© 2024 Alpha Spread Limited. All Rights Reserved.
www.alphaspread.com is not operated by a broker, a dealer, or a registered investment adviser. Under no circumstances does any information posted on www.alphaspread.com represent a recommendation to buy or sell a security. The information on this site, and in its related newsletters, is not intended to be, nor does it constitute, investment advice or recommendations. The information on this site is in no way guaranteed for completeness, accuracy or in any other way. In no event shall Alpha Spread Limited be liable to any member, guest or third party for any damages of any kind arising out of the use of any content or other material published or available on www.alphaspread.com, or relating to the use of, or inability to use, www.alphaspread.com or any content, including, without limitation, any investment losses, lost profits, lost opportunity, special, incidental, indirect, consequential or punitive damages. Past performance is a poor indicator of future performance.
You don't have any saved screeners yet
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Good morning, and thank you for joining us today for QBE's 2020 full year result presentation. I'll start by briefly discussing the key features of the result, give you a feel for the current pricing environment and the implications for GWP growth and margin, then provide a quick update on COVID-19 before handing over to Inder to talk you through the details of the financials. I will then close with our 2021 priorities before opening up for Q&A.Firstly, I should open by stating that the Board and management of QBE recognize that last year's result was extremely disappointing and well below our expectations. I should also say that today's results, both at a headline and underlying level, are very much in line with, and hopefully in some instances, ahead of the expectations we established with our December announcement and our subsequent business interruption update in mid-January. There was obviously plenty of detail to work through but there should be no negative surprises in today's release.I'd now like to turn to Slide 3, the performance update. Given that we have preannounced much of the result headlines, I won't dwell on the performance update or result summary slides for too long. Our headline combined operating ratio of 104.2% was obviously heavily impacted by $655 million of COVID-19-related costs as previously advised. Excluding COVID-19, the combined operating ratio was 98.6%, up slightly from 97.5% in 2019, with a further material improvement in the attritional claims ratio and a modest improvement in the large individual risk claims ratio, more than offset by elevated catastrophe claims and adverse prior accident year development.In addition to the pandemic, 2020 also saw significantly heightened catastrophe activity, including bushfires, hail and storm losses in Australia, derecho and wildfires in the U.S. and of course, a record Atlantic hurricane season. Although a number of major reinsurers described 2020 as the fifth worst year on record, it was a much tougher year for primary insurers because it was driven by frequency rather than severity. There were 28 insured loss events of $1 billion in 2020, which is easily the highest ever and compares with only 18 in 2017, which is widely regarded as one of the costliest catastrophe years in history.I think this context is important and helps explain why our catastrophe costs finished 1.1% above our allowances in 2020. I understand that one of the most disappointing aspects of our 2020 result is the adverse prior accident year development, which amounted to $366 million or 3.1% relative to net earned premium.I will comment briefly on prior accident year development in a moment while Inder will comment in more detail later. Premium rates have improved further in 2020, with the group achieving an average annual renewal rate increase of nearly 10%, up from over just over 6% in 2019. Momentum accelerated across 2020 and especially the second half with rate of 12.6% achieved in the fourth quarter. In the Northern Hemisphere, this is the strongest premium rate environment we have seen for at least a decade.And to preempt the inevitable question, what is driving the pricing cycle and how long might it run for? I would like to point to the following factors: firstly, the general core profitability following more than a decade-long soft cycle; secondly, catastrophe losses have been elevated in recent years, including more significant unmodeled losses; next, favorable reserve development that subsidize results during the soft cycle is coming to an end; fourthly, social inflation and the propensity for litigation remains a material industry cost and a sort of significant lingering uncertainty. Underwriting margins have to adjust to compensate for dramatically lower interest rates; and finally, COVID-19.While we remain confident with our ultimate net cost of $785 million, the eventual cost and timing of emergence of COVID-19 losses remains a significant risk for the industry and is undoubtedly contributing to rate increases. It is easier to point to the causes of the current hardening cycle than it is to predict its duration. But rates are expected to keep increasing for at least the remainder of 2021. Many of the factors driving the current cycle are not quick fixes, and some will take likely multiple years of rate to rectify before the industry returns to an appropriate risk-adjusted return on capital. While investment returns were clearly well down on 2019 for obvious reasons, we enjoyed a strong rebound during the second half to finish 2020 with net investment income of $226 million, equating to a 0.9% return.Our asset allocation hasn't really changed since the interim result. We remain cautiously invested, reflecting still significant global economic uncertainty. But also frankly, with interest rates where they are, we currently see better opportunities for capital deployment across the group's underwriting business. Inder will speak to the balance sheet in more detail later but suffice it to say that we're in good shape. Regulatory capital is above the midpoint of our target range, S&P capital in the AA territory and pro forma gearing has improved significantly to be comfortably within our target range.If we now turn to Slide 4, which is the result snapshot. While I don't want to harp on about growth, having just posted a $1.5 billion loss, after an extended period of remediation and disposals, which has seen GWP shrink from over $18 billion in 2012, it is pleasing to report meaningful GWP growth which you can see has increased sharply to $14.7 billion. I will talk about where we're achieving growth in a later slide.I mentioned rate momentum earlier and the premium rate chart in the middle of the slide speaks for itself. The rate we achieved in 2017 was entirely driven by Australia Pacific, with North America and international still in the throes of a soft cycle and reporting rate reductions. While pricing conditions improved in 2018 and 2019, momentum clearly accelerated in 2020, driven by the Northern Hemisphere. And as I will show you in a later chart, that momentum accelerated steadily across 2020 and especially in the second half.While it is early days, rate increases have continued into 2021, albeit there are some signs that the level of rate increases might not continue to accelerate like they did in 2020. On claims inflation, we are currently allowing for around 3% in both Australia Pacific and international and 5% in North America, which equates to nearly 4% across the group, the higher level in North America, reflecting social inflation and more cautious current accident year loss picks. We are now achieving quite meaningful rate that is comfortably ahead of loss cost inflation. And as we enter 2021, we will benefit from more rate on rate. While this is indicative of margin expansion, there is still much uncertainty for all insurers to navigate.I mentioned the high-level drivers of the year-on-year rate movement in combined ratio earlier so won't dwell on it here. The significant COVID-19 impact on the 2020 result is self-evident, and I will talk more about COVID-19 in a later slide. In 2020, we achieved a further 2.9% improvement in the attritional claims ratio, which largely reflects the earning of past premium rate increases. Given the premium rate increases we achieved in the second half of 2020 and the early rate trajectory in 2021, I would expect to see further improvement in our attritional claims ratio in 2021.I've already provided some context for the spike in our catastrophe costs during 2020, so we'll instead focus on the trend in our large risk claims ratio, which is the bottom segment of the bars in the large risk and catastrophe claims ratio chart. On a like-for-like reinsurance basis, the net cost of large risk claims has improved from 8.5% in 2017 to 7.9% in 2020, reflecting rate increases but also improved risk selection and derisking initiatives, including, as an example, multiple actions taken to reduce the risk profile of European financial lines portfolio over the last 36 months as well as less significant derisking in other classes such as international liability and property catastrophe exposed lines. I mentioned the European business in this context as this is where most of our large risk claims allowance is.Turning now to prior accident year development. Adverse prior year development ended up being $366 million or 3.1% of net earned premium in 2020. We had a small amount of net releases in Australia Pacific, which was more than offset by around $80 million of adverse development in international and just over $300 million in North America. I won't dwell on Australia Pacific other than to say that we believe industry reserve redundancies are largely depleted, so pricing will arguably need to continue to adjust to reflect current accident year loss ratios that are no longer being subsidized by significant prior accident year releases.International experienced a modest amount of adverse development, largely due to strengthening in financial lines, some loss creep in 2019 catastrophes and strengthening in some North American inwards reinsurance treaties. North America accounted for the majority of the group's adverse development. And I probably don't need to remind you but 2020 is not the first year that North America has experienced adverse development.I will let Inder discuss the specific issues that contributed to North American claims development and will instead preempt the obvious question: Have we got it all? And can we guarantee that there won't be any further adverse development? The short answer to that question is no, I can't give any guarantees. But what we have done is the best job we can, and I'm hopeful that in addition to strengthening in response to observed development and the explicit $100 million provision largely for social inflation, coupled with improved new business quality and strengthened current accident year loss picks, this has reduced the risk materiality. Inder will talk later to our current accident year loss picks in North America, which have been materially strengthened.I'd now like to turn to Slide 5, which is the pricing momentum. As I said earlier and is even more evident from the chart on Slide 5, in 2020, we saw strong and accelerating rate momentum in most markets around the world but especially in the Northern Hemisphere. Momentum across 2020 was strong, with new group achieving an average rate increase of 8.7% in the first half of the year, rising to 11.3% in the second half, including a final quarter of 12.6%.The notable exception was Australia Pacific, where rate increases slowed across the second and, especially, third quarters as we temporarily suspended rate increases for SME, commercial motor and A&H renewals. Rate momentum recovered to 6.4% in the fourth quarter as COVID-19 pricing relief initiatives rolled off and as we implemented further performance-based rate changes across a wide range of commodity classes, including business packages, farm and householders. Although early days, the positive fourth quarter rating momentum has continued into 2021.Key portfolios in Australia Pacific, where we have seen strong rate, include aviation at plus 14%; commercial property at plus 13%; and householders and strata at plus 8%. Premium rate increases in Australia Pacific are clearly above our 3% allowance for claims inflation. However, given recent catastrophe experienced in Australia, I'm pushing the team to achieve even stronger increases for catastrophe-exposed businesses.As I mentioned earlier, the Northern Hemisphere is experiencing the strongest premium rate environment in at least a decade, perhaps best evidenced by international's fourth quarter rate increase of 19%. In North America, rates momentum improved in every quarter of 2020, with the business achieving an average rate increase of 9.5% in the first half of the year, which increased to 10.8% in the second half, including a final fourth quarter increase of 10.9%. This compares with our allowances for claims inflation of around 5%, which importantly allows for social inflation in impacted classes to continue at high single digits to low double digits over the medium term.Some of our U.S. peers report rate increases excluding workers' compensation because up until quite recently, comp rates had been falling in the U.S. On that basis, our North American business achieved 12.1% rate increase in the third quarter, which increased to 12.4% in the final quarter.Key portfolios in North America where we have seen strong rate include professional and financial lines of plus 19%; aviation at plus 18%; property cat programs of plus 17%; specialty and casualty programs at plus 12%; and finally, A&H at plus 11%. January renewal rates in North America remain encouraging.For international, the quarterly picture speaks volumes for market conditions, with rate momentum building strongly across the year. The business achieved an average rate increase of 10.1% in the first half, which increased to 16% in the second half, including a final quarter of 19%, underpinned by international markets, which achieved a final quarter rate increase of 24.7%. This compares with our allowances for claims inflation of around 3%.Key portfolios in international where we have seen strong rate include in the International Markets business, which achieved 19%. As part of that, financial lines at plus 28%, liability at plus 24% and property at plus 17%. In the U.K., financial lines achieved plus 34% and property plus 16%. And our Canadian portfolio achieved plus 21%. Although they doubled, 2020 rate increases for QBE Re were still disappointing at only plus 5%. And although up appreciably in casualty and specialty at the recent 1/1 January renewals, rate was still below expectation for property catastrophe classes where we remain cautious.While Asia remains a laggard, rates have improved off a low base and now be plus 7%, underpinned by a rate of plus 12% in Singapore. January renewal rates for European operations remain positive in all lines and geographies, with increases at least in line with our planning assumptions.Finally, I should point out that the group has seen a steady improvement in premium retention across 2020 and indeed the past 18 months. This is partly due to completion of all the heavy remediation of recent years but also reflects our increased customer focus and improved service levels and bodes well for premium growth in future years, which comment is a good segue to the next slide, which is Slide 6, which is gross written premium.On a constant currency basis and adjusting for 2019 disposals, gross written premium grew by 10%, reflecting the strong pricing environment, improved policy retention and targeted new business growth. If you allowed for portfolios such as crop and CTP, which are not included in our average renewal rate increases data, and our premium retention ratio of around 80% -- 82%, underlying real growth, i.e., growth in excess of lapsed business and premium rate increases was around 4%.Although pricing has improved across most lines and across all regions, we remain disciplined and selective, only targeting new business that is appropriately priced and within our risk appetite. Our priority remains to use this period of repricing to drive technical pricing across all portfolios to ensure appropriate risk-adjusted return, allowing for currently low investment returns. North America achieved GWP growth of 13%, with growth centered on classes where we have strong market franchises and core competencies, including: A&H plus 20%, crop plus 14% and property programs plus 23%. International achieved GWP growth of 12%, underpinned by growth of 22% in international markets, where growth was strong across all lines but especially Canada, financial lines, liability and marine.Several competitors continue to adjust their risk appetite by reducing capacity and underwriting authority, which is generating some quality new business opportunities, especially in London. Growth in the U.K. regional business was more modest, reflecting a significant reduction in commercial motor premiums associated with reduced activity due to the COVID-19-related lockdowns.QBE Re achieved growth of 11%, reflecting new business associated with the recently opened Dubai office, coupled with rate and new business in specialty and casualty lines. Gross written premium in Asia fell by 10%, reflecting the impact of COVID-19 on Hong Kong travel and cargo insurance, as well as derisking initiatives in trade credit taken at the beginning of last year. Australia Pacific achieved underlying growth of 6%, underpinned by strong growth in strata, householders, LMI and New Zealand, while the Pacific Islands business contracted due to the severe impact of COVID-19 on the region's economy.In a hardening market like this, it is important to not only see rate in those portfolios where risk-adjusted returns are insufficient but also to optimize our portfolio by driving growth in higher-margin books of business. In this regard, in 2020, we achieved GWP growth of between 10% and 20% and often in excess of that in portfolios that are amongst the most profitable in each division, including, for example, in international and in International Markets, our natural resources business, liability and financial lines, Continental European liability, QBE Re casualty and the Tracker facilities; in North America, our A&H and crop businesses; and in Australia Pacific, we see good growth in New Zealand, farm, commercial property and general aviation.Before I move on, I should add that growth in net earned premium was not as strong as growth in gross written premium in 2020. This was partly due to a large increase in reinsurance costs, which included the derisking reinsurance initiatives undertaken at the onset of COVID-19. But there is also a $500 million lag in the earning of significant rate-driven growth in gross written premium. Unless top line growth accelerates again, our net earned premium trajectory to be closer to gross written premium in '21 as the 2020 growth is fully earned.If we now move to Slide 7, which is the attritional claims ratio. The benefit of the improving premium rates is clearly evident in the trajectory of the group's attritional claims ratio. Having improved by 2.7% in 2019, the group's attritional claims ratio improved a further 2.9% to 44.6%, with improvements seen across all divisions but especially international. Despite the impact of derisking reinsurance initiatives and more prudent current accident year actuarial assumptions, North America's attritional claims ratio improved 2.4%, reflecting the improved pricing environment and the nonreoccurrence of attritional weather that occurred in 2019.International's attritional claims ratio improved by 4%, reflecting the increasingly favorable pricing landscape supported by targeted underwriting actions in our European and Asian businesses. Australia Pacific's attritional claims ratio reduced by a further 1.7%. Improvement was observed across most portfolios but especially CTP, New Zealand and farm, partially offset by attritional weather, which adversely impacted the household and commercial property books. Given the earning of 2020 rate increases in 2021 and the ongoing 2021 rate increases themselves, we expect further improvement in the group's attritional claims ratio in 2021.If we now turn to Slide 8, which is the COVID-19 update. Firstly, just let me say, there is certainly no new news on COVID-19 front today. Our estimated ultimate net cost of COVID-19 is unchanged at $785 million and comprises the $655 million cost recognized in 2020 and a further $130 million of largely claims expense yet to be incurred. The $785 million is net of reinsurance and includes claims benefits offset as a result of reduced frequency associated with the COVID-19 lockdowns, mainly in the U.K. and Australia Pacific motor portfolios.As can be seen in the chart, the $320 million charge during the second half was dominated by a further $185 million risk margin charge we announced last month, which brings the COVID-19 risk margin to $300 million. While the U.K. Supreme Court ruling did not directly impact QBE's net claims position, the additional risk margin announced last month ensures that we have allowed for the net cost of Australian business interruption claims even in extreme scenarios. The remaining $135 million of net charges during the second half largely reflected additional claims in areas that should not surprise, including trade credit, casualty, LMI driven by a $60 million provision to protect against a deterioration in unemployment to 8% concurrently with a 5% fall in house prices and Australian business interruption.The $130 million of largely claims costs expected to be incurred in 2021 are likely to be in areas such as LMI, trade credit and casualty lines. Trade credit claims to date have been far less than the more extreme scenarios that we felt were possible at the onset of the pandemic. This reflects significant portfolio derisking initiatives, coupled with a more resilient economic backdrop in Australia and the government reinsurance backstop implemented in the U.K. last year that has been extended until the 30th of June this year. Nevertheless, we do expect additional trade credit claims to emerge as fiscal stimulus around the world is eventually unwound.Similarly, our LMI business has performed better-than-expected due to the management of the pandemic and the resilience of the Australian economy and housing market. COVID-19-related hardships are down to around 1% of active policies from around 6% during the August 2020 peak. Underlying arrears are broadly stable. Unemployment is relatively stable at around 6.5%, and the local housing market remains surprisingly firm, supported by a recovering economy and record low interest rates.In light of this, I would hope that the $60 million LMI provision taken up at year-end will prove unnecessary, but we felt it prudent to do so, given the long-tail nature of this business. In addition to the significant reinsurance protection, the $300 million of COVID-19 risk margin gives us confidence that the $785 million net ultimate cost of COVID will prove sufficient. The majority of our claims expense today comprises IBNR, and it will take some time for claims to develop and even longer for claims to be finalized. In light of that and even in more optimistic scenarios, I would caution you not to expect risk margin releases for the foreseeable future.If we now turn to Slide 9, which is business interruption. The insurance industry's potential exposure to COVID-19-related business interruption claims have been the topic of greatest concern for investors and the source of the greatest uncertainty for the industry. In North America, we are currently of the view that our exposure is minimal. All our properties, policies require physical damage to trigger business interruption coverages. And the vast majority, more than 99% of policies, include the extra protection of a virus exclusion.To date, the vast majority of the U.S. court ruling, something like 121 out of 131 at the federal level on the subject of whether the virus causes physical damage to property, have been in favor of the insurance industry. U.S. business interruption wording is far less affirmative than the U.K. and, to a lesser degree, Australia. So while we cannot completely rule out the risk of widespread social justice-based court ruling, as proved to be the case in the U.K., we regard the possibility as remote.But if you want to play devil's advocate, as is the case in the U.K. and Australia, nondamage business interruption claims are protected by our catastrophe reinsurance. While the result -- recent results in the U.K. Supreme Court ruling was surprising and overturned established case law precedent, it does bring more certainty to the U.K. business interruption landscape. And the industry can now commence evaluating and settling valid claims. As we said in our ASX release last month, while our gross U.K. business interruption claims costs increased as a result of the ruling, our net claim cost remains $17 million as a result of the various reinsurance treaties.The situation with respect to business interruption in Australia remains far more uncertain. The ICA has applied to the High Court of Australia for special leave to appeal the judgment with respect to the initial test case surrounding the incorrect reference to the Quarantine Act. Regardless of the outcome of that application and any ensuing appeal, there will be further test cases to resolve legal issues around the interpretation of common business interruption policy wordings. And those cases will take some time to be heard and even longer for the rulings to be handed down, though it is difficult to see any clarity emerging with respect to Australian business interruption in the first half of 2021. And the time frame could easily extend towards the end of 2021 and beyond.QBE has set aside material gross provisions for potential Australian business interruption claims, that is considered appropriate in the circumstances, having regard to the group's potential exposure and plausible claim scenarios. In light of those gross provisions, together with substantial risk margin and significant reinsurance protection, the group considers that it has allowed for potentially extreme, albeit highly unlikely, Australian business interruption claim scenarios.Again, to preempt the question, I thought it appropriate to make a few comments about the recoverability of nondamaged business interruption claims under our various catastrophe reinsurance treaties. We have been consistent all along on our position in this regard. Since the onset of COVID-19 and especially since the U.K. Supreme Court judgment last month, we have spoken on many occasions with our reinsurers, the reinsurance brokers who placed the covers and specialist external reinsurance legal counsel, including again after the recent U.K. test case ruling.Moreover, we have renewed our reinsurance treaties with the same reinsurers. Although there can be no absolute 100% guarantees, our position is unchanged after all those conversations and considerations. I will now hand over to Inder to take you through the financials in more detail.
Thank you, Richard. Good morning, all. As Richard has highlighted, the headline result is very much in line with our December announcement and our subsequent business interruption update in mid-January. Whilst we're disappointed with the headline result, the operating momentum in our business remains strong, and I'm confident that we can deliver significant further improvement in the underlying margin beyond the trends that are evident in the 2020 result.As Richard has updated you on the impacts of COVID-19, I'll focus my remarks on performance, excluding these impacts. I'll start with the overall group P&L on Slide 11. Gross written premium increased by 10% on a constant currency basis and excluding the impact of 2019 disposals. The combined operating ratio deteriorated by just over 1 point to 98.6%, reflecting above-average catastrophe activity and adverse prior accident year claims development, primarily in North America. Net investment income was $226 million compared with more than $1 billion in 2019. Our investment portfolio recovered strongly in the second half, and the unrealized mark-to-market losses on our investment-grade credit book more than fully reversed as markets rebounded and credit conditions stabilized.The adjusted cash loss after tax for the year was $863 million, reflecting combined impact of COVID-19, elevated cat and adverse prior year development. Our financial position remains strong with the PCA multiple above the midpoint of our target range. Pro forma gearing remains well within our target range despite the $600 million in noncash impairment and amortization charges we took in the second half. Given the statutory loss, the Board elected not to declare a final dividend for 2020. Subject to stable global economic conditions, we expect to resume dividend payments up to 65% of cash profit in conjunction with the 2021 interim results.I'll now step you through some of the key movements in our combined operating ratio on Slide 12. There is a fair amount of noise in the headline results, and what we're trying to do here is to isolate some of the nonrecurring elements to better demonstrate the improving underlying trends that we're seeing across the business. On the right-hand side of the top chart, you can see the reported combined operating ratio of 104.2%. As you work towards the left, COVID-19 had a 5.6-point impact on the combined operating ratio. We expensed $655 million in 2020, with the remaining $130 million of our estimated ultimate COVID-19 costs expected to be incurred in 2021. The combined operating ratio was also impacted by 40 basis points due to a modest increase in risk margin unrelated to COVID-19, mainly reflecting FX and the impact of lower discount rates on the central estimate. The probability of adequacy of net outstanding claims increased to 92.5%, which is right at the top of our range.The next block along represents a 3.1-point impact relating to adverse prior accident year claims development and I'll expand on this shortly. As Richard referenced, the 2020 net cost of catastrophes was just over 1 point above our allowance for the year. Normalizing for these items, the underlying current accident year combined ratio of 94% was more than 4 points better than the prior period. And what's really encouraging is that this improvement was underpinned by sustainable, repeatable drivers that will continue to support a stronger, more resilient earnings profile going forward.The waterfall chart below sets out some of these drivers. The attritional claims ratio improved further by around 3 points, with each division contributing to this positive trend. We are pleased with the progress we're making on large individual risk claims. By their very nature, these claims are considerably more volatile than attritional claims. And so we expect some fluctuation from one period to the next and especially from one half to the next.Importantly, the large individual risk claims ratio improved further during 2020 despite absorbing some headwinds that particularly impacted the second half. Firstly, additional reinsurance in response to COVID-19 reduced net earned premium by $325 million and adversely impacted the large risk claims ratio by 20 basis points. Secondly, we booked more cautious current accident year loss picks, including IBNR as we exited 2020, particularly in North America and I will touch on this a little later. On operating expenses, our efficiency program has delivered cost reductions ahead of schedule, and similarly, I will pick this up in more detail later. It's just worth noting on crop, whilst we're showing a year-on-year improvement in this waterfall chart, we've had 2 very disappointing years in this business, with combined operating ratios of around 109% and 98% in 2019 and 2020, respectively. As flagged in December and in view of recent performance, we have revised the assumed crop combined ratio in our plans to 92% from 90% previously.I'll now turn to divisional performance in more detail, starting with North America on Slide 13. Gross written premium of $4.8 billion was up 13%, adjusting for the sale of retail personal lines. As you can see on the chart, the normalized current accident year combined ratio improved by around 3.5 points to 100.3%. Within this, the attritional claims ratio improved by 2.4 points, driven primarily by the strong rate momentum across the business. The cost of large risk claims increased by 2.3 points, reflecting increased severity in our general aviation book and materially stronger current accident year loss picks across the portfolio.These strengthened current accident year loss picks largely reflect additional IBNR. And to give you an indication of magnitude, large risk loss IBNR for 2020 accident year is 40% higher than the IBNR we held against the 2019 accident year as we exited 2019, and it's up nearly 130% on 2017 on the same basis. These are very significant increases. Even with the benefit of hindsight, the IBNR we would have held on these old years is not as much as we are now reserving for the 2020 accident year. Similarly, excluding COVID-19, the total IBNR we are holding against attritional large risk and cat claims for the 2020 accident year is around 8% of NEP higher than the prior period.The main disappointment in North America was obviously the adverse prior accident year development. We booked an additional $190 million in the second half, bringing the full year number to $305 million or 9% of earned premium. On a full year basis, this adverse development related to 4 main areas: firstly, $70 million related to the discontinued excess and surplus lines portfolio. This is a book of business from a couple of years ago, where with the benefit of hindsight, we didn't have the necessary expertise and ended up with a portfolio of largely habitational business that wasn't well underwritten or appropriately priced.Around $60 million of development related to general aviation. This largely reflects a single, very large claim involving a private jet that entails significant loss of life and substantial property damage, where the Transport Bureau made a surprise finding of pilot error rather than engine failure. Around $30 million related to 2017 catastrophes, particularly Hurricane Irma. The entire industry has experienced adverse development on 2017 catastrophes and especially in relation to Hurricane Irma.And lastly, we've made an additional explicit provision of around $100 million to address more systemic issues such as social inflation and the potential for higher severity loss trends in casualty lines. Importantly, this $100 million explicit provision is more forward-looking and not just a direct response to the specific underlying development observed in 2020. Separately, we've made good progress on our operational efficiency program in North America, with around $60 million of net cost savings realized over the last 24 months. This has been driven by our exit from personal lines and the associated simplification and rationalization of technology infrastructure, operating model and regional footprint. Moving forward, we continue to see opportunities to further improve our operating efficiency in North America.Turning now to international on Slide 14. Gross written premium of $5.9 billion was up 12% on a constant currency basis and excluding the impact of asset sales in Asia. This growth was underpinned by the strong premium rate increases in international markets, around 20%; U.K., around 15%; and Continental Europe, around 8%. As you can see on the chart, the normalized current accident year combined ratio has improved by a remarkable 7.8 points to 90.6%, with meaningful improvements in both the attritional claims ratio, down around 4 points, and the large loss ratio down around 3 points. This improvement clearly reflects the earn-through of the significant premium rate increases as well as our continued focus on underwriting discipline and risk management.At a portfolio level, we have seen marked improvements in the profitability of many of our larger books, including within International Markets, our natural resources book, liability book and financial lines portfolio, our commercial motor and liability books in the U.K. and commercial property book in Continental Europe.On prior accident years, whilst there was positive development across much of international, the overall net adverse development of $80 million or 1.7 points was driven by 3 main areas: firstly, financial lines, predominantly U.K. professional indemnity and, to a lesser extent, U.K. management liability for 2017 and prior years; second, reserve strengthening on some North American treaties within QBE Re. These treaties are now reported within international as we have consolidated all global inventory insurance business within QBE Re; and lastly, loss creep on 2019 Japanese typhoons and Hurricane Dorian. This is not unique to QBE.On operating expenses, the expense ratio improved by 140 basis points from 15% to 13.6%. This reflects continued improvement in operating efficiency in both Europe and in Asia, with disciplined cost containment enabling us to benefit from operating leverage as the premium rate environment improves.Lastly, our home market of Australia Pacific on Slide 15. Gross written premium of $4.1 billion was up 6% on a constant currency basis and excluding the impact of asset sales. This is despite the suspension of rate increases on selected lines during Q2 and Q3 as we provided COVID-19 relief to customers. The normalized current accident year combined ratio was strong at 91.5%, a 50 basis points improvement from the prior year. The attritional claims ratio improved by a further 1.7 points, and in aggregate, has improved by more than 14 points since we commenced detailed cell-level performance management in late 2016.The large risk claims ratio improved by 20 basis points, supported by strong performance in commercial property, partly offset by some volatility in CTP and, to a lesser degree, in New Zealand. A significant improvement in large losses since 2017, alongside improvement in the attritional claims ratio, has put the Australia Pacific business on a much stronger, more sustainable footing with a vastly improved quality of earnings. On prior accident year reserves, releases from CTP and commercial property were partly offset by minor strengthening in public liability and New Zealand professional indemnity. Overall, we remain comfortable with the reserving trends in liability lines.The reported expense ratio was 80 basis points higher than 2019. This included nonrecurring risk and regulatory costs and a new South Wales CTP profit normalization charge of $61 million. This CTP charge is based on our estimated performance relative to industry profitability and the schemes benchmark margin and is consistent with the favorable experience of lower claims frequency that has supported our reserve releases.On lenders mortgage insurance, lead indicators and credit metrics have trended better than expected, with most bank customers having resumed mortgage repayments after brief COVID-19-related payment holidays. Having said that, this is a long-tail business and the economic outlook, while stabilizing, remains uncertain and supported by significant stimulus programs that are yet to roll off. Given this backdrop, we are being cautious about the economic assumptions underpinning the LMI reserving model. And accordingly, we booked an incremental reserve of $60 million for COVID-19-related impacts. Including this additional reserve, the LMI business reported a combined operating ratio of 103% for financial year '20.Turning now to Slide 16. You'll recall that in December 2018, we announced a 3-year operational efficiency program, targeting around $200 million of gross savings, $130 million of net savings and an expense ratio of less than 14% by 2021. 2 years into the 3-year schedule of work, the program has progressed ahead of plan. As you can see on the chart on the left-hand side, our 2020 exit run rate costs are estimated at around $1.69 billion, which equates to an underlying expense ratio around 14.3%.This exit run rate excludes one-off items, both positive and negative, including the elevated risk and regulatory costs that we flagged at the end of 2019, restructuring charges, the CTP profit normalization charge, significantly lower variable rem and other one-off net savings. We are pleased with the meaningful progress we've made in rationalizing our technology estates, retiring legacy applications, simplifying our operating model and reducing third-party consulting, travel and discretionary costs. Additional savings were also realized from the disposal of retail personal lines in North America and related reduction in our regional footprint. As we transition into the next phase of our program, we see meaningful opportunities to improve operational leverage by minimizing expense growth as we take advantage of the organic growth opportunities offered by the current trading environment. We recently commenced our IT modernization program, which includes moving the majority of our IT estate to the cloud and exiting legacy data centers, consolidating technology infrastructure services under a common provider and automating our IT operations, driving further digitization and process automation throughout the organization.More broadly, we also see opportunities to continue to refine our operating model, our ways of working, including where and how work gets done to drive year-on-year improvements in operating efficiency. As part of the next phase of work, we'll incur a $150 million restructuring charge, primarily related to the IT modernization work to be expensed over the next 3 years. This will support a target expense ratio of 13% by 2023.Turning now to investment performance on Slide 17. We have obviously seen extraordinary market conditions over the course of 2020, with severe dislocation and elevated volatility especially immediately following the onset of the pandemic. As we move into 2021, the outlook for the global economy and capital markets appears more stable. Following our derisking actions in the first half of 2020, we have maintained a conservatively positioned investment book with a high-quality investment-grade credit portfolio. 85% of fixed income is rated A- or better. We've had fewer corporate bond downgrades in the market. The BBB portfolio is resilient with fewer exposures on negative outlook compared to the market. We have no [indiscernible] and none of our credit is trading distressed.It is worth noting that the total balance of cash and investments at the end of 2020 was $27.7 billion, up significantly from the prior year. Looking forward, the developments through 2020 have clearly reset the outlook for investment returns, as is clearly evident in the chart on the right-hand side of this page. Our fixed income running yield is now around 40 basis points and this was more than 5x higher in 2018 at 2.2%. We have seen an incredible leg down in global risk-free rates while credit spreads are modestly tighter, supported by the extraordinary policy stimulus we've seen over the last few months. Based on our asset allocation today, this translates to a FY '20 exit annualized return of around 90 basis points or 10 basis points lower than we referenced at our December update, reflecting further spread compression.I will conclude with some observations around our reinsurance profile, balance sheet and capital position on Slide 18. In early January, we announced the successful renewal of the group's reinsurance program. This program was placed broadly in line with our expectations at terms slightly better than allowed for in our planning assumptions and strikes an appropriate balance between cost, capital and earnings volatility protection.The chart on the left-hand side here sets out key movements in our reinsurance spend, which increased from $1.6 billion in 2019 to $2.3 billion in 2020, mainly reflecting our Crop business. In 2020, we purchased additional quota share reinsurance to seed 90% of crop hail exposure, and combined with significantly reduced MPCI recoveries, our property insurance expense increased by $438 million. Crop reinsurance expense is highly variable and difficult to predict, with quota share sessions impacted by variability in gross writings and MPCI recoveries impacted by scheme profitability in any particular year.Reinsurance expense was also impacted by the additional cat cover we purchased to reduce North American peak retention to $150 million from $400 million previously as part of derisking initiatives we executed in April last year. The PCA multiple is above the midpoint of our target range. This is despite a material COVID-19 impact, with the vast majority of our $785 million net ultimate cost of COVID-19 either incurred in the FY '20 result, accounted for in risk margin or reflected as a deduction from capital through premium liabilities.Our capital position has also absorbed the risk charges associated with the expansion of our balance sheet over the course of 2020, with cash and investments up 9% in constant currency terms, unearned premium up 10%, gross outstanding claims up 16%, including the impact of greatly reduced risk-free rates. Similarly, our S&P capital position remains strong with a modest surplus to the AA minimum capital requirement.Allowing for the prefunded subordinated debt repayment that will be completed in early March, our debt-to-equity ratio is around 32%, down from 38% a year ago and 41% in 2017, while head office liquidity is around $1.2 billion. With the operating environment remaining uncertain, we are maintaining a rigorous focus on capital and risk management, with our decision-making informed by dynamic, forward-looking stress and scenario analysis. With that, I'll hand back to Richard to discuss our priorities and outlook for 2021.
Thanks, Inder. Before we move to Q&A, I'd like to quickly mention my priorities for 2021. Performance is and has to be my #1 near-term focus. The market conditions we are currently experiencing are rare, and so I'm determined that we maximize the opportunity for both margin expansion and growth. I'm not envisaging any changes to the group's strategic agenda. Cell reviews and Brilliant Basic program remain integral. However, I am reinvigorating and evolving them, increasing the emphasis on capital allocation and targeting growth in portfolios generating the highest risk-adjusted return.We will challenge ourselves to raise the bar in all aspects of performance management, from risk selection to pricing to claims management. I will ensure that all our QBE colleagues realize the importance of these activities. I will talk about my other priorities at the 2021 interim result, including our modernization journey, which we are optimistic will facilitate material additional efficiency savings over the medium term.Finally, if we move to Slide 22, which is the outlook. Given the considerable uncertainty as a result of the pandemic and its lingering impact on the global economy, we have decided not to provide any results targets for 2021 financial year at this stage. That said, and as Inder discussed on Slide 12, we would regard our underlying 2020 current accident year combined ratio to be around 94%. If you adjust this for the recently flagged increase in our group XOL reinsurance costs and the increase in the catastrophe allowance to $685 million, but also assume a normal crop combined operating ratio, the 2020 exit combined operating ratio is around 95%. That is as good a base level from which to assess the likely margin expansion in 2021.While premium rate increases are clearly well in excess of loss cost inflation in most of our portfolios, there is still lingering uncertainty around claims inflation and the ongoing level of catastrophe activity. With that, I will open the call to Q&A.
Thanks, Richard. First question this morning is from Kieren Chidgey of Jarden. [Operator Instructions] Over to you, Kieren.
I've got a couple of questions. Richard, I might just start with you around GWP. Obviously, it's a very strong rate backdrop and you talked to disciplined growth and flagged around 4% volume growth in FY '20 on an underlying basis. Just wondering if you could give us some thoughts around sort of volume growth prospects for the business through '21.
Yes. Look, we -- our plan, we haven't assumed everything will grow as much as it has in the last 12 months, but I still think that we will see strong -- whether we see double digits, I don't know. But the fact is we'll see material GWP growth across our portfolios. We'll see a bit more in Australia because we'll get more rate. I don't think one can expect to see quite as much strong rate and growth in some of the London market businesses. But I think some of it will carry on in the U.S. So the benefit of having a widespread portfolio is that 1 or 2 may not achieve quite as much growth as in 2020 but others will. So no, we're not giving the guidance around that, but you can assume that will be high single-digits growth.
Right. And volumes specifically within that, which was sort of more where I was aimed, still looking for positive volume growth ex rate?
Yes, I think so. Look, the best rate we can get is rate growth. The best growth we can get is rate growth because that goes to the margin. And it's -- and if we get rate on renewal business, that's a business we understand, so we can feel far more comfortable about the underlying performance of that business. But we're not an organization to say that we're going to grow indiscriminately.We don't view it's a market where you just write anything that moves. We're still going to be careful and selective. And frankly, QBE's made too many mistakes in the past by pushing growth in the wrong place. So our growth will be deliberate and targeted around where we're comfortable, we understand the business and we understand the margin rather than just going indiscriminately for any type of growth. So again, you would like to see some growth above rate. I can't give you what the number would be but it would be nice and we would expect to see some.
My second question, just in regards to this 95% underlying combined ratio exit run rate. Just hoping, Inder, you could unpack in a bit more detail what you're assuming around the reinsurance expenditure sort of maybe as a proportion of GP or NEP -- sorry, GEP and also around the cat budget and large risk allowance for '21.
Yes. Look, Kieren, maybe just a couple of comments. I think in terms of the reinsurance spend, as I referenced in my remarks, it is volatile in aggregate because of the crop movements, albeit those crop movements don't actually go to margin. So what we try to do is the element that does go to margin, we've shown you, it's about a 30 to 40 basis point drag year-on-year in terms of the additional cost on what we call the group covers, which is, in essence, most of the XOL program. So that's the relevant bit.Clearly, the writings in crop will be driven by where we end up on commodity prices, what happens to the vol factors, et cetera. So -- but look, I wouldn't stress too much about that. I mean, that tends to net out but the margin impact we've shown you very clearly. In terms of cat and large, I think we've been pretty clear on cat, where we think the allowance sits and we feel good about that. The large, as I referenced, it is volatile from one period to the next. We are pleased with the progress and the trajectory. We've got moving in the right direction. That's despite some of the, what I'd say, fairly cautious current accident year loss picks we struck at the end of 2020.So look, the working assumption is that, that ratio will come down a bit from where we've exited but not a lot, right? We're going to be cautious, just given that some of the inflation is a bit spiky. So albeit inflation is in that 3% to 5% zip code, it is a bit spiky. So we're being careful in some of the areas, in particular.
All right. And my last question, just around the expense target you put out there for FY '23 of 13% and the $150 million of one-off costs to achieve that. Just wondering if you can sort of talk through the profile. I know you're saying you're exiting at 14.3% ex COVID. How back-end weighted are those benefits expected to be and how are you going to account for the $150 million of one-off costs?
Yes. So in terms of -- the focus really is somewhat different to what we've done in the phase 1, I'd call it. The phase 1 has been the operating model adjustments that we've made, some of the technology rationalization that we've done. The phase going forward is very much centered around IT modernization so moving a lot of our estate to the cloud. And what that's going to help us do is invest in the right areas but contain the cost growth. So what we're really looking to do, Kieren, is to get some jaws as we get the earned premium come through and contain cost growth materially below revenue growth.But we do that through driving efficiencies primarily in technology in the first instance but we've got some other opportunities we're working on that will give us greater confidence as we execute into the program. So I'd say we'll give you a further update at the half year as to how we're tracking. But we should see those benefits come through fairly evenly between now and what I'd say, 2023.In terms of restructuring charges, we are going to apply the same treatment like we've done previously, which is we'll treat them separately as one-off below the line, and they're adjusted for an adjusted cash profit that we use to consider what our dividend payment is. So that's broadly the consistent treatment around those restructuring charges. And in terms of [indiscernible] impact of those, I mean, they'll be a bit lumpy. Probably some of that is going to be a bit more front-ended in 2021 but then sort of in 2022 and 2023.
Next question is from Andrew Buncombe at Macquarie. Go ahead, Andrew.
Just the first one, can you give us a bit of insight into how much headroom you have left in the FY '20 aggregate cover, please?
Richard, I'm happy to take that.
Yes. Inder, you take it.
Andrew, we have some headroom left. It's, I'd say, sort of north of $50 million-ish, probably less than $100 million-ish.
Got it, excellent. And then I can see the commentary about the cat allowances for '21, but maybe can you just confirm that the large loss allowance is unchanged into FY '21, please?
Yes. I mean, we're putting a fair amount of steer out. I mean, I think we're exiting 2020...
[indiscernible]
Okay. And then just -- sure. And then just a final one from me. Can you just comment on the provision gross of reinsurance for Australian BI losses? I'm sort of aware that the net is capped at $5 million per occurrence, but can you just make some comments on what that gross number would be for Australia, please?
Richard, do you want to...
Yes, we're not sharing the gross number. We don't think it's appropriate to do that when we're involved in potential litigation and where it is. So I think at the moment, that's not something that we'd be in a position to share.
Our next question is from Matt Dunger at Bank of America. Over to you, Matt.
If I could just follow up on the U.S. PYD. The reserve strengthening appears to be an industry issue, particularly when we exclude some favorability from workers' compensation. Can you comment how confident you are that you won't see further adverse PYD in the U.S. and elsewhere?
Look, there's a few approaches to that. As Inder mentioned, we have significantly strengthened the exiting year's IBNR, a higher level we've ever held. So obviously, the stronger we hold our exiting year, then the more secure that is, number one. Number two, we've had 2 stabs at it when we went through this in December. One was we responded to activity, which is what QBE has done in the past. But more importantly, we provisioned another $100 million, which was a forward-looking vision, which actually didn't respond to activity but was a provision looking forward for social inflation issues.We're comfortably aligned with our external actuaries and everything. So we think with our expiring year accident picks and with the picks that we put in our current year and with the provision for $100 million forward-looking provision rather than responding to deterioration, then we think we're in a lot better position.
Great. And if I could follow up with a question on the crop COR, which you've budgeted at 92%. Now what impact is pricing of the reinsurance program, in particular, hail, having on that and any potential additional covers that you might purchase on reinsurance?
The reinsurance on hail helps. Hail is the element of that business that runs poorly historically. We've had to see some MPCI premium to get that deal away last year. That remains consistent with this year. So the 92% is all in. It includes the impact. There is going to be no further impact from additional reinsurance that's going to come into that.The only thing that changes really mechanically between now and when we sort of set the pricing, et cetera, is just what happens to commodity prices, what happens to vol factors. And then we'll make the decisions around what we seed into the central fund. But 92% is inclusive of the reinsurance that we placed externally on hail, et cetera.
Okay. And if I could just ask, Richard, a final question. Richard, given your performance agenda and planning for the August result, can you give us any update on CEO appointment? And would you consider the role permanently?
Well, I've always made it clear that I'm not considering the role permanently. That's always been pretty clear. My agenda is I'm going to run 2021 with the leadership team. And my understanding is the Board is progressing well with the search for a new CEO.
Our next question is Ashley Dalziell at Goldman Sachs. Over to you, Ash.
Just a couple of follow-up questions on the new cost targets, maybe initially, the 13% target that you have out for the '23 year. Could you at least maybe just sort of give us a couple of the building blocks there, for instance, what the gross cost-out is that you're targeting and then potentially sort of what top line assumptions you have behind that 13% target as well?
Look, broadly speaking, we have assumed some level of top line growth so that's different to what we did in the phase 1 of the program. Obviously, we've got some good line of sight on 2021 with the earned premiums started to come through from the writings of the back end of 2020. So we've assumed some decent level of growth in 2021 but probably moderating in '22 and '23. It really depends on how well the cycle runs, et cetera, but we're not being too optimistic. We'll just have to wait and see how that pans out.In terms of gross to net, look, on the IT components, we think there's a meaningful growth opportunity around the $50 million, $60-odd million range. But really, we're focused on making sure that the net cost growth is contained. So whilst we're going to go after the growth very, very hard, what that's really allowing us to do is invest smartly and really connect the activity that we're doing to the value that we're generating. And so that's really important for us.There's opportunities ongoing. I mean, the cost structures in the industry more broadly could be described as archaic in many parts of commercial insurance around the globe. So we think there's plenty of opportunity for us to go at. But we also need to be mindful that we're investing smartly, given some of the growth opportunities that are emerging.
Okay. Just a follow-on then, I guess, maybe a little more conceptual, but in terms of how we think about QBE and operating leverage, you noted, obviously, good line of sight on top line growth into '21. But if that indeed does sort of continue at least into '22, should we be thinking 13% is the overriding factor? And any sort of incremental improvements that are coming through on operating leverage sort of beyond that get reinvested? Or is there scope for further margin expansion beyond the 13%?
13% is a target rather than a floor. Look, I think the reality with operating leverage, Ash, is that we haven't had any floor for like a decade, right? So for us to be sitting here and going we're absolutely confident about it beyond 2021 is -- we want to be realistic when we're setting these targets. We want to make sure they're achievable. We make sure we've got a pathway to do that. And look, with the operating leverage, let's see how we go. We can give you an update at the first half but we haven't experienced that for like a decade. So we don't want to get ahead of ourselves.
And I think the approach is multifaceted. We're obviously going to have the IT modernization, which in itself saves money and creates efficiency. We think we'll get further growth at the top line. And as Inder said, there's more to do with the general efficiency and cost base in the organization. So this is not like a one-trick pony approach to improving the ratio. There are several levers that we can pull and we will pull to go on that journey.
The next question is from Andrei Stadnik from Morgan Stanley. Go ahead, Andrei. We'll come back to Andrei. The next question then we'll move on to Nigel Pittaway from Citi. Go ahead, Nigel.
Okay. So as we went through -- as you went through, Richard, you were saying that you felt there were some signs that the premium rate environment might not continue to accelerate. So I was just, first of all, interested in what are the most obvious signs that you're seeing to make that comment.
Look, I think some -- the areas where we've seen the dramatic acceleration, I think it will probably slow up a little bit. We've got a bit of a 2-speed market in some ways. Anything that's financial lines, cat exposed, large specialty, is still seeing really good rate. Anything that's more mid-market in retail, you would find is maybe a little bit more moderate. And by that, I'd say what you're seeing in the Australian market, is moderate at around 5%, 6%, 7% compared to what we would be seeing elsewhere, but it's staying at 5%, 6% to 7% rather than accelerating.And if financial lines is at 20%, it will -- it's more like holding at 20% than going on to 25% or 30%. So it's just the fact that some of them are at high levels and to get a repeat of that level is a really good outcome. You don't necessarily need to get -- to go from 20% to 30% in subsequent years. So we're not saying the rates are coming off anyway because we're not seeing rates reducing. We're just saying we're seeing similar levels of rate that we saw last year and maybe some of them topping out a little bit.
Okay, that's clear. Second question, I mean, obviously, you've exited the year in the fourth quarter with average 12.6% rate rises. You're saying that your sort of average claims inflation sort of assumption is around about 4%. So prima facie, I mean, I know there's things like sort of high reinsurance costs and cat allowance to factor in too, but prima facie, that means that margin expansion, if anything, should be accelerating from what you've achieved this year. Is there anything we're missing in terms of drawing that conclusion?
Well, no, you -- and I mentioned in my commentary, I think you get rate on rate as well. So long as our loss cost assumptions are right and we don't get any material increases in reinsurance costs and so on, then, yes. But as you know, it's not quite linear, Nigel. It doesn't always go. You can't do 1 minus the other creates a margin.But I think what we're saying to you guys, we're not giving you guidance, but what we are giving you is better data points around where the margin -- not exactly what it is going to be but how much this could be good confidence that the margin is improving better than it probably ever has done in the last 10 years.
Fair enough. And again, that's clear. And then maybe just finally, just a smaller point. But on -- I noticed that in the COVID provision, you've not made any allowance for U.K. business interruption losses in '21, where I understand you don't have any reinsurance cover. Is that because you are confident that you had reworded all the policies by first of January? Or can you say anything about that?
Well, we're about 75% excluded in the U.K. by the first of January. But I think the more important point is, our view this has been a continuation of a single event. And most of our policies are these policies and virtually all of them have a short indemnity period like 3 months. So on that basis, they can't exactly have an extended indemnity period. So it doesn't necessarily get them into the second lockdown unless it's a second event.What I can tell you by the time the second lockdown started at the beginning of November, we were 67% excluded at the beginning of November, and we were 75% excluded at the end of December and 85% excluded as of today. So if they could come back again, if they had 1 claim and use their limit and try it again, it's almost impossible for them. Nigel, if there's any follow-up, you're on mute.
Andrei at Morgan Stanley again. Over to you, Andrei.
I wanted to ask 2 questions. First one, just about the rate increases by region. If we have a look at the attritional improvements in FY '20, international saw about 4% attritional improvement versus only about 2.5% in America. And yet, the rate increases are running much higher international and America. So does that mean that you need to reaccelerate America to get the returns that you want?
Look, we'd always like to get more rate in North America. It depends on business mix and so on. But we have seen the most dramatic rate increases in our London Market International business. And that's where the really material reduction in attrition comes because we've had such major rate increases, higher than anywhere else in the organization.And we have -- we write quite a lot -- and we've always written quite a lot of high-risk business in that market. So that's why you would see that attrition improve materially. And actually also, the U.K. book is benefiting from many years of significant rate increase in commercial motor, which were kicked on because of Ogden and PPOs and so on. So we've seen very good improvements in our commercial motor book in the U.K. because of those 2 or 3 years of very good rate increases rather than just 1 year.
If you can still hear me okay, just want to ask my second question, kind of a broader question, but how do you think the London market will evolve post COVID, given the losses and also given the learnings on underwriting?
London market, well, it's difficult because we actually haven't got much COVID exposure in our London market business, our Lloyd's business because we don't write contingency. And a lot of the other businesses. The real exposure to COVID in the U.K. is pretty much in the U.K. market, the traditional U.K. players because it's very much an SME, mid-market play, which is not really Lloyd's.Yes, I think, look, everyone's excluded it. We started excluding COVID and various other diseases in March. We're putting for rate. And actually, the other thing we're doing is we're changing our wordings because if the court is going to interpret our wordings in that way, then we need to be a lot stricter on the cover within our offer because we can't rely on the courts upholding what we think the policy should be covering.
Our next question is Sid at JPM. Go ahead, Sid.
I had a couple of questions, if I can, please. The first is just on your guidance. In your market release, you say that you expect margin expansion on the 95% underlying exit COR for FY '20. I just wanted to be sure, that does include the $130 million of COVID losses you guide to as well? Just a point of clarification on that.
The COVID's on top of that.
Yes. So that doesn't.
Right. Okay, okay. Fair enough. Okay. If I could just ask another question just about your U.S. operations. If you could just make some comments just about what you'd like to see to ensure commitment to this business. I suppose we've seen quite a lot of reserve additions for many years. Maybe you could just give us what your thoughts are in terms of commitment to this and also just what kind of time frame you're giving to see a reasonable set of numbers to come from that division?
Yes. Well, look, Sid, we haven't done a very good job running the North American business for several years, so we're acutely aware that we need to make progress and more quickly and more sustainably in the past. There's -- so number one, we've made sure that the plan and the loss picks are sensible. We're targeting our growth into areas where we can control so we brought in a new team from Berkshire Hathaway to build up the financial lines business.We're very exposed to weather, obviously, with crop and the program cat business so we need to make sure that we have some balance. The team is also building out a retail business, which is more the middle market, more the heartland of the Travelers and the Hartfords and the Allstates, which is a massive market and there's opportunity for us to penetrate that, which is a more predictable, less cat-exposed -- it is cat-exposed but less cat-exposed than other sectors. And we're making a nice start there.But we're going to stick at what we're good at, where we're good at the specialty programs on the cat and on the casualty. We're constantly looking at how we can refine the portfolio and where we can grow and where we should shrink. The crop business, we still think, has long-term good prospects but we can't be over-reliant on it. But it's going to be ruthless execution around underwriting discipline, pricing focus. And when Inder talks about cost, there's clearly cost work we need to do in North America because we've got to pull that lever as much as we've got the pull the underwriting lever.
And just one final question for me. Just on your recoveries on business interruption. I was just keen to know, have you actually started paying any claims for BI post the U.K. FCA test case? And if so, have you actually put any of those claims to your reinsurers and have they accepted those claims?
Well, we were paying claims beforehand because there were some of our business, which we accepted had coverage before the test case, even before the High Court, not the Supreme Court. So yes, we are paying claims but we haven't got the level of attachment with our reinsurers yet, so we're not in a position to actually submit a recovery to our reinsurers.
Our next question is Matt Ingram at Bloomberg. Go ahead, Matt.
Just a couple of interest rate-related questions. The first is on the Tier 2 and Tier 1 debt that you've got out there. Obviously, it's quite high cost, given where rates were when you wrote the debt. I just wonder what your likely renewal costs would be, if you could give us any thoughts on that. Clearly, rates have stepped down a lot. And whether you're obviously in the position to call those when they come first due.The second question, you put 1.75% out there last year as a reduced investment yield target. Richard, I know you didn't specifically give any guidance for this year but I wondered if you could comment on whether that still stands.
I'll let Inder cover that one.
Yes. Look, on the debt cost, you've seen what we've done over the last 3 years or so, right? We've been -- where it makes sense, we're refinancing at current rates, which are obviously meaningfully lower. So as the calls come due and as we find sensible opportunities to manage our liability profile, we will do that, right? Obviously, the contra to the fact that these are high-yielding legacy instruments means that they're also trading at reasonably healthy premiums, right? So we're just going to make sure we're sensible from a shareholder point of view as we think about the overall cost of debt and also obviously, the overall weighted average cost of capital.I think on the investment yield, we've sort of given you the construct around the exit running yield on the fixed income book. We've also given you some steer as to where we think, if you include some sensible assumptions on our portfolio of risk assets, that gets you to about 90 basis points, which is kind of the exit. Now that doesn't include additional opportunities to rerisk. We said in the past that we could rerisk to around 15% risk assets over time, but that isn't a near-term focus for us, just given the opportunities, as Richard referenced in his opening -- in his remarks that there's opportunities for us on the underwriting side to deploy capital and lock in long-term sustainable margins. So that's really a priority for us.
Sorry. It keeps muting itself. So that 90 bps included any kind of mark-to-market or whatever from risk assets? I guess, I just want to clarify that.
Yes. Yes, it does, yes.
We're down to the last 2 questions. Next question, Brett Le Mesurier from Velocity. Over to you, Brett.
Inder, you were talking about increasing the reserves in the North American business for the 2020 underwriting year. But when I have a look at the claims development side, I see that the original estimate for ultimate claims is the same for 2019, 2020 were pretty much the same, only $20 million apart. I wonder if you can reconcile that observation with the comments you made before in the reserving of North America.
Yes. Brett, I think the simple answer is the reported incurred claims, so the case reserving, et cetera, we're actually seeing meaningfully less activity, right? So the loss activity is down a lot, and therefore, we're picking additional IBNR, right? And when we look at where the ultimates for '15, '16, '17, '18, '19 have ended up, about 2020 is at least, if not higher than that, right, given that the business has shrunk over time and we've re-underwritten and the mix has got better.
Can you also comment on why the 2019 underwriting year is the worst development? 2018 doesn't look so bad. You increased the reserves last year, but 2019 increase in reserves is quite significant. What's different between the 2018 and 2019 underwriting year?
Yes. Look, I think the $200-odd million deterioration on 2019, if you look at the aviation loss I mentioned, that's around a $50-odd million all in for that particular loss. We've got some of that E&S business I referenced. That was very early 2019 accident year business, some of that so that's a contributor to that.Obviously, we had some cat events. I mentioned Dorian, et cetera, that -- and the Japanese typhoons have deteriorated a little bit. So there's elements. We're disappointed with that deterioration, Brett, but those are the main drivers. Obviously, '18 has held up comparatively better.
And the final question is from Kieren Chidgey again at Jarden. Over to you, Kieren.
Just had 2 very quick follow-up questions. Inder, what is the outlook for the tax rate going forward?
Yes. Look, when you look at the blended tax rate for the company, Kieren, it's, in the jurisdictions we operated, it sort of rounds to about 23%, 24%, maybe a shade higher. We've obviously, given the loss situation, accumulated off-balance sheet tax attributes so we'll roll some of those in, hopefully, as we go through 2021, assuming the profitability emerges in the right areas. And so we're thinking kind of 20% is probably a fair planning assumption for 2021, a bit lower than our 23%, 24%, 25%, call it, sort of blended rate. So there's a bit of a -- short term, it's a bit lower; long term probably reverting back to that sort of low to mid-20s range.
That's all the questions for this morning. Now I'll hand back to you, Richard, to close.
Thank you, Tony. Thank you, everybody, for joining today's session. And I hope you all have a good day and look forward to speaking to you individually soon. Thank you very much.