Description

Permanent TSB (IL0A ID): long equity, target price of 3.71 per share (+95%)

Permanent TSB is an Irish retail community bank in the final stages of a multiyear restructuring process. The former institution, Irish Life & Permanent, was nationalized by the Irish government in 2011 after the bank failed as a result of bad commercial real estate loans like nearly every other Irish bank at the time. Post nationalization, the government brought in a new management team, incurred over €3bn in loan loss provisions, and right sized the cost structure. In early 2015 the government IPO-ed a 25% stake, which subsequently experienced a 60% drop in share price due to a litany of investor concerns:  fears around an inability to dispose of legacy noncore UK assets, fears around the performance of Irish CRE and residential NPLs, fears around the impact of Brexit on the overall Irish economy, the bank’s dependence on wholesale funding, and increased regulatory costs. However, in the two years since becoming public the company successfully completed its disposal program of commercial real estate and UK assets, normalized its funding base, and returned to growth. The last remaining issue is the topic of Irish residential NPLs. TSB has €5.8bn of Irish NPLs with €2.3bn of loan loss provisions (40% coverage on average). Unlike the traditional NPL definition, the majority of TSB’s NPLs are largely paying cash principal and interest (at least in part) and are classified as impaired or non-performing solely due to housing values and the fact that the loans have been restructured versus their original contract terms. In fact, the re-performing or “treated” book of NPLs is more profitable than performing tracker mortgages. I argue that PTSB’s stock could easily trade to 0.8x TBV (+95%) on simple blocking and tackling on NPL sales (e.g. reduction of uncertainty) and that tangible book value itself has upside in the event of successful NPL resolution. From a pattern recognition perspective, the investment is similar to long US subprime & Alt-A mortgages in 2012-2013 as the bottom is already in for the local housing market and the investment benefits from falling uncertainty around the performance of restructured / modified mortgages.

Thesis points:

• Huge potential for valuation normalization on NPL resolution as the Irish market starts clearing bad residential property loans in a similar fashion to the US RTC episode in the early 1990s, US subprime assets post GFC, and Irish commercial property loans in 2012. The precedent case studies would indicate that the initial sales yield attractive IRRs to the buyers, which brings in more capital to the opportunity, compressing returns on purchases to the benefit of the NPL sellers (e.g. to PTSB). The initial residential NPL sales in the market were in 2013-2014 to private equity firms such as Lone Star, who subsequently have recouped their investment via sales into the securitization market in 2016 & 2017.

• PTSB has 5800mm of Irish residential NPLs. 53% of the book has been restructured, with 93% reperformance rate and 3.6% cash yield on the loans. 47% are still to be restructured but still generate a 2.0% cash yield.

• Provisions on the NPLs include forced sale discounts of 15%, as well as workout costs of 5%. Furthermore, provisions assume a 45% peak-trough decline in home prices versus an actual current peak-trough decline in home prices of ~30% (roughly a 15% buffer on home prices versus the home price indices). On the whole, the loans mark the residential properties at a 35% discount to values implied by the government’s CSO home price index.

• AIB recently sold a portfolio of 400mm Irish buy-to-let mortgages (BTL) to Goldman for 50% of par – PTSB carries a similar book at 47.5% of par. Sales of mortgages to reduce the headline NPL ratio and free up risk weighted assets seem highly likely (http://www.irishtimes.com/business/financial-services/aib-sells-non-performing-buy-to-let-loans-to-goldman-sachs-for-200m-1.3048092) – “Dubbed Project Cypress, the tender process for the portfolio sale attracted strong interest from a number of potential purchasers. This is thought to have included Cerberus, the US private equity fund that has acquired a large chunk of loans and assets from Nama in recent years.” http://www.independent.ie/business/irish/ptsb-may-launch-1bn-loan-sale-as-early-as-the-end-of-this-year-35974900.html

• Lone Star securitized a portfolio of 100% NPLs in the buy-to-let segment in April 2017 at a 50% advance rate (ERLS 2017 NPL1 Mtge). If private equity can buy PTSB’s BTL book at 50% of par, then securitize out their entire cost base (e.g. given the 50% advance rate) to keep all upside on home prices and mortgage reperformance for free, then arguably the BTL book will clear at higher prices than 50%. The main difference between now and the initial loan sales in 2014 is that the securitization market is willing to afford leverage on these assets, allowing private equity to reach mid-high levered teens returns without extreme discounts to par. The securitization market is open to Irish NPLs because they generate cash income, contrary to the traditional conception of what “non-performing” means. See also ERLS 2016-1, which is 33% NPL, 67% re-performing

• Mortgage-to-rent scheme introduced by the Irish government in 2017 has the potential to clear out the supply of “untreated” mortgages in the owner-occupied segment. In this program, the bank sells the mortgage loan to a private fund who receives ownership of the property. The local municipal housing authority enters into an agreement with the private fund and pays them a quasi-sovereign income stream and deals with the tenant directly.  The MTR scheme thereby transfers a distressed property loan into a de facto social housing asset, with the tenant being able to wipe out his secured debt and remain in his own home, dealing with his local housing counsel as a landlord rather than the private fund. The difference between the rent paid to the private fund and the rent paid by the tenant to the housing association would be a straight fiscal transfer / social program. This will allow banks to unload distressed loans and release the forced sale discount and foreclosure cost discount associated with the provisions. http://bizplus.ie/arizun-jumps-mortgage-rent-scheme/

• Irish home prices are rising around ~11% per year and are the primary determinant of the loan loss provisions since every restructured loan is classified as NPL (e.g. with a 100% probability of default). Irish homebuilders like Cairn (CRN LN) trade at 1.5x NAV because of market expectations around embedded profits in their land bank due to home price appreciation through 2020 – although, if you want to get long Irish home prices, PTSB is the direct way to do it given its provisions will be directly reduced by rising home prices. The contrast between the homebuilders and PTSB could not be more severe – for the one group the market extrapolates home price increases multiple years into the future while for the other it fears that they cannot unload housing loans at any price. Furthermore, the contrast between PTSB’s own debt and its equity could not be more severe. The AT1 bonds trade at 105% of par, the equity at 40% of tangible book.

• Huge potential for normalization of the Irish residential mortgage origination market (loan volumes). 2006 mortgage loan disbursements were €40Bn compared to 2016 levels of €5.6Bn (-86%). As a local Irish thrift, PTSB is a pure play on increased mortgage originations in a consolidated market with 50% less players today than in the last cycle. PTSB should have significant operating leverage on increasing loan origination volumes given that the marginal cost to originate is de minimus, its operating costs are stable, and the credit quality of the new loans is pristine.

• Resumption of the dividend in 2019, based on 2018 numbers (post NPL sales, if we are right on the prices achieved)

• Massive asset-sensitivity to rising ECB interest rates (the majority of PTSB’s mortgages are “tracker”, which means floating rate loans tied to the ECB base rate). Given its deposit funding and floating rate assets, PTSB is one of the most asset sensitive banks in the world.

• Significant excess capital in the event NPLs are sold near current marks.

• Potential acquisition of the company by KBC for further consolidation of the Irish market, as has been repeatedly discussed in the Irish press off and on over the last few years

• http://www.independent.ie/business/kbc-bank-confirms-it-will-stay-in-ireland-and-will-look-at-possible-acquisitions-35436650.html

• http://www.thetimes.co.uk/article/kbc-eyes-merger-with-permanent-tsb-nnwmx2lfn

Bull Case scenario: No writebacks but PTSB trades to 0.8x book value as the market perceives the company earning its cost of capital due to gradual dividend payments, growth of the mortgage loan book and operating leverage. Return of excess capital would greatly speed the process of generating a higher RoE post NPL dispositions (with 100% risk weighting, NPL sales quickly decrease the denominator of the required capital ratio). The company will have 400-500mm of excess capital compared to the current market capitalization of ~900mm if our NPL resolution thesis is correct.

• 2017 TCE of 2112mm x 0.8x TCE multiple = 1689mm (3.71 per share, +91.5%)

Bear risk scenario: PTSB takes a large hit on NPL wind-downs of 500mm, and the market prices the stock at 0.4x TBV – where most banks bottomed out in core Europe during the sovereign crisis in 2011. Note, we are not assuming the repeat of the Irish sovereign crisis given the significantly improved state of government finances (NAMA bond repayment, BKIR repayment, AIB & PTSB Coco / pref repayment and partial equity repayment). Essentially this scenario would be PTSB sells loans below the values that could be advanced in the securitization market, in the process taking a large writedown, followed by no mortgage loan growth, no interest rate hikes, no excess capital, no operating leverage, forever (e.g. 4% ROTCE in perpetuity). I would note the company already trades at 0.4x on current TBV, but in our estimation this is the market already attempting to price in a large degree of losses on NPL resolution; the bear case is that the losses are worse than feared and the company/market/economy/interest rates never recover. This degree of losses would NOT cause the company to violate capital ratios or face the prospect of bail-in; capital ratios are 17.1% transitional and 15.0% fully loaded. For those who would argue that the bank could be politically bailed-in "just because", note also that the Irish government is pari passu to the equity position with a 75% ownership stake.

• 2017 TCE of 2112mm less 500mm = 1612mm

• 1612mm x 0.4x TCE multiple = 645mm (1.42 per share, or -23%)

• For sake of conservatism, this trading level is below the post-Brexit selloff when the market worried that the company would never be able to sell its non-core UK assets (it sold them at their guide price a few months later)

Uber-bull case scenario: In an environment of rising European rates, PTSB is one of the most asset sensitive banks in the world given 90%+ of its loan book is tied to Euribor floating rates. In this scenario we model in 200mm of writebacks on NPL disposition, given the 47.5% starting price for the BTL book and 50% assumed take-out price, 20-40mm annual writebacks from the treated book, and ~275mm of arguably excess IBNR on currently performing mortgages. 1.5x TBV is where US community banks traded to when the rising rate thesis began to first take hold, despite them having RoEs well below the cost of capital at the time.

• 2017 ending TCE of 2112mm + 89mm consensus net income in 2018 + 200mm writebacks = 2401mm

• 2401 x 1.5x = 3601mm (7.92 per share, +317%)

• EDSF would indicate that Euribor rate increases start in 2H 2019, hence the valuation at end 2018 TCE

Business description:

• Permanent TSB Group is an Irish retail bank that was nationalized during the Irish sovereign crisis during 2011. The central government still owns 75% of the shares outstanding and will likely sell them around 4.50 per share, similar to the initial IPO price done in May 2015.

• Permanent TSB has undergone a very long deleveraging process, selling all of its noncore loans in 2014-2016 in the Irish CRE lending and UK residential mortgage lending segments

• Consolidating Irish Banking market:

• Compared to 2008, the Irish banking market has consolidated down from 12 to 5 players: BKIR, AIB, Ulster, KBC ireland, and TSB

• Anglo Irish Bank went into liquidation in 2013 after the government receiver took over the assets. EBS and Irish Nationwide were merged into larger institutions and wound down.

• Lloyds (Bank of Scotland Ireland) withdrew from the market completely, as did Rabobank (ACC bank subsidiary) and Danske (National Irish bank)

• TSB built very large provisions against its stock of Irish residential mortgages during 2011-2014 given the fall in home prices. TSB assumed a 50% peak-trough decline for its provision coverage model, and as Irish home prices continue to rise the case can be made for gradually releasing those reserves. The central bank of Ireland eliminated any cure rate assumptions related to Irish residential mortgage home loans, so TSB still has to carry reserves and high RWAs against modified & currently performing loans. Disposition of those loans as part of a broader NPL workout could release large amounts of capital

What happened? At the time of IPO in May 2015, the selling stakeholder (the Irish government) had a story around a re-rating to tangible book value with a 10% ROE and resumption of dividends in 2018, based largely around a cost of funds re-rating (deposit pricing falling) and falling complexity of the bank as noncore assets were sold. However, almost immediately after the IPO at 4.50 per share the stock tanked due to a combination of political and regulatory interventions as well as faster money investors exiting the stock as the re-rating story appeared delayed.

• Regulatory cost burdens:

• Banking sector levy: The Irish banking sector levy is a tax imposed only on financial institutions based off their respective deposit market share as of 2011. The original tax was supposed to only exist between 2014-2016 as a point in time contribution to the Irish economic recovery. At the time of the IPO the government promised that the tax of 27mm to TSB would end in 2017, and shortly after the deal was complete they extended the tax to 2021. Given changes in the respective sizes of the Irish banks since 2011, TSB’s share of the tax has fallen to 24mm, which we expect to continue into perpetuity regardless of what the government currently says.

• European Central Bank regulatory taxes: Since the time of the IPO the ECB has imposed a tax on all European financial institutions related to the creation of the Single Resolution Fund (centralized bank recapitalization facility) and the Deposit Guarantee Fund. PTSB’s expenses related to European regulatory issues moved up from 8mm per year to 24mm, which is notably primarily in the context of PTSB’s relatively low pre-tax profit numbers in 2016/2017 anyway (the burden falls proportionately over time due to mortgage loan growth and operating leverage)

• ECB TRIM model review – Post the 2011/2012 banking sector crisis the ECB decided that it no longer believes any individual bank’s NPL methodology or risk-weightings – for good reason. Hence the ECB has installed auditors in each of the banks to oversee their NPL recognition methodologies and assign “appropriate” risk weights as a result. PTSB noted that it expects required risk-weighted assets to increase by ~2.8bn in 2017 simply due to higher risk weightings on existing exposures (actual assets do not increase, but instead are falling), furthering the desire for them to sell or otherwise transfer assets off balance sheet. The market assumes that these increased risk weightings will derail the excess capital story, without giving credit for the RWA reduction that comes from selling NPLs at 100% RWA. AIB and BIRG do not disclose the impact of TRIM on their capital ratios, so the market has chosen to fixate on PTSB.

• Central bank “macro-prudential” lending limits on mortgages: right after the IPO, the central bank imposed limits on debt/income for new mortgage loans, immediately curtailing the supply of credit in the overall mortgage market. The central bank subsequently relaxed these restrictions in

• Mortgage tracker settlement / redress: In mid 2015 TSB made 35mm of redress payments to holders of tracker mortgages. Some mortgage holders had the ability to fix their interest rate for a period of time and revert to the tracker rate at the end of the fixed period. However, they exited their fixed mortgage early and when they tried to claim the right to the lower tracker rate option when the ECB cut rates, TSB said no. Lawsuits followed, which took the case to High Court. PTSB finished its tracker review in 1H’17, although at the cost of having to delay its mortgage modification activities (e.g. you cannot foreclose on homes when your foreclosure process is being investigated by the government). Now that the review is over, loan modification and NPL resolution will proceed apace in 2H’17

• Political pressure to reduce mortgage rates: During 2014-2015 there was intense political pressure applied by the Finance Ministry to the Irish banking sector to proactively lower standard variable rate (SVR) mortgage cost to consumers. As 100% owner of AIB, the government initiated a mortgage price war with AIB taking 35% of the new mortgage origination market in 2015. After early jostling amongst the banks and a slew of new production introductions, by mid 2017 pricing pressure seems to have abated with relative stability in the market, although this is a persistent and valid concern amongst investors. However, given that the ECB is requiring increased risk weightings on performing Irish mortgages as part of TRIM vs all other European countries, it would seem reasonable that pricing remain higher for Irish mortgages to earn a reasonable return on RWAs (e.g. if the ECB dictates that Irish mortgages require more capital than all other European countries given the legacy of the crisis in Ireland, then it would be predictable for Irish lenders to charge higher rates for those same mortgages than is charged in other European countries).

Risks:

• PTSB sells its NPLs at a loss – the focus here would be the 1700mm book of “untreated owner-occupied” mortgages, carried around 60% of par. It’s always easy to say that private equity will only make very discounted bids for the loans given that PTSB is a “forced seller” except that the company could simply securitize this book and spin it out to shareholders around 50% of par, allowing current holders the opportunity to benefit from future HPI and mortgage modifications (instead of the 20% returns required by PE). Once again, the key distinction here versus other NPL situations is the cash yield being generated by the portfolio even in the absence of loan modification.

• Price competition on new mortgages – AIB has reportedly “looked at” 3% asset yields on new mortgages, down from ~4%. Small scale new entrants Pepper and Dilosk have attempted to enter the market on brokered mortgages with little success to date. Many sellside banks compare the asset yields on Irish mortgages versus those in other European countries and conclude that the spreads must fall. However, they fail to incorporate the significantly higher prospective risk weights of Irish mortgages per TRIM and longer reposession time lines, which justify the higher spreads. In essence, analysts double penalize PTSB in that they assume asset spreads most come down at the same time RWA weightings significantly rise. My view is that RWAs will increase to ~50% on performing residential mortgages per PTSB's TRIM guidance (already in current numbers), but that spreads remain the same.

• Government passes the Variable Rate Mortgage bill, and the central bank actually uses its power. The bill aims to regulate mortgage pricing in the standard variable rate segment (e.g. mortgages priced off bank’s own prime rate) by empowering the central bank to intervene in the event of “market failure”. In essence, this is a bill designed to give local politicians room to pressure the central bank to forcibly intervene if the politicians believe the bank prime rate is “too high” – the CBI has the power to cap the bank prime rate.

• The Irish finance minister, the ECB, and CBI itself all expressed significant reservations about the bill and do not want it passed, or used if it were passed.

• Brexit induced recession

NPLs:  There are a number of definitional issues at play with regard to NPL classification, which the market overlooks in entirety after losing faith in management teams during the last cycle (2011-2012). Irish banks have very high NPL ratios in the overall context of European banks, which is partially why they trade at lower multiples than the rest of the sector. However, these NPLs have come under intense scrutiny many times (the Irish PCAR, PLAR, Bank solvency exercise, IMF analysis, etc) and are in the final stages of being worked out – wherein lies the opportunity. The market will never “believe” bank management teams as long as the NPLs are classified as such, so the burden of proof falls on management to simply remove them off balance sheet if, in fact, they are conservatively provisioned for as I believe.

Definitions of NPLs across Irish banks: Bank of Ireland includes as NPLs all residential loans > 90 days in arrears, whether or not they are impaired. AIB’s definition is the same, except that they exclude loans > 90 days in arrears that are not impaired (e.g. recovery value > loan balance). PTSB includes all loans that are impaired, all loans that are > 90 days in arrears, and all loans that have bullet repayment schedules post restructuring (e.g. loans that are current but have been restructured). Hence PTSB shows the highest level of NPL / assets of any Irish bank simply due to definitional reasons alone.

Types of NPLs

•  Split mortgages – the entirety of the mortgage is classified as NPL until the 3 year review period has passed given it has been modified, even if the loan is current and paying. Given that some of the treated NPLs have some reduction in principal and some reduction in interest, combined with a bullet repayment at loan end, they remain classified as NPLs for the life of the loan. Others will cease to be classified as NPLs after the 3 year review term has expired

• Part-capital, part-interest: these are loan modifications where TSB restructures the principal and interest schedule to more affordable monthly payments. Given that the payments have been restructured and generally have a balloon maturity (e.g. 30%-40% of principal due at final date, which typically means refinancing at loan maturity) TSB must classify these loans as NPLs under the European Banking Authority definition even if they are performing on the new schedule.

• Any other type of treatment – temporary changes interest and principal schedule are removed from the NPL definition via cure if they meet 12 months of payments on the new schedule. The previous loan modification techniques incorporate permanent changes to the original loan terms and hence are permanently classified as NPL regardless of subsequent re-performance. Deferred / capitalization of  interest modifications, one-year interest only periods, temporary reductions in payments, term extensions, and other less substantial modifications can cure to performing if the borrower is current for 12 months (albeit with a greater propensity to redefault than the previously described modifications). PTSB has been following the optimal economic loan modification strategy as evidenced by re-performance rates, albeit a punitive one from a “NPL classification” perspective since the loans remain NPL forever.

What determines provisions on NPLs?

• Underlying Home Prices – all banks using peak-trough discounts as the short-hand for their home price assumption. The provision values were established in 2013, at the trough of the housing market in Ireland, when peak-trough discount was 55%. Since then, home prices have risen considerably, and are now at a 30% peak-trough discount.

• Forced sale discount (to the above) – PTSB assumes a 20%-35% discount to “underlying” home prices. The change from a 10-20% discount assumed in 2014 to a 20-35% discount assumed in 2015 drove a total impairment of 135mm, which more than offset provision release from cures.

• Workout costs (in addition to the above) – PTSB assumes another 5% of principal consumed in workout costs

• Time to Sell: PTSB assumes 6 years to sell owner-occupied homes, 4 years to sell BTL, given the extensive delays in court-restructuring processes. PTSB has been pursuing these mortgages since 2012, which is why there is a large number of the untreated NPLs at the end of the

• Cure rates – unknown, except that PTSB did build another 284mm of provisions in the last two years related to this metric in spite of evidence to the contrary

3100mm of treated NPLs – yielding 360bps, 93% performance rate on modified loans. These loans are actually more profitable than performing tracker mortgages and should generate an “annuity” of 20-40mm annual provision writebacks. The presence of these loans is helpful in generating operating leverage for the company but in a worst case scenario could be securitized and spun out to shareholders if politicians or regulators demanded it.

2700mm of untreated / un-restructured NPLs – cash yield of 200bps. There are 1000mm BTL in this group, and 1700mm owner-occupier. The BTL portfolio will most likely be sold in the next 3-6 months per the press article.

• 700mm is “closed”, which means that they have possession of the house, or have a court order for possession but are awaiting resolution of a title issue.

• 400mm are technically held, which means that they are paying full principal and interest (generally owner-occupier mortgages) but which are classified as NPL because the same person defaulted on a BTL mortgage with the same bank or a different BTL mortgage if there are multiple to the same borrower; since they defaulted on one of the loans, the ECB requires PTSB to classify the other mortgage as NPL regardless of actual payment status.

• 1500mm are in various states of treatment and workout. The 500mm of untreated BTL would generate a gain on sale if realized at the same price as AIB. The 1000mm of owner-occupied untreated mortgages is the most likely candidate for generating losses on sale.

• However, it is important to note that the 600mm treated BTL book would likely generate a substantial gain on sale if sold, the 2500mm of treated owner-occupied will generate small consistent annual writebacks and the company appears to have ~330mm of IBNR / collective reserves on legacy non-impaired Irish mortgages, which is multiples higher than any other European bank (and hence could probably be released as well).

• http://www.permanenttsbgroup.ie/~/media/Files/I/Irish-Life-And-Permanent/Attachments/pdf/2017/2016-annual-results-presentation.pdf? – slide 35

Side note:

IFRS 9 – this is a new accounting pronouncement that will come into effect on Jan 1 2018 for all European banks. Provision accounting moves from an incurred loss model to an expected loss model. Going forward all performing loans will have their expected loss provisioned each year, regardless of outcome. Loans with significant deterioration (e.g. delinquent or restructured) but which are not yet NPL will be hit with lifetime provisions, which is the large change. Non-performing loans will still carry lifetime provisions as per usual (e.g. PD is already 100%, loans should be carried at workout value).  For PTSB, they are already carry the industry’s most conservative definition of NPL since they include restructured loans that are currently performing (e.g. split loans and capital and interest loans), hence they should have minimal capital hit from the new accounting rules. PTSB has guided that since all restructured loans are already classified as NPL, this should not impact them unlike AIB and Bank of Ireland who classify many restructured loans as performing.

What happened in the last month?

PTSB reported 1H’17 results that were largely in line on current results (excluding a 10mm restructuring charge) but failed to address investor concerns around NPLs. The apparent lack of progress on the topic of NPLs triggered capitulation by the sellside with nearly every analyst covering the stock downgrading it in a 5 week period, citing uncertainty and a permanent delay of dividends and excess capital returns (on the assumption that PTSB will generate substantial losses on NPL resolution).

• For the core business, pre-provision profits were up 18% y-y on NIM expansion, mortgage loan growth of 60%, and cost control. Provisions were 6bps, meaning the company generated capital in the period.

• The market was expecting the company to make an update regarding NPLs, as a few sellside firms had pre-emptively (and wisely it would seem) downgraded the stock leading up to the announcement given their concerns about NPL uncertainty. Management did not give any detailed guidance on the topic, preferring to give “actions, not words” on the topic later in the year. The lack of detail on the topic further stoked market fears about losses and triggered additional downgrades and capitulation selling by the remaining bulls. The general thinking is that since PTSB has not done anything on the NPLs year to date, that they cannot do anything on the NPLs going forward, and must be facing a large loss. The reality is more complicated:

• PTSB was finalizing the regulatory tracker review process in 1H’17, as they had previously noted in March 2017. You cannot modify & foreclose on loans when the government is investigating your modification and foreclosure process – that is over now and they are moving forward with foreclosure and NPL cures in 2H’17. The Street has been highly skeptical of this explanation given BoI and AIB worked out NPLs in the commercial book and continued to modify residential loans in the half year. However, looking at the pace of new NPL formation and cure activity in FSTNT monthly data (PTSB performing and re-performing securitizations) one can see that cure activity picked up in May and June to take net NPL formation negative in those months after building early in the year. Hence, one would expect the pace of NPL reduction to increase in 2H’17 even without portfolio sales given the resumption of loan modifications. The reason as to why only PTSB followed CBR review guidelines in halting modification activity in the early part of the year is much less clear. Regardless, loan modifications and cures have resumed.

• Given the rising Irish home price environment (+11% in 2017), each month they wait to sell the better the price will be. Furthermore, they are pushing the loans in the untreated book through the final stages of legal process after 4-5 years and the further along the loan is at the time of NPL sale, the lower the discount on legal process uncertainty required by the buyer.

• It also seems fair to say that the company was a bit shocked by the share price reaction, hence the late breaking news on selling the 1bn of BTL NPLs a few days after they were berated in the press on “zero progress on NPLs” http://www.independent.ie/business/irish/ptsb-may-launch-1bn-loan-sale-as-early-as-the-end-of-this-year-35974900.html

• PTSB arguably does have a worse-than-usual BTL book, but it also carries a higher than usual collective allowance against it. On specific impairments alone, the NPLs are carried at 47.5% (specific provisions on impaired loans) at 6/30/17, but including the collective / IBNR reserve and all NPLs, they are carried at 40.5% of par. If one excludes the “NPL, not impaired, no arrears” category (most likely performing loans where the same borrower has defaulted on a different loan, hence classified as NPL, e.g. “technical NPLs”) then the book is marked at 34% of par including collective reserves. Selling that book even at 45% would generate a net gain, albeit via a loss on the specific marks but a release on the collective reserves. Hence the importance of looking at the full picture on reserving. Having said all of this, if PTSB sells in this book in the near future I would argue for higher prices than 47.5% simply due to cash generation from the book, AIB's precedent sale price, and the availability of securitization funding; clearly this would change if the market window closed.

• PTSB gave guidance on a 250bps fully loaded CT1 hit from the TRIM exercise, which is an estimate in isolation without any impact from selling NPLs at 100% risk weightings. For example, if PTSB sells its 1000mm BTL book per the press commentary, that would significantly offset the 2800mm increase in RWAs due solely to model changes on performing loans.  As a side note, there is a very reasonable debate to be had on whether the ECB should punish all Irish banks with 50% risk weightings on residential mortgages simply due the legacy of the Irish housing bubble; other European jurisdictions carry risk weights of half that amount and the idea of TRIM is to harmonize models across countries. In any event, all the TRIM update did was further discourage the bull case around excess capital and isolate the bank as the most impacted since their peers declined to provide guidance on the topic.

In summary, the perception today is that PTSB cannot and will not address its NPL situation. Securitization data provides the insight on funding advance rates and actual NPL performance that is being missed by Street analysts who have left the company for dead. Curiously, in a world where private capital is struggling to generate attractive returns on incremental long investments and NPL managers in Europe cite rising competition & prices on new portfolio sales, the market has decided that this company cannot dispose of Irish residential NPLs at any price. Never mind that the asset class in question is rising 11% per year and the loans pay a current cash yield. All the company has to do is start selling the loans.

Catalyst

The company sells a portfolio of BTL loans in 2H'17