Description

Long ALL

OVERVIEW.  We would recommend building a long position in ALL.  The key debate for ALL coming out of their 2Q print is currently around the cadence and magnitude of recovery in their personal auto insurance margins.  The bears are concerned about the potential for further reserve charges and margin deterioration on the back of soft 1H results in personal auto, but we are constructive on ALL because we suspect that their 2Q auto margins may be at or near trough levels, and we see a path to normalized profitability by early 2024.  If our thesis is correct, then we see potential for ~30% base case upside over the NTM (or $170/sh) supported by both positive revisions and re-rate as investors see a cleaner roadmap back to normalized earnings power.   

For brief background on the situation, ALL has been under-earning in their U.S. personal auto business since mid-2021, primarily as a result of elevated loss severity driven by the industrywide spike in used car prices.  ALL's stock had recently sold off following their disappointing 2Q22 preannouncement, which had revealed that the inflationary headwinds to their auto margins would be more severe, and the recovery timeline more protracted, than many investors had originally anticipated.  These results had suggested that ALL should continue to under-earn well into 2023 – although we suspect this is generally well-understood within the investor community by now, despite not yet having been fully baked into sellside estimates (we are still -15% below bbg cons for 2023 eps).

But looking past the headline miss, ALL’s 2Q results have made us more comfortable that their auto margins could be at or near trough levels, with a reasonable outlook to normalized earnings by 2024.  (i) We believe ALL has built up a significant cushion in their personal auto reserves over 1H22, which reduces the likelihood of future adverse development and puts them in a better position to move past their recent margin issues.  For example, our reserve analysis suggests that ALL’s reserves may now be in a similar position as PGR’s, which is significant because PGR’s auto margins have already troughed and their reserves are considered best-in-class, with some likelihood of proving redundant over the coming years.  (ii) Macro indicators for used car pricing and broader core inflation have recently begun to roll over, which should provide a tailwind for auto margins into 3Q.  (iii) ALL's first cohort of implemented rate actions will finally begin to translate into more material margin expansion in 3Q22, and the pace of margin expansion should further accelerate into 4Q22 and 2023 as later cohorts of implemented rates earn into the P&L as well, which will altogether provide another meaningful margin tailwind.  (iv) We have also become more comfortable with ALL’s latest loss trend assumptions, and putting this altogether, our latest rate-vs-trend math suggests that ALL can return to normalized auto margins by early 2024.  

Our base case price target of $170/sh is based on our 2024 eps estimate of $16.00 (or msd% positive revisions vs bbg cons of 15.18) capitalized at 10.5x PE.  Our positive revisions are supported by the normalization in auto margins, the roll-through of higher new money rates into their investment portfolio, and ALL's recently announced cost takeout initiatives for their G&A and loss adjustment expenses.  We believe 10-11x PE is an appropriate multiple for normalized earnings based on (i) ALL's historical trading range, adjusted for the divestiture of the majority of their lower-multiple life & annuity business in 2021; (ii) comparison to peers (PGR and KMPR have traded at 14x and 13x average through-the-cycle PEs, respectively); and (iii) our longer-term DCF of the business.  We also believe ALL's current risk/reward is skewed to the upside: our bear case valuation is $110/sh representing $12.00 of 2024 eps at 9.0x PE (or depressed earnings at a depressed multiple), while our bull case valuation is $190/sh, leading us to ~3x r/r for the stock.   

DISCUSSION.  First, we suspect ALL’s 2Q22 auto loss ratio should be near trough levels, with a path to normalized margins by early 2024.  We are only limiting our discussion to some of the key points below because the modeling for ALL’s reserves and margins can be somewhat technical.  

·     (i) We suspect ALL has recently built up a significant cushion in their personal auto reserves, which should put them in a better position to move past their margin issues.  As one example, the table below shows ALL’s average incurred, paid, and ultimate loss reserves-per-claim for the first 12 months of loss development within their auto property damage (PD) class over several accident years, which is a proxy for how much severity trend has been baked into their reserves over time.  Our takeaway is that ALL’s ultimate losses-per-claim for accident year 2021 (AY21) had originally looked deficient relative to peers, but after adjusting for ~$400m of reserve charges taken over 1H22, ALL’s auto PD reserves now look similarly conservative to PGR’s reserves, where margins have already troughed and reserves may prove redundant in the coming years.   

More specifically, one of the measures of reserve adequacy that we can observe is around how ALL’s AY21 first year ultimate reserve-per-claim has progressed relative to their pre-covid accident years, on both an absolute and relative basis.  For example, after adjusting for their 1H22 reserve development, ALL’s AY21 ultimate reserves-per-claim have increased by 25-30% since AY17-18, relative to a similar increase in incurred losses and only a 20% increase in paid losses-per-claim.  Similarly, PGR’s AY21 ultimate reserves-per-claim have increased 35-40% over the same period, relative to a 30% increase in paid losses-per-claim (for context, PGR’s paid losses have increased more than ALL’s because PGR has been shifting their mix towards higher-income, bundled policies).  In other words, after adjusting for ALL’s 1H22 reserve charges, there is a 5-10%pt delta between the paids- and ultimates-per-claim between the AY21 and pre-covid cohorts for both ALL and PGR, which suggests that both companies have built up a similar level of reserve cushion in auto PD. 

Tying this and a few other pieces altogether from our reserve analyses, we can observe several points that give us more comfort around ALL’s reserve position heading into 2H22.  (a) ALL’s original AY21 ultimate reserves-per-claim had screened as significantly weaker than PGR’s, which implies that ALL had been reflecting relatively little conservatism in their auto PD severity assumptions.  But following their 1H22 reserve charges of $400m, ALL’s AY21 reserves now seem to baking in a similar level of reserve cushion as PGR’s.  (b) We can also infer from both ALL’s auto PD and consolidated P&C paid-to-incurred trends (the former is shown below) that their 1H22 reserve charges were taken as a prospective reserve build, rather than a catchup to reflect any deterioration in paid loss trends; the former should be interpreted positively by shareholders, and the latter negatively.  This is a somewhat subtle but important point which provides a necessary foundation for our conclusion that ALL has been prospectively building reserves in (a).

And (c) the comparison to PGR is meaningful because PGR has built a strong balance sheet over the past year which has since led to strong margin results, and we suspect ALL’s margins should soon begin to follow a similar progression now that they have shored up their reserves.  For context, PGR had moved earlier than ALL in aggressively building reserves; PGR’s reserves and loss cost trend assumptions have generally since held up well over the last few months; PGR’s auto margins have already begun to inflect off of their cyclical lows; and we suspect PGR’s reserves could prove redundant over time, which means that they may gradually transition from under- to over-earning in auto.  

·     (ii) Macro indicators for used car pricing and broader core inflation have recently begun to roll over, which should be a tailwind for personal auto margins into 3Q.   Used car inflation is the most important consideration for ALL’s loss trend, as mgmt has previously estimated that higher used car prices have driven 80% of their severity headwinds since covid.  The Manheim Used Car and CarGurus Used Car price indices have recently stabilized in 2Q and into 3Q, and Manheim expects pricing to decline modestly by YE22 and more substantially in 2023 as supply chain improvements become more consistent.  Similarly, American Recycler data for scrap metal prices and used car components have peaked in 2Q and are rolling over in 3Q, almost universally across geographic regions.  Other broader measures of inflation have been moving slowly in ALL’s favor as well (i.e., commodities, housing, etc).  

·     (iii) ALL had begun to file for higher auto insurance pricing in late 2021, and these rate actions from 2021 will finally translate more meaningfully into margin expansion beginning in 3Q22 given the lengthy lead time between when rates increases are approved by state regulators, and when they are earned into the P&L.  From there, the pace of margin expansion should accelerate because the earn-in of rate actions from the 2022 and future cohorts will compound on top of the implemented rates from the 2021 cohort.  We estimate ~1pt of auto loss ratio improvement in 3Q from the earn-in of higher rates, which should build to ~3pts cumulatively by 2023 and ~4pts by 2024.  

·     (iv) Putting these factors together, our rate-vs-loss cost trend math suggests that ALL can achieve an auto combined ratio of approx. 103% in 2H22, 100% in 2023, and 96% in 2024, which would bring run-rate margins in-line with mgmt’s longer-term guidance of mid-90s CR.  Our base case margin assumptions reflect (a) 2.5% per quarter of average implemented rate increases through 2023, plus separate adjustments for CA and NY which run unique regulatory approval processes.  Our implemented rates flow through into earned pricing on a 6-9mo lagged basis; (b) 13.5% average incurred severity trend through 2022, based on their 2Q loss trend commentary plus a modest cushion which we assume to embed conservatism, followed by a modest decline in each of 2023 and 2024 to reflect inflation rolling over; and (c) average incurred frequency remaining approx. 90% of pre-covid levels, reflecting some sustainable frequency benefit from the shift to work-from-home. 

Our resulting rate-vs-trend math suggests that ALL could be at or near trough auto margins in 2Q.  We believe our rate-vs-trend math is directionally reliable based on a rough back-testing against ALL’s recent quarterly results.  For example, our base case assumptions above lead us to calculate an underlying loss ratio (ULR) of 80% in 2Q22 for their Allstate Brand division, which is consistent with the ~80% ULR that we suspect their 2Q results imply (for reference, ALL had stopped disclosing segment-level ULRs following their acquisition of National General in 2021, so we must make certain assumptions to derive segment-level margins). 

Notably, there was a meaningful catch-up component in ALL’s 2Q22 ULR to retrospectively reflect the losses which should have been incurred in 1Q22 based on mgmt’s updated view of loss cost trend for 2022.  After adjusting 2Q22 results for this catch-up component, we believe ALL had an ULR of ~78% in the quarter, which is in-line with our rate-vs-trend implied of ~79%.  Importantly, our rate-vs-trend framework leads us to calculate a ~79% ULR in both 3Q and 4Q for Allstate Brand.  This is sequentially flat vs our derived 79% in 2Q, which implies that margins should stabilize in 2H if we are correct about ALL’s reserve quality, and our assumptions around rates, severity, and frequency trends are reasonable.  Of these assumptions, we believe that the largest source of uncertainty is around severity trend, and we can gain comfort around this factor because (a) we believe mgmt’s updated 2Q assumptions seem reasonably conservative on both an absolute and relative basis, and (b) as noted above, broader inflation metrics have recently begun to roll over.  

Additionally, we would briefly call out several other factors which could contribute to positive revisions, or otherwise give us more comfort around ALL’s earnings profile and risk/reward.  

·     We see further upside to 2024 eps from the earn-through of higher new money yields into ALL’s fixed income investment portfolio, as well as their recently announced cost takeout initiatives.  With respect to their fixed income portfolio, we suspect ALL could be investing new credits at close to 4.00% yield compared to a 2Q22 book yield of 2.78%.  This could lead to at least +100bps of yield expansion by 2024 if ALL gradually turns over their credit portfolio over several years, and we think there could be further upside to yields because ALL has historically been more aggressive in turning over their credit book.  The sellside has generally been slow to reflect the tailwind from rising rates across the broader P&C universe, so NII upside is not a lever which is unique to ALL, but we believe it should nevertheless contribute to positive revisions over time. 

Additionally, ALL had recently introduced a new cost savings initiative to reduce their Property-Liability adjusted expense ratio from 26.0% to 23.0% by 2024.  This opportunity is difficult to fully dimension because part of the savings will come from loss adjustment expenses (LAE), which we and the street do not model separately due to lack of sufficient granularity in their financial disclosures.  However, we would expect the majority of the 3pts of planned cost savings to come from the expense ratio given the split between G&A and LAE, and we are only forecasting less than 1.5pts of ER improvement in our base case based on the historical cadence of margin improvement.  As such, we believe additional margin upside is possible relative to what we have baked into our estimates, especially given ALL’s strong track record around implementing cost savings, and each 100bps of ER improvements would add approx. $1.50 (+10% revision) to our 2024 eps.

·     Some bears have raised potential concerns around margin headwinds in ALL’s homeowners portfolio, but we do not see these concerns as having a significant impact on our earnings outlook for the outer years.  (a) First, ALL’s 2021 consolidated home ULR seemed elevated by ~5pts relative to their historical trend, but we suspect ~2.5pts were simply due to mix shift from a recent acquisition rather than margin headwinds, and the rest should partially be attributable to the inherent volatility in homeowners margins.  We believe the acquisition of National General in 1Q21 has structurally changed the ULR profile within homeowners, but this is not obvious because neither ALL nor NatGen have given granular financial disclosures, and the margin profile of NatGen vs legacy ALL requires some assumptions to back into.  (b) Second, we believe ALL’s results implied a 2Q22 home ULR of approx. 44% for Allstate Brand, which is in-line with the PYQ of 44%, so we do not see clear evidence of margin deterioration.  (c) Third, although ALL has not implemented much rate in home, the inflation kickers within their policies have effectively added high-teens % rates to help keep pace with replacement costs.  (d) And finally, we believe ALL has one of the best homeowners franchises in the U.S., with their strengths in pricing, underwriting, and reinsurance purchasing having underpinned above-average profits and below-average volatility over the last decade. 

·     We would expect ALL to be defensive in an economic slowdown because used car and raw material prices would likely decline further, which would accelerate the timeline of ALL’s margin recovery in auto.  Additionally, rising unemployment has historically been a tailwind for accident frequency, particularly within the nonstandard segment of the auto market.  Peers such as KMPR have similarly expressed that a recession would likely be a net tailwind for their businesses. 

VALUATION AND RISK/REWARD.

Base Case.  Our base case is $170/sh (+30% upside) representing 16.00 of 2024 eps at 10.5x PE, or approximately normalized earnings at a through-the-cycle multiple.  Our base case assumes that ALL can achieve an auto combined ratio of 103% in 2H22, 100% in 2023, and 96% in 2024.  Our bull case also assumes a 91% home CR in 2024, which is above their longer-term guidance for sub-90%.  This translates into a 94.2% Property Liability CR in 2024, which is roughly in-line with their 20yr average of 94.7%. 

We believe 10.5x Y3 PE is a reasonable through-the-cycle multiple.  ALL has historically traded at a 9.5x average Y3 multiple over the last 5, 10, and 20 years.  However, ALL also had a large life & annuities business which contributed ~20% of pre-tax earnings before corporate expenses over most of this period, until they divested a large portion of this business in 2021.  If we assume an average of 4-7x PE for the life business based on public life & annuities peers, this implies an average of 10-11x PE for the P&C units.  Our 10-11x PE range is also consistent with our DCF based on the present value of ALL’s longer-term capital returns.  

Bull Case.   Our bull case is $190/sh (+45% upside) representing ~$18.00 of 2024 eps at 10.5x PE, or cyclically elevated earnings at a through-the-cycle multiple.  Our bull case assumes that ALL can achieve an auto combined ratio of 103% in 2H22, 100% in 2023, and 95% in 2024.  Our 2024 auto CR is in-line with their longer-term historical average of ~95% in Allstate Brand, as well as their longer-term guidance for mid-90s CR.   Our bull case also assumes an 89% home combined ratio in 2024, which is consistent with their longer-term guidance for sub-90s CR.  This translates into a 93% CR for Property-Liability, which is below their 20yr average of 95%, but there have been many years in ALL’s history where they achieved a 93%-or-lower CR (most recently in 2017-2019).

Bear Case.  Our bear case is $110/sh (-15% downside) representing $12.00 of 2024 eps at 9.0x PE, or depressed earnings at a depressed multiple.  Our bear case assumes that ALL can achieve an auto CR of 104% in 2H, 104% in 2023, and 99% in 2024; and a 96% home CR in 2024.  Altogether, this translates into a 97% CR for Property-Liability, which is above their 20yr average of 95%.  Our 9.0x PE is roughly a bottom-decile multiple relative to ALL’s 5, 10, and 20 year averages, adjusted for their life & annuity sale.

Catalyst

Trough margins in personal auto insurance.  Over the near-term, the stock may move around ALL's monthly rate filing disclosures (ALL reports monthly results on the third Thursday of each month), as well as PGR's monthly results (PGR reports monthly earnings).  Inflation continuing to roll over.