全球保險(xiǎn)行業(yè)：全球凈資產(chǎn)收益率拆解2022

1、April 16, 2021 04:00 AM GMTInsuranceGlobal RoE Decomposition 2022MORGAN STANLEY & CO. INTERNATIONAL PLC+Jonathan Denham, CFAEQUITY ANALYST HYPERLINK mailto:Jonathan.Denham Jonathan.Denham+44 20 7425-3551Jon HockingNon-life insurers are the most defensive names in the sector owing to their lower asse

2、t leverage and reliance on underwriting profit. We repeat our analysis from last year andEQUITY ANALYST HYPERLINK mailto:Jon.Hocking Jon.HockingMORGAN STANLEY & CO. LLCMichael W. PhillipsEQUITY ANALYST HYPERLINK mailto:Michael.Phillips Michael.Phillips+44 20 7425-2307+1 212 761-3640break down the Ro

3、E of 36 insurers (across Europe, the Americas and Asia) with meaningful non-life exposure.MORGAN STANLEY MUFG SECURITIES CO., LTD.+Mia NagasakaEQUITY ANALYSTWe often find ourselves comparing the P/B and RoE for different non-life insurers across our global coverage. In this report, we explore a bit

4、further, to get a better picture of relative earnings quality. It is possible to generate the same RoE with different sources of earnings (Exhibit 2 and Exhibit 3). We decompose the RoE of insurers listed in Europe, the US, Australia, Japan, Hong Kong and Latin America into: non-life underwriting, n

5、on-life investing, life insurance and other. We show the drivers of the underwriting contribution (underwriting margin and underwriting leverage), as well as the drivers of the investment contributions (yield and investment leverage) Exhibit 1. HYPERLINK mailto:Mia.Nagasaka Mia.NagasakaMORGAN STANLE

6、Y AUSTRALIA LIMITED+Andrei Stadnik, FIAAEQUITY ANALYST HYPERLINK mailto:Andrei.Stadnik Andrei.StadnikMORGAN STANLEY ASIA LIMITED+Jenny Jiang, CFAEQUITY ANALYST HYPERLINK mailto:Jenny.Jiang Jenny.JiangMORGAN STANLEY & CO. LLCJorge KuriThe market is still rewarding insurers with a higher share of unde

7、rwriting profitwith a lower cost of equity (we assume no growth in our calculation, Exhibit 8). HYPERLINK mailto:Jorge.Kuri Jorge.KuriJorge Echevarria+1 212 761-6341The correlation between performance and share of underwriting and otherRESEARCH ASSOCIATE HYPERLINK mailto:Jorge.Echevarria Jorge.Echev

8、arria+1 212 761-8015profits is positive when looking at performance over the past 3 years or sinceYE19 (Exhibit 17 and Exhibit 20). However, the reverse is true, on average, YTD (Exhibit 19).The (simple) average expected 2022 RoE for the 36 insurers we analyse globally is 8.9%, or 12.8% when excludi

9、ng Lemonade and Root. The absolute RoEs and sources of earnings differ dramatically (Exhibit 4). Root has the lowest expected 2022 RoE (-95.3%) and Admiral the highest (40.7%).A high return on underwriting remains key to a high RoE. The three highest RoE insurers have a return on underwriting 20% wh

10、en including Admirals underwriting-like earnings (Exhibit 4). Two of these are in the top three by return on underwriting (Exhibit 10). Admiral is not in this group owing to its reinsurance structure. All three insurers have little or no life insurance earnings.Though there is a fairly strong correl

11、ation between P/B and RoE, there remain clear outliers. Overweight-rated names including AXA, Direct Line, Everest Re, Lancashire, PICC P&C and Sompo screen as undervalued, whereas Gjensidige (Underweight) looks overvalued (see Exhibit 5 and Exhibit 6). The regional sections have more details and ou

12、tline our respective investment theses.The Nordic insurers are among those with the lowest implied cost of equity, whereas the Chinese, Scor, Beazley, Sompo, Lancashire, AIG and AXA the highest (Exhibit 7). The higher the share of earnings from underwriting, the lower the cost of equity, on average

13、(Exhibit 8).We present a global non-life insurance comps sheet at the end of the note.InsuranceEuropeIndustryViewIn-LineMorgan Stanley does and seeks to do business with companies covered in Morgan Stanley Research. As a result, investors should be aware that the firm may have a conflict of interest

14、 that could affect the objectivity of Morgan Stanley Research. Investors should consider Morgan Stanley Research as only a single factor in making their investment decision.For analyst certification and other important disclosures, refer to the Disclosure Section, located at the end of this report.+

15、= Analysts employed by non-U.S. affiliates are not registered with FINRA, may not be associated persons of the member and may not be subject to FINRA restrictions on communications with a subject company, public appearances and trading securities held by a research analyst account.EQUITY ANALYST+81

16、3 6836-8406+61 2 9770-1684+852 2848-7152MethodologyWe decompose the RoE of 36 insurers under our global coverage that have meaningful non-life exposure. These include insurers listed in Europe, the US, Australia, Japan, Hong Kong and Latin America.We outline how we do this in Exhibit 1. We split the

17、 RoE into:non-life return on underwriting we additionally split this into the non-life underwriting margin (i.e. 1 - combined ratio) and the underwriting leverage (net earned premiums/equity).non-life return on investing we also split this into the investment yield and the investment leverage (non-l

18、ife financial assets/equity).life contributionother contribution e.g. central costs, retail income/reinsurance commissions for UK motor names, associates.taxExhibit 1: Our RoE decomposition methodUnderwriting contributionUnderwriting margina = 1-CORReturn on underwritingc = a*bUnderwriting leverageb

19、 = NEPs/shareholders equityInvestment contributionInvestment yielddReturn on investingf = d*ePre-tax RoEPost-tax RoEInvestment leveragei = c+f+g+hk = i*(1-j)e = financial assets/shareholders equityLife contributionEffective tax rateg = life earnings/shareholders equityjOther contributionFinancial le

20、verageh = other earnings/shareholders equityl = debt / debt + equitySource: Morgan Stanley ResearchWe look at 2022 earnings (based on 2022 combined ratio, investment yield, tax rate, etc.) on 2021 equity (as well as other 2021 balance sheet items, e.g. financial assets, debt).It is (clearly) possibl

21、e for insurers with very different business models to have the same RoE. However, differing earnings quality would suggest that these companies should not necessarily trade on the same P/B. We illustrate this through some theoretical examples in Exhibit 2 and Exhibit 3.Whilst 2. Higher leverage in E

22、xhibit 2 has the same return on underwriting and return on investing as our base example, the higher leverage means that it will be more impacted by changes in the combined ratio or investment yield.3. Higher yield in the chart illustrates the trade-off between investment income and underwriting pro

23、fit. For example, the period of low interest rates in Europe has caused the source of earnings for the Nordic non-life insurers to shift further towards underwriting profit and has led to a re-rating.Not all insurers in this analysis are pure non-life players: composites tend to have lower non-life

24、underwriting and asset leverage, with life insurance contributing to the RoE (see 4. Composite).An underwriter with a lower return on underwriting and lower investment return can make up this difference by being based in a low-tax regime (e.g., in 5. Low tax).Exhibit 2: The hypothetical insurers all

25、 have the same RoE but different sources of earnings we would expect them to trade on a different P/B multipleRoE decompositionTaxOther contributionLife contribution14%12%10%8%6%4%2%0%-2%-4%Return on investingReturn on underwritingPost-tax RoEBase2. Higher leverage3. Higher yield4. Composite5. Lowta

26、xSource: Morgan Stanley Research estimatesExhibit 3: We can see that examples 1 and 2 are actually different, despite looking the same in the prior exhibit1. Base2. Higherleverage3. Higheryield4. Composite5. Low taxUnderwriting margin8.0%4.0%4.0%8.0%3.6%Underwriting leverage75%150%75%38%150%Return o

27、n underwriting6.0%6.0%3.0%3.0%5.4%Investment yield4.0%2.0%6.0%4.0%1.8%Investment leverage150%300%150%75%300%Return on investing6.0%6.0%9.0%3.0%5.4%Life contribution0.0%0.0%0.0%6.0%0.0%Other contribution1.0%1.0%1.0%1.0%0.9%Pre-tax RoE13.0%13.0%13.0%13.0%11.7%Effective tax rate20.0%20.0%20.0%20.0%11.1

28、%Tax-2.6%-2.6%-2.6%-2.6%-1.3%Post-tax RoE10.4%10.4%10.4%10.4%10.4%Financial leverage20.0%20.0%20.0%20.0%20.0%Source: Morgan Stanley Research estimatesGlobal ConclusionsThe average expected 2022 RoE for the 36 insurers we analyse globally is 8.9%, or 12.8% excluding Lemonade and Root. The absolute Ro

29、Es and the sources of earnings differ dramatically (Exhibit 4). Admiral has the highest expected 2022 RoE (at 40.7%) and InsurTech companies Lemonade and Root the lowest at -17.4% and -95.3%, respectively.Whilst, as one would expect, there is a fairly strong correlation between price/book and RoE, t

30、here are clear outliers. Gjensidige (rated Underweight) looks overvalued, whereas AXA, Direct Line, Everest Re, Lancashire, PICC P&C and Sompo (all Overweight) look undervalued see Exhibit 5 and Exhibit 6, as well as sections Europe, Americas and Asia for more details and respective investment these

31、s.The Nordic insurers are among those with the lowest implied cost of equity, whereas the Chinese, Scor, Beazley, Sompo, Lancashire, AIG and AXA have the highest (Exhibit 7). The higher the share of earnings from underwriting, the lower the cost of equity, on average (Exhibit 8).A high return on und

32、erwriting is key to a high RoE. The three highest RoE insurers have a return on underwriting 20% when including Admirals underwriting-like earnings (Exhibit 4). Two of the three are in the top three by return on underwriting (Exhibit 10). Admiral is not in this group owing to its reinsurance structu

33、re it reinsures a large part of its business, generating a substantial part of its return through profit commissions.The 2022 RoE of the insurers included in our global analysis varies from -95.3% to 40.7%, on our estimates, with an average expected RoE of 8.9%, or 12.8% when excluding Lemonade and

34、Root (Exhibit 4).Nine insurers included in our analysis are expected to have an RoE 15% and three have an expected RoE 20%: Admiral (Underweight), Topdanmark (Equal-weight) and Gjensidige (Underweight), all of which have little or no life insurance business.Seven insurers have an expected RoE of les

35、s than 7%. Life insurance tends to be more important for these names, although we expect challengers Lemonade and Root to make large underwriting losses.European insurers in our analysis have the highest expected 2022 RoE (at 14.9%), with insurers in the Americas averaging 12.0% (excluding Lemonade

36、and Root) and Asian insurers averaging 9.1%.European insurers have the highest return on underwriting at 9.0%, closely followed by the Americas (ex-challengers) at 8.3%, with Asian insurers at just 4.6% on average.This is because Europes high expected underwriting margin offsets lower underwriting l

37、everage. Europe has an expected underwriting margin of 10.1%, the Americas at 6.8% (ex-challengers) and Asia at 4.0%. Underwriting leverage is higher in the Americas (ex- challengers) and Asia than Europe (NEPs 120% vs 95% of shareholders equity, respectively).A mix between long-tail and short-tail

38、business drives the level of asset leverage and thus the relative importance of investment returns.The insurers in the Americas (ex-challengers) generate the highest return on investing, on average, due to the higher asset leverage combined with a higher investment yield. In general, investment leve

39、rage is similar for the European and Asian insurers. However, the higher average investment yield means that the average insurer in Asia generates a higher return on investing than the Europeans included in our analysis.Return on non-life investment assets accounts for around half of PBT in the Amer

40、icas (ex-challengers) and Asia vs just around a quarter in Europe, on average.The European insurers may have had to raise premiums to compensate for low yields. Non-life underwriting is a larger component of earnings than non-life investment income for 15 of the 17 European insurers vs 5 of the 9 Am

41、erican insurers (ex-challengers) and just 3 of the 8 Asian insurers (just the Australians).Exhibit 4: Whilst the average 2022e RoE is 12.8% (excluding challengers Lemonade and Root), we estimate that 3 insurers in our analysis will have an RoE 20%ecomposition (2022e)TaxReturn on investingOther contr

42、ibutionLife contributionReturn on underwritingPost-tax RoERoE d55%50%45%40%35%30%25%20%15%10%5%0%-5%-10%China Re (UW)MS&AD (EW)Aviva (OW) AIG (OW)Suncorp (EW) Generali (EW) Axis (EW) QBE (OW) AXA (OW)Travelers (UW) Sompo (OW) Tokio Marine (OW) Allianz (EW) Swiss Re (EW) Munich Re (OW) IAG (EW)Scor (

43、EW) RenRe (EW) Everest Re (OW) Sampo (OW)W.R. Berkley (EW) Zurich (OW) PICC P&C (OW) Allstate (EW) Hiscox (OW) Direct Line (OW) Lancashire (OW) Hannover Re (OW) Porto Seguro (UW) Beazley (EW) Progressive (EW) Gjensidige (UW) Topdanmark (EW) Admiral (UW)-15%Source: Morgan Stanley Research estimatesAs

44、 one might expect, there is a fairly strong correlation between P/B and RoE; however, there are clear outliers (Exhibit 5). For example, Gjensidige (Underweight) looks overvalued, whereas PICC P&C and Lancashire (both Overweight) look undervalued.When zooming in, we can see other names that we view

45、as attractively valued, such as Overweight-rated AXA, Direct Line, Everest Re and Sompo (Exhibit 6). Whilst a simple P/B vs RoE regression is not enough to support a stock call, we outline our investment thesis for these stocks in their respective regional sections.Exhibit 5: As we would expect, the

46、re is a fairly strong correlation between P/B and RoE, though there are outliers, such as Gjensidige, PICC P&C and Lancashire2021e P/BOverweight Equal-weight UnderweightADMTOPGJFPGRSUN AGXTASOSARKLMEIVOURNRVNERALL PSQTBEV ZURN HNRIASGAWMRPBO DLGMS&ADCHINA REAIG SOXMAPSOCORHSXBEZAVLREPICC P&C9.08.07.

47、06.05.04.03.02.01.00.00%10%20%30%40%2022e RoEExhibit 6: and when we zoom in, we can also see other outliers, such as AXA, Direct Line, Everest Re and Sompo2021e P/BOverweight Equal-weight UnderweightTOPGJFPGRIAGSAMPOWRBTRVZURN HNR PSBEZSUN QBE ALVMUVRERNR HSXMS&ADCHINA REAVGTOKIOSRENAXADLGALLLRE SCO

48、RAIGSOMPOAXSPICC P&C4.54.03.53.02.52.01.51.00.50.00%4%8%12%16%20%2022e RoESource: Morgan Stanley Research estimatesSource: Morgan Stanley Research estimatesWhen looking at the implied cost of equity (2022e RoE divided by 2021e P/B, assuming no growth), these discrepancies become more apparent (Exhib

49、it 7). PICC P&C and China Re are trading at an implied cost of equity of around 20%, whereas Topdanmark, Gjensidige, Admiral, IAG and Progressive are 6% or below. Scor, Beazley, Sompo, Lancashire, AIG and AXA also have a relatively high implied cost of equity.Exhibit 7: The Nordic insurers are among

50、 those with the lowest implied cost of equity, whereas Chinese, Scor, Beazley, Sompo, Lancashire, AIG and AXA have among the highestImplied cost of equity 25%20%15%10%5%China Re (UW) PICC P&C (OW)Scor (EW) Beazley (EW) Sompo (OW) Lancashire (OW) AIG (OW)AXA (OW)RenRe (EW) Everest Re (OW) MS&AD (EW)A

51、llstate (EW) Hannover Re (OW) Generali (EW) Allianz (EW)Direct Line (OW) Tokio Marine (OW) Porto Seguro (UW) Axis (EW)Munich Re (OW) Swiss Re (EW) Hiscox (OW) Aviva (OW) Zurich (OW) QBE (OW)Travelers (UW) Suncorp (EW) Sampo (OW)W.R. Berkley (EW) Progressive (EW) IAG (EW)Admiral (UW) Gjensidige (UW)

52、Topdanmark (EW)0%Source: Morgan Stanley Research estimatesThe higher the share of earnings from underwriting earnings, the lower the cost of equity, on average (Exhibit 8). In certain cases, underwriting earnings are seen as higher quality and more sustainable than other sources of earnings. For exa

53、mple, the Nordic insurers trade at a premium, since a large proportion of their earnings comes from underwriting profits generated in oligopolistic markets with high barriers to entry andhigh customer retention. This source of earnings is less subject to market performance than life insurance or non

54、-life investment income.Exhibit 8 is before adding in other underwriting-like sources of earnings. For example, underwriting accounts for just 12% of Admirals PBT and the company has just a 5.2% cost of equity. Were we to include other items including reinsurance commissions, commutations and retail

55、 income, this would be 95% of earnings, making it look less expensive on this basis (Exhibit 9).Exhibit 8: The higher the share of earnings from underwriting earnings, the lower the cost of equity, on averageExhibit 9: this remains the case when we include other sources of earnings, such as retail i

56、ncome, profit commissions and central costsImplied cost of equity25%Overweight Equal-weight UnderweightImplied cost of equity25%CHINA RE20%15%PICC P&CSOMPOSCORBEZCHINA RE20%15%SCORPICC P&CMS&ADAIGAXAGLRERNRAIGMS&ADSOMPORNRBEZLREAXSALVRE ALLGAXATOKIO RE ALL10%AV PSHSXQBEAV 10%AXSHNRHSXDLGTOKIOSREN HN

57、R DLGWRBSRENMUVALVPSADM5%ZURNMUV TRVSAMPOSUNPGRTOP GJFIAGOverweight5%TRVWRBZURNQBESAMPO PGRSUNIAGADM0%-60% -40% -20%0%20%40%60%80% 100%Equal-weight Underweight0%GJFTOPSource: Morgan Stanley Research estimatesUnderwriting share of PBT (2022e)-60% -40% -20%0%20%40%60%80% 100%Underwriting and other sha

58、re of PBT (2022e)Source: Morgan Stanley Research estimatesUnderwriting is the key driver of superior RoEsAll of the three highest RoE insurers all have a return on underwriting 20% when including Admirals underwriting-like earnings (Exhibit 4).Two of the three (Topdanmark and Gjensidige) are in the

59、top three by return on underwriting (Exhibit 10). Admiral is not in this group, but this is a function of its reinsurance structure. The company reinsures most of its UK motor business, and thus generates a large part of its return through profit commissions and commutations.Interestingly, Progressi

60、ve (which has a relatively high return on underwriting) does not have a superior combined ratio when compared to the overall group: it makes its superior return on underwriting through being able to run with the highest underwriting leverage (its net earned premiums are 2.5x its shareholders equity)