saw margin improvement, which was then offset by the sort of changes to the portfolio. Can you talk a bit about going forward, do you expect continued portfolio changes that could depress margins further or certainly hold back some of the operational improvements that you're seeing? And how we should sort of think about that going forward? Secondly, within insurance in Risk & Business Analytics. You talk about growth in some of the adjacencies around sort of autos. Can you talk about how much those various areas make up now within your insurance business and also how you're seeing the penetration of those versus where the sort of opportunity set is? And then finally on ThreatMetrix, if you could just give a little bit of color in terms of the -- you've talked about making continued investments in that business and therefore unlikely to sort of contribute to earnings accretion this year. Can you just give a bit more color about what those investments are likely to be?

Okay. Well, I'm going to ask Nick to take the first one and then I'll come back and talk about the others.

So when I see our margins going forward, clearly it depends what comes into the portfolio or may get out of the portfolio. It is typically true. Of course, if we are buying fast-growing businesses that perhaps on all of that mature, they would tend to be a drag on the margin. And then on the disposal side, it will depend what we sell. Sometimes we do sell relatively high margin, but low growth already declining bits of the portfolio. So it will depend.

Okay. So insurance adjacencies. Well, the rule of thumb we've done before for Insurance, broadly speaking, is -- and these are not exact numbers, but broadly, we say that in the -- if you look at our overall Insurance business, roughly 40% of Risk, we talk about that as saying that a portion of that then, of course, it's international, which we've talked about. And then in the U.S., where order of magnitude sort of 20%, 25% of what we're doing is not what you would consider traditional auto revenues, and that's the scale of it roughly. And it has been growing marginally faster than the core of the other business, of course, because they're from a lower base. So on an overall basis, they're contributing well to the business and to the growth, and we see them continuing to grow. Not much ahead of the rest, but perhaps a little bit, right. So at this point, then we don't know what will happen in each one of those market segments going forward. So I'm again not going to make a forecast for future growth rates, but you look at -- we don't think we're anywhere near the saturation point in those, let's put it that way. So I don't see that this would -- this is something that's limited in the short term. ThreatMetrix, the way we look at this is this is a very well-established company in this segment, but we see significant opportunities in this segment going forward because we have so much history ourselves in physical biological identities of individuals, and now we are combining this with their sort of leadership in digital and device identities. And we think and we've seen that, that adds significant value to customer sets that we already serve. And we're going to continue to expand that value by continuing to broaden out the product suite and do integrated services and to build out our sales efforts. So that's why we think in the first year of ownership, we don't think this is going to have a material impact on the overall profit picture, if you look at it that way, that's the way we think about it in the first year. Want to keep going?

This is Ian Whittaker from Liberum. Two questions, please. First of all, just if you look at the sort of overall group sort of underlying revenue growth, it has sort of accelerated over the years up to 4% and so forth. Just wondering in terms of your views sort of what gets it to accelerate further after that 5% and beyond, are there any particular areas that you think will drive this? Would it be sort of just a weighting of the revenues that you get more from Risk and then we'll have a natural sort of calculation effects? Does Legal gets a little bit better or so on? So your thought might be quite interesting. And then second of all, I think, just in terms of ThreatMetrix, you mentioned the point that in terms the acquisition type, it fits into the profile, it's a business you already know, you already worked with and so forth. Maybe just for our benefit, could you sort of highlight sort of any areas the business where you particular sort of have sort of relationships in terms of sort of other partners sort of is there any sort of division, which is particularly sort of you've got similar sort of relationships so that they have with ThreatMetrix or sort of just generally across the whole group?

Okay. Okay, yes. The overall growth of the company, the way we think of it overall is that at any given point in time our growth rate is going to be impacted by, to some extent, what it is we do, right. And that's how we add value to our customers, and that's where we spend all our time focusing on. On the other hand, any one -- any growth rate in any one market in any one time period is also going to depend on the environment we're selling into, general economic environment, the customer, the health of the customer segments we're selling into and so on, and we don't try to give specific forecast for the second part. We spend all of our time focusing on how it is we add value to our own customer in our segments. And we believe that the way we will drive further growth in each one of our segments is identical, which is that we are spending most of our time internally focusing on how can we take existing content sets, additional data sets, how can we integrate those and leverage those with analytical tools on top, predictive analytics and decision tools so that we shift further and further from information lookup to integrated decision-making tools where we can track, measure ideally and increase the value we provide to the customer at that decision point. And the closer we get to those decision point and the more that we can quantify the increased value from that decision, the higher value we're providing to them. And if you focus on the increased value to the customer, it shouldn't be a surprise that we then assume that they will want more of that service and use more of it and somehow we will capture a portion of that value. So that's how we think about growth driving, growth acceleration across all our customer sets. Then, of course, we still have a little bit here and there or what I call sort of traditional drag. That's going smaller and smaller, whether that is fueled printings here or books and magazines here and there, but they're becoming smaller and smaller, and therefore that's not our main strategic focus. Our main strategic focus is add customer value and drive accelerated growth that way. The second question was on ThreatMetrix. The way -- you asked where do we see the combination having real value?

Sorry, it wasn't so much that. It was just -- so when you talked about the acquisition, you said you'd acquire -- sorry, you said you'd acquire it because it sort of fits to the profile in terms of its business you already knew. So what we're saying just for our sort of our benefit, I just wondered if there's any sort of other partnership sort of similar that you have in the business to what you have ThreatMetrix to just identify for us.

I'm sorry, I misunderstood. I thought you meant what ThreatMetrix do. No, no, we have small partnerships around the business everywhere, and in particular in the risk business, we have lots of licenses and partnership agreements with different data sets, different providers, so content or sometimes distribution partners. We have that, of course, primarily in the U.S., which is where these businesses originated, but also in different parts of the world. So we constantly get to know lots of these companies that have data sets or distribution or an integration opportunity. And we, as you know, do very few acquisitions on (inaudible). It's really when we see that this really add significant -- significantly more value when you can offer them in combination. And for ThreatMetrix, that's the case across financial services industries, many different subsegments, of course, also in different places where the identity of an individual is important and you can link it to the device or the digital identity in terms of online transactions also or accounts, online transaction accounts opening and so on in a nonfinancial (inaudible), okay? Okay, let's keep going. Yes?

It's Adam Berlin from UBS. Three questions if I could. In the last, say, 12 months, you have signed quite a lot of deals on general subscriptions. Can you just talk broadly about what kinds of concessions on open access you've been asked to make if any and what you've agreed to? Second question is on STM, how much of the revenue in STM is now from author payments or gold open access? And thirdly, you mentioned in RBA that some of the market conditions have improved. Can you talk about which metrics you are referring to and how we should think about that?

Sorry, let me answer the last one. Say it again.

You mentioned in your speech that in RBA, you've seen some improvement in some of the underlying market conditions in insurance. Can you just talk about what type of metrics you mean by that?

Maybe I'll address that first because you -- when you say improvement in Insurance, we say that this year, the beginning of this year, is the same as what we saw at the end of late in the fourth quarter last year, which is that the market environment has returned to our historical average metrics when we look at market attractiveness. And that's a combination of the things that we see in the marketplace through our systems and through other indicators that relate to insurance industry, price changes, marketing behavior, shopping behavior, right, and switching behavior. And what we see out there in the general base market as a whole, right, and these are indicators where you can track that in combination, adapt to basically use the market. So we can track a bit the market cycles. It is not something used to predict market cycles, but we can actually track and see what is happening as it is happening. So that's what we're saying that it goes -- it fluctuates a little bit over time. The beginning of last year was slightly, slightly less attractive than it had been the prior couple of years. And then by the fourth quarter, it returned to what we consider historical average. And we're starting off the year so far, 1.5 months, at what looks like historical averages, right? Now like you said, the first one you said about our contracts. As -- just to put this into scaling, as I said there, we have over 5,000 institutional customers, right, that have broad collection subscriptions with us.

About every year, because some of those are 1-year and many of them are multiyear, that means that every year, we probably renew -- renegotiate, I should say, renegotiate about half of those each year. So we're talking about well over 2,000 individual renegotiations every year. And in those, in all of those, you asked me about the concessions in a way. We don't see open access as a concession at all. We see it as we are a service provider. We are here to help our customers in their pursuit of advancing research, advancing science that sometimes include health care. We are here to help them. And we provide the services to them that they can choose to get from us or anybody else. We provide those services to them at, as you know, above-average market quality and below-average market pricing. And price trends over time, ours have been below-average market price trends. That's how we think of ourselves, right, a service provider that offers a full range of services whether that is just institutional that nobody publishes, services for preprint, whether that is a subscription-based request or whether that is an open access article, sort of fee per article-based request. And we're happy to provide those services to all our customers on a market basis. And we're happy to do it with our philosophy about price and quality. When we go into a contract renewal, when we get into a negotiation, we always do everything we can to make sure that we sit down with all these institutions every year and make sure we understand what it is they are trying to achieve in terms of research, in terms of science, what information they need, what tools they want to use. And we try to work with them to figure out a way that we can help them reach their objectives in a way that they ultimately find economically attractive solution for them. That's what we work on and that's how we approach our business. And we look that way for all along. So we did not see any of these questions about is one bad, one good? We do not try to influence on what's good or what's bad. And we see them as -- we see ourselves as a service provider. We're happy to act for all of these services. So the second question you asked was what percent of our total article share is roughly is gold open access as you referred to, the sort of fee per article. We are now approaching, let's call it, we are above mid-single digits but not yet at double-digit percent of our article share internally. And that's what you asked, yes, okay. Next, okay, so let's go over here.

I'm Sami at Exane. Historically, you reported the revenue growth for the journals at being between 3% and 4%. Would you comment on whether 2017 was in the same range and how you see the outlook for that subsegment of Elsevier, please? Secondly, historically, you used to comment on the international risk business. Can you update us on how big the international risk business is and the growth rate you've had there? And lastly, you've made a few disposals. Can you share with us the revenues that you've disposed of annually, annual revenues and what the revenue trends were for the asset's disposal?

Okay. I'm going to ask Nick to cover the last one, but I'll first cover the other ones. Yes, I mean, as you know, we have not given specific growth rates for any one of our subsets in numeric terms. We give indications of roughly how our segments are doing. And I think just like I did today, I gave sort of ranges and so on. But I think maybe it would be helpful for you to, when you talk about the journals business, to help all of you just how to think about the scaling of that and the segments. So if you look at it this way, of Elsevier's total revenues, we have shown you that just over half is the primary research business. When we say primary research, we include all content sets, all primary research, everything you would have called historically the journal business, right, but it's primary research. Of that, a few percentage point is transactional. They're transactional, it's backfile sales, it's one-off licenses and other things that have nothing to do with what you think of as a subscription business or the pace -- so gold open access fees and other types of related individuals. So once you take those out, you're down to almost exactly half the revenues, sometimes a little more, a little less. But we're talking about now 50% of Elsevier is what you would think of as subscription revenues for journals or for primary research. Of course, that 50% is a global average. And that means that in the countries that are fast-growing countries or very large Asian countries, where we don't have a lot of historical local presence, that number is significantly higher than 50%. So many Asian countries, many developing markets, it's significantly higher than 50% is the subscription base for large broad research collections. Of course, the other side of that is then, therefore, in the U.S. and some other large developed Western economies, that number is, therefore, below 50% on the subscription -- subscription business as a part of our total Elsevier business. And then of course, we often here end up talking about the academic customers. And I think you also have to understand that corporate customers are also large subscribers on a global basis. Of the total subscription business, maybe 10% is corporate customer subscriptions. And of course, there are significant geographic differences. In major Western industrialized mature countries, where you have corporate headquarters, this is often twice that, closer to 20% of the total subscription base. So that might help you scale how to think about that 50% of the company, of Elsevier that's subscription-based. And also of course, don't forget in there, we also have individuals that subscribe, individual scientists, health professionals, often related to some health and biomedical content and some society-related. So that's also a portion of it. So when we talk about the institutional academic subscriptions, that might help you frame it, right? So when we talk about journal growth, you add up the whole thing. That's what we're talking about. And every year, there's some ups and downs. But as I told you before, if you look over the last 5 years, every year is little bit of ups and downs and ins and outs. There's always somewhere in the world, some countries that slow down a bit and some that pick back up. And every year, that goes a little in and out. But if you look at it on average, we're continuing to run in the same general trend as we have over the last 5 years. And don't forget also, and I've said, we are talking about renewals and renegotiations each year and we're also talking about new sales each year, right? New subscriptions, that's what we do in all our business across all our segments, right? So the broad range, you have to look at over a few years. And as you can see, Elsevier's overall growth rate is similar for the last 5, 6 years. And the trends in all our major subsegments are broadly the same over that time period, right? We have similar trends in books, so even though there are fluctuations in books. But we're not going to give a specific percentage growth of any one of the subsegment for any 1-, 6- or 12-month period.

You reported an acceleration in new sales, basically. What's driving that?

Well, as I said there, new sales come and go in different years at different times. And I'm just giving you across all our major segments that I did in Insurance, that I did in Business Services and Legal and so on, where are we as of today in 2018 because I've heard that some of you expressed an interest in where we are right now at the beginning of this year. And as I said in our major individual markets, Legal is very much similar to us as we said before. Insurance is in line with historical trends, Business Services, in line with historical trends. And research, renewals in line with historical trends, new sales higher than last year, right? So that's where we are at this time, okay? International risk business, yes, as I said last year that we were heading to a year where it was roughly $100 million, the total of all the business. This is what we define as the international expansion from the old U.S.-based business. Of course, we have lots of other things around the world that originate from acquisitions after that or on the old RBI side. So that's not what we're including here. This is the expansion from the U.S. risk business, now about $100 million. And it continues to grow as we hope to. Then disposals, I'm going to let Nick cover.

So Sami, I mean, you can tell from the fact that the underlying growth was 4% and the constant currencies got to 2%. There's a gap of 2%, which obviously there's some cycling in there. There's acquisitions in there. But most of that is the disposals. So about a 2% drag on the overall group revenue, which GBP 150 million is coming from order of magnitude is coming from disposals. Obviously, some of that is a part year effect because it has some of them in the part of the year. So that could give you a sense of the scale of it. The trend, the businesses (inaudible), of course, (inaudible) have been things like print magazines and about the New Scientist Elsevier magazine, for example. And typically, those are declining modestly. So that depends on which particular segment you're looking at. But generally, they are going down but at a modest rate.

Okay. And over here.

It's Patrick Wellington from Morgan Stanley. A couple of questions loosely associated with STM. If I buy a sofa from DFS, I believe I'd get interest-free credit for about a year. If I get journals in Germany as part of the deal consortium, I seem to get interest-free credit for more than a year. At what point does supplying a customer with a product for nothing begin to annoy other customers who are paying for that product? Secondly, 180 is a number familiar to us in Germany. I think you said 5,000 institutional customers of Elsevier. So 180 divided by 5,000 is 3.6%. Is that an accurate reflection of Germany's significance in your Elsevier business? And then thirdly, do you have any views on ResearchGate and your ability to restrict the availability of your copyrighted information on ResearchGate?

Let's see here. So you asked the first question of STM. Well, as you know, I'm not going to talk about any one individual customer. But I can help you describe our approach and how we work this in general, which is that we are here, as I said before, to serve our customers. We are a service provider. When we go through these renegotiations that take place, we spend all our time and effort interacting directly with the individual institutions and trying to understand what their objectives are. And we help them reach those in a way that is that they are happy with in terms of substance and that they see as economically attractive. We do that every single year and we want to continue to do that. We have been around for well over 100 years. And we are very, very patient and very customer-focused company. And these things will take the time it takes when we go through those. We are here to serve our customers and not to pride sort of an argument. When you look at the proportion that you say in general, you might want to -- you might also be aware of a very different number, which is that around the world, over the last now approaching 20 years, institutions have formed large consortia, right? And we have around the world now over 170 large consortia, most of which are national or near national, right, very large established consortia. Most of those were launched 15 years ago or 10 to 15 years ago. And every time you start to get a consortia together, it is a slightly more complicated process than it is to just renew an existing one of those 170, right? So of course, because they're multiyears agreement, this means that probably just under half of those we are fully renegotiating every single year. Now some countries or most countries formed these a very long time ago. And as you know, not going to get into in detail, but there are these formations are still at the tail end. And when they take place, they are more complicated discussions. And the question you asked about ratios, I think you can, from my previous description of the size of our overall business and the size of the academic -- the size of subscription part and the corporate part and the geographic differences, combined with the chart we have shown you, that Europe in total for STM, including the U.K., is about 25% of the revenue. And we've also said in the past, U.K. is probably around 5%. So if you take the 40 countries that we count in Continental Europe, represent that 20%. And you heard what we said before about the scaling of subscription business and corporate and other things. So I think you can figure out the relative importance of any one customer whether it's this time or any other time without me commenting on your exact calculation. The last one, ResearchGate or any of the other sites that are -- that have been out there or will continue to pop up there or anywhere else. As you know, I'm not going to comment again on any one individual. But we are in the intellectual property business. Every company in an intellectual property business will have to deal with IP protection, copyright, illegal activities or illegal postings or violation of intellectual property rights. Like all other industries, we have industry associations in each one of our market segments that deal with these and are experts on these. And typically, what happens when something comes up is that through industry associations or affiliated groups, they contact any one of these sites and they inform them of what the rights are and what you're allowed to do. This also happens, of course, in consumer industries whether that's movies or music or anything. And many industries go through a period of adjustment to how to work with sort of within the law. And that's the process that we expect to continue to go through in every one of our segments all the time is primarily through industry association and industry groups and it's with intellectual property rights. And it's a relatively standard process. And sometimes it goes very quickly and sometimes it takes a long time until you end up to some place that's functional and that respects the law. So next?

Sorry, just a quick follow-up, more positively, if you've got a customer who hasn't paid you for something for 18 months and then they decide to pay you, do you write off the 18 months? Or do you get a lump sum for the 18 months (inaudible)?

Again, I'm not going to talk about any one individual customer. As I said to you, we work directly with every one of these institutions that renew. And we're talking, depending on how you look at it, well over 5,000 institutions or close to 200 large, complex consortia. And we want to work with them to get them -- to help them reach their objective in a way that they think is economically attractive. And we've done this for many years, and we're going to continue to work it that way. And we do it directly with them and we're very patient. And we are a service provider.

It's Ruchi Malaiya from Bank of America. On the risk division, which bit are you personally sort of most excited about there? Is it the opportunities to expand that core Insurance business more globally? And is it sort of the identity fraud product expansion, so particularly in light of the threat

(technical difficulty)

I'm personally excited about both of those and a few others because I see significant opportunities for us to take our assets which relate to content sets as well as analytical algorithms, mathematical tools as well as technology platform capabilities to combine the information with the (inaudible) with the technology tools to drive significant value. And I see significant opportunity to do that over time. And you asked about Insurance globally. And I see significant opportunities on the others that are risk side. But we also know that these are complicated businesses and individual locations are different, which means that they are not very quick buildups. It takes a long time to work your way into -- if you look at the Insurance side, if you work yourself into a local environment to build up the datasets, to build the customer trust. We can transfer the analytical tools, our industry knowledge and our technology platform relatively easily. But then to apply that to local datasets and demonstrate value to a customer, so they can see and quantify and then build on that, that's a different -- that takes a bit longer. You could argue that as you said, some of the other ones that are related to ThreatMetrix, that there, that you could argue that it's easier to establish a local presence. But there are lots of local regulations, lots of local restrictions on what you can do that you have to work your way through one location at a time. But we think there's significant opportunity in these -- in the U.S., where they started, they are still left. So we're not dependent on a certain pace of international expansion even though we're excited by it.

It's Steve Liechti from Investec. Erik, you talked about the large consortia. I think you referenced over 170, then you said 200. So just to double check, which is the...

Yes, I said under 200. It's in the 170.

Okay. And in the last 10 to 15 years, of those, let's call it, 170 consortia that were presumably put together, how many of those failed to reach agreement with you? So was there sort of 200 and now some of them never reached agreement with you?

I don't want to try to guess if there is one that I'm missing. But let's put it this way, the number of consortia have continued to grow over time in a steady stream. The difficult situations that we've gone through with them are long and drawn-out. And normally, when they're combining, when they have -- when they're working through their role and how they can add value and how we can add value, and we are fully supportive of all of them using all the tools available to them to get the best deal possible in their mind. We are a very large provider. We publish probably 17% of the world's articles, more like 25% of the world's citations. And when you do that at above-average quality and below-average price and below-average price trajectory, it shouldn't be a surprise that the demand for our services keeps growing. And therefore, our number of submissions keeps growing and our rejection rate increases every year. And we're trying to manage our volume growth and quality combination within those parameters. But if you were any customer of that, it would be really, really important for you to make sure that you get the right deal from a company like ours. And we have full respect, full understanding for that. And if some of them feel that taking a bit longer and being a little bit more vocal about it helps them, we have full respect for that. I respect and I understand it and we support it. But we are going to work with them directly and not in any other form. So have these things formed over time? They continue to form. The numbers have continued to grow. Once in a while in some country somewhere, they split up again because they didn't see the need to be together, right? That's also a question of do they have similar objectives? Are they trying to do the same thing rather than are we offering them something?

I know I'm being simplistic here. But sort of to ask the question a different way, has any one not reached agreement with you ever in the last 15 to 20 years, apart from people who speak German?

I'm sorry? Well, ultimately, we always end up with some form of customer relationship. Every institution in the world that conducts research is a customer of some part of our content sets. Because we have such a significant range of services, 2,500 individual journals, right, most research institutions in the world want to be involved in that somehow. I mean, as you might know, we have over 10 million unique monthly research users in the world. We publish -- we received last year around 1.6 million article submissions to the company. We published quite a bit over 400,000 articles last year. We operate a network with 20,000 -- over 20,000 editors, over 80,000 editorial board members. 1 million reviewers are involved every year. And every year, we have these more than 10 million users in every month downloading last year well over 900 million full-page articles. We are talking about in that all these numbers keep growing. So clearly, there is something that we're doing that everybody thinks supports science. And our objective is to continue to do that. But we also want for our customers to get value from it and feel that they're doing it in a way that suits their objectives and their economic structure.

Okay, great. Can I have one more? Just on Lexis Advance, you said that's rolled out now completely across the U.S. And you always said about the decommissioning and what that has the effect on margin. Is there any sort of like one-off boost once you get it to 100%? Or shall I think of it as part of the sort of gradual effect?

It's part of a journey because U.S. Legal customers are now effectively completed their usage migration, when we first moved the customers on and then we moved the usage up, right? And then effectively, all the usage has now rolled over. But U.S. Legal customers are only one customer set that uses and shares the overall platform. So what it means is there are certain portions of the platform we can now turn off. But we're not talking about one platform as either on or off. When we talk about the platform, we're talking about a few hundred individual components, content sets, databases, software tools. And at this point, we've already turned off several dozens of those. And yes, there will be a few more this year. But it's not one big step. And we don't foresee one big step in the near future either.

It's Matthew Walker from Crédit Suisse. The first question is would you accept the argument that it has -- in these negotiations, it has become easier for people to resist because they can use Sci-Hub, because they can use ResearchGate or because they can get interlibrary loans? Has it become easier for them to hold out? Would you accept that? The second question is what is the percentage of readership on the journal side to articles that publish within the last 12 months, i.e., the new material that people will in theory not have access to, versus the old data which they presumably keep on having access to? And then last question is you seem actually pretty confident that STM can sustain the growth rates that we've seen in '16 and '17. So where are you finding the additional revenue to allow you to do that? Is it coming from China? Is it coming from other parts of Asia? What's actually happening to bolster the revenues when you do have this hiatus with certain customers?

Well, let's see here. First question, you said, yes, "Is it easier, interlibrary loans and so on?" Interlibrary loans as tool has been available for decades, right, and it's still there. And interlibrary loans work. But one-off, and this is what most customers understand when they work this through in detail, that to work that way with limited subscription, interlibrary loans and so on, it takes a lot longer and it's actually more expensive. What most of our customers have discovered, and why is it that we now have well over 5,000 institutional customers to our broad collections, is that over time, over the last 20 years now since ScienceDirect was introduced, you have seen customers go from subscribing to a portion of our 2,500 titles, a portion of those titles that they thought were targeting their specific areas of research focus. And then when we move to a broad-based technology platform to supply those, they then continued to supply -- to subscribe to those selected journals. Then when we offered additional viewing rights only, which has a different interesting history, that you subscribe to certain content sets and you pay and you own those, but then we started to offer very long time ago, before my time here, the broad collections, where you say in addition to that, you can pay a limited amount of money you get access to all the other content on an annual basis, right, to view all the other content. And those are the broad collections that everybody has been talking about over the years for a very long time. And those are normally the ones that are up for the big renegotiations. Most customers that moved into those collections over time tried it for a year or 2 or 3 years. And typically, the most typical number is that they discover that between 40% and 50% of the usage of the full content set came from the additional broad collection that was not in the original subscription list as a typical conversion result in the first couple of years. And that means that you have significantly broader access to information sets. You have significantly broader ways to link and access everything. It's very fast, it's very easy. You know that no matter what topic you're researching today and putting together something for tomorrow, you know that you have 100% of the world's research accessible. And we have, as you know, a linking hub, which means that you can link from one of ours to somebody's article, right, so you can go in and around. So therefore, most of the customers have that started it have then continued to stay with it and expanded it. And that's why we're now well over 5,000 of those, having started, of course, with 0, literally 0 on the broad collections when they were introduced. Every year, a few people add to this collection around the world, a few institutions. But if they don't see the need for all this because they're very special (inaudible), they might very well then move out of that broad collection. So overall, we added a few hundred customers last year. But every year, there are a few moving in, a few moving out, a few are testing and so on. But the net, it has been a very steady, significant increase over a very long period of time. And if you look at cost per effective article access or cost per article download, I mean, we are down now on a global basis, where I think the cost for customers is basically -- I mean, it's below $2 an article, right? So if you look at that level, anything that you can come up with that is more specialized and more focused or different is not usually that attractive for most research institutions in the world. And we are not trying to force anybody to take it. It's a choice. You can choose whatever you like. We are a service provider. We'll offer our services in any range or any subset that you would like as a customer. But it's what's most appropriate for you in terms of usage and help you reach your objective and what's most economically attractive to you.

On your confidence on the...

Sorry, yes. So yes. The way I think I have to look at it is that we have all these different subsegments that I talked about before, right? I mean, we talked about 50% of the business being subscriptions, 30% of the business, database and tools and electronic reference, 10% print, less than 5% of print format. And we have lots of subsegments in that, right, that vary a little a bit every year. But overall, we are pursuing improvements in our product sets, improvement in our customer value in all of those segments at the same time. And there are always some subvariations, also subvariation by customer group, right, by customer group as well as by product sets. There are regional variations and there are sort of timing-related variations. And that's why when you look at this, you have to understand the broad product portfolio that we sell and the broad geographic portfolio we have and the broad customer sets. And I think I helped you before to scale the different portions of it. So we have seen over the last few years that there's always a little bit of variation any 1-, 6- or 12-month period. But overall, as you've seen, the Elsevier growth rate has been very stable for the last 5 years. Okay? Okay, we have one over here.

It's Chris Collett from Deutsche. A couple of quick questions. One was just on Legal, you made a comment a couple of times now about the early stage legal analytics. Now obviously, adoption of analytics is somewhat slow. But just wondering, at what point does that become part of the customer workflow so that we can actually see an acceleration in the growth rate coming from Judge Analytics? Second was on STM but not about the consortium. It was really just about some of the acquisitions that you've been making over the past few years, some of the sort of preprint tools, the platforms, the workflow tools and so on. What is the overall strategy there? Are you going to run them as separate businesses? Are you looking at integrating them in with primary research? And then lastly, just on the collapse of the DLC structure, I just had a question about the -- you're going to continue to be listed in Amsterdam. But is that a -- is that going to be a depository receipt listing or some other format?

Okay, I'm going to let Nick answer that one. But I'm first going to cover the others. You said Legal. Yes, we have spent a significant amount of time and effort in building out our analytics offering over the last few years. You are aware of some of the acquisitions we made, which we talked a little bit about in our investor seminar on Legal now, that's probably bit over a year ago. And we are -- we think this is a very significant, increasing value add to Legal decision-making over time, right, that to involve analytics and decision tools on top of the information, we think, has significant value over time. However, at the beginning, it is small. It's growing very rapidly. So I wouldn't say it's slow. But I would say it's small and growing rapidly, which means that, by our standards and impact to the whole business, it's not material to what we're doing today. But I do think it's a material part of the value proposition to Legal customers over time. And I think it's a very important piece. And I think we are extremely well positioned to do this for the legal industry based on how we do this across our several market segments and the capability we have in predictive analytics related to the risk division, which we also apply in the same way to the science and STM side. And that goes to the second question you said, which is, "What are we doing with all these other tools?" Well, that is exactly in the same direction. Yes, this company started historically as a publisher in many of our segments. We do not view ourselves today in any of our segments as only a publisher. We are a publisher. We publish content. We take ownership of that content, we stand by the content. We do retractions, we do corrections and so on. So we take our role as a publisher of content very seriously. But we believe that if we're going to add more value to our customers across all these market segments all the time, we are moving significantly from providing information and content to actually helping our customers make better decisions, get better results from those decisions and to be more productive while doing it by offering a suite of alternative tools that you can use standalone. So here's your answer to your precise question, that you can use standalone and you can only go in and use one of them at a time, if you would like, that's your decision. But also if you're a user of several of them, we plan to have them in almost all our market segments integrated with each other and linked, not forcing them together but linking them. So you can move more seamlessly between them and get better results while being more productive because you're eliminating the complexity of stepping in and out, right? And therefore, you can follow the decision-making process, not just reading an article, which is one thing, but then taking the content in that article, working with that article, linking it to other similar articles or linking it to datasets in others on our platform, which you wouldn't do if you had a printed-out article. You're actually doing research inside the content sets on our platform. That is very different from the traditional reading a text on a page, right? And it adds significantly more value, we believe. And over time, when you get -- when you integrate those tools and you provide them to different types of research center here or in different types of legal work, it's a much higher value-add solution set for our customers. And that's what we're going with the acquisitions as both standalone and as integrated solutions. Okay?

Two things. The question about the share listing. So one major depository receipt is direct listing of the shares in Amsterdam. London will be the primary listing, but there will be an additional listing in Amsterdam. Unlike today, of course, it will be entirely fungible there. You can move the shares between the 2 markets and trade in either one with the same share.

Okay. We have one more in the back.

It's Richard Eary from UBS. Just two questions. Just the first one, STM. I mean, Erik, you gave some sort of breakdowns in terms of primary research percentage. I think looking back to the last time you gave some breakdown in 2015, that number was 59%. I'm just trying to make sure that those numbers are comparable between the 50% and 59%? And the second question is just on the RBA side, I think in the notes, you talked about an expansion in terms of RBA margins due to some portfolio shifts. Can you just elaborate on those portfolio shifts and whether those are sustainable into sort of '18, '19, particularly with the acquisition of ThreatMetrix?

Just to make sure I understood the second one, could you...

So in the notes for the results, there was a slight improvement in the RBA margins. And I think the comments around that was a change in the portfolio mix. So I'm just trying to understand what was the change. Is that sustainable?

Yes, okay. No, I got it. And Nick, you got that question, I'll let you explain that one. Yes, okay. Yes, so I -- we have shown you so many charts of our revenue breakdown over time for all our different divisions. And I can't remember exactly which year we had the one that you refer to as you saw it looks like 59%. We have shown a few of those. It wouldn't surprise me if it was something very close to that number in the past. I think if you look at it today, and we put some charts in the back that have the revenue breakdown by division, that have the revenue breakdown by area of different types of revenues, and I think that it might possibly be -- if I do that today, if I did that today, I think we might be few percentage points lower than that, right? That's the way we look at where we take the total research -- primary research business, which you would refer to as the journal business with all its different subcomponents. So I know what the number is today, right, and it's probably a few points below that but not many, not many points, just a little bit. You can see in the chart in the back, so it's very close. So I'm not sure why it's exactly why it has changed a couple of points. So whatever it might be over that time period, it might be that on the margin, there was some marginally different definitions or it's currency, currency valuation differences between the countries that are higher or lower dollar-based versus because we take it spot at that time. And as we have some very significant portions of our revenue in different part of the world and we have different proportions, the proportion of the research is very different by geography. That's why there are some natural built-in fluctuations if I do a static pie at spot rates. So these will always vary a little bit a few points, right?

So the margin question, Richard, in '17, the portfolio pluses we had were coming from because we're selling lower-margin businesses and the print magazines, for example, were coming out of RBA, which gave us the boost to margins. If you look forwards, then, I mean, you're absolutely right, acquisitions, particularly even though the sort of businesses we tend to buy quite tend to be quite immature and would typically not be at the average margin that risk makes today in the high 30s, so they would tend to be a drag on margin. And ThreatMetrix would be in that category. But if you look at the 2018, you've obviously got the full year effect of the things we did last year and anything else we do on the disposal front. So I think you'll have to wait and see exactly how it pans out.

Okay. Thank you all for listening. I look forward to seeing you all again soon.