Category: Regulatory Capital

Contingent Capital

OSFI Revises Rules for LRCNs; Finally Provides Information for Insurers

The Office of the Superintendant of Financial Institutions Canada (OSFI) has announced:

OSFI has published an update to its July 18, 2020 capital ruling on the Limited Recourse Capital Notes (LRCNs). The revised ruling now addresses LRCN issuances from institutions of all sizes and across different industry sectors.

The revisions include a number of clarifications to the ruling’s conditions and limitations, which are part of OSFI’s prudent approach to assessing the quality and quantity of financial instruments used as regulatory capital. This is consistent with our mandate to protect the rights and interests of depositors, policyholders and financial institution creditors, while also allowing financial institutions to compete effectively and take reasonable risks.

The ruling continues to conclude that federally regulated financial institutions may recognize the LRCNs as regulatory capital subject to the capital treatment, conditions and limitations set out in the revised ruling. Should you have any questions, please contact CapitalConfirmations@osfi-bsif.gc.ca.

The new limits are set out in the appendix to the new rules:

LRCN Issuance Limitations by FRFI Sector
Note: For life insurers, the following limitations supplement and are subject to any existing capital composition limits set out in OSFI’s capital guidelines. P&C insurers and mortgage insurers should consult OSFI’s Capital Division in respect of the limitations applicable to any prospective LRCN issuances.

FRFI
Sector Regulatory
Capital
Treatment LRCN Issuance Cap Floor

Deposit-
Taking
Institutions AT1 Greater of $150 million, 0.75%
RWA, or 50% of the
institution’s aggregate net AT1
capital Lesser of 0.30% RWA or 20% of
the institution’s aggregate AT1
capital

Life Tier 1 Capital
other than
Common
Shares Greater of $150 million or
12.5% of Net Tier 1 capital 5.0% of Net Tier 1 capital

The “Floor” has the following effect:

The Cap may be removed with the prior approval of OSFI’s Capital Division. In seeking this approval, a FRFI must demonstrate that it has issued institutional preferred shares and/or other Additional Tier 1 capital instruments (other than LRCNs) targeted towards institutional investors that, in aggregate, are no less than the applicable limit, or Floor, set out in the Appendix. If the FRFI’s outstanding institutional preferred shares and/or other Additional Tier 1 capital instruments issued to institutional investors were to subsequently drop below the Floor, the FRFI would not be permitted to issue additional LRCNs in excess of the Cap until it has re-established compliance with the Floor. The Floor will not apply where the LRCNs are issued exclusively to a FRFI’s affiliates.

Contingent Capital

S&P Revises Bank Outlook to Stable on Federal Complacency

Standard & Poor’s has announced:

•We continue to evaluate the likelihood, degree, and timeframe with respect to which the default risk of systemically important Canadian banks may change as a result of the government’s progress toward introducing a bank bail-in framework.
•We now expect that the timeframe could be substantially longer than we had previously assumed. We see the absence of the topic from the new government’s Dec. 4 Speech from the Throne as recent, incremental evidence in this regard.
•We now do not expect to consider the removal of rating uplift for our expectation of the likelihood of extraordinary government support from the issuer credit ratings (ICRs) on systemically important Canadian banks until a point beyond our standard two-year outlook horizon for investment-grade ratings, if at all.
•When and if we remove such uplift, the potential ratings impact will also consider uplift for additional loss-absorbing capacity, as well as any changes to our stand-alone credit profiles on these banks.
•As a result, we are revising our outlooks on all systemically important Canadian banks to stable from negative.

RATING ACTION
On Dec. 11, 2015, Standard & Poor’s Ratings Services revised its outlooks on the Canadian banks that it views as having either “high” (Bank of Montreal, The Bank of Nova Scotia, Canadian Imperial Bank of Commerce, Royal Bank of Canada, The Toronto-Dominion Bank), or “moderate” (Caisse centrale Desjardins and National Bank of Canada) systemic importance, to stable from negative (see ratings list). The issuer credit ratings (ICRs) on the banks are unchanged.

RATIONALE
We believe that the potential negative ratings impact of a declining likelihood of extraordinary government support, at least within our standard two-year outlook horizon, has subsided. This reflects our updated view that there could be an extended implementation timetable–2018 or later–for the proposed Canadian bail-in framework. Importantly, at the point we would consider removing any uplift from the likelihood of extraordinary government support from our ratings, we would also consider the potential ratings impact of any uplift for additional loss-absorbing capacity (ALAC), as well as any
changes to our stand-alone credit profiles (SACPs) on these banks. In our view, the extended timetable introduces some potential that either the presence of ALAC or fundamental changes in credit quality at individual banks might come into play more than under the previously contemplated timetable.

We had revised our outlooks on systemically important Canadian banks to negative chiefly in reaction to the former government’s “Taxpayer Protection and Bank Recapitalization Regime” consultation paper of Aug. 1, 2014, as we then expected a bail-in regime could be fully implemented by 2016 (see “Outlook On Six Big Canadian Banks Revised To Negative Following Review Of Bail-In Policy Proposal,” published Aug. 8, 2014, on RatingsDirect). A number of subsequent developments have caused us to re-evaluate this expectation:

•In its April 2015 budget proposal, the former government affirmed its intention to introduce a bank bail-in regime in Canada, but it provided only very limited additional information relative to what it had outlined in its 2014 consultation paper; nor did the government make substantial subsequent public statement on the topic; nor did it specify timing for the announcement of its fully-developed (post-consultation) legislative proposal.
•The Oct. 19 federal election changed the party in government to Liberal (center-left), from Conservative (center-right). The former government’s proposed bail-in regime did not feature prominently in election debates.
•The new government’s Dec. 4 Speech from the Throne made no mention of the proposed bail-in framework, nor were any of the legislative priorities enumerated therein closely related, in our opinion. We believe this indicates the introduction of a bail-in framework is not among the immediate priorities of the new government.

Moreover, with Canada experiencing no government bank bail-outs, nor large bank failures, for decades, we believe the political incentive to rapidly end “too-big-to-fail” is less in Canada than in the U.S. and several EU countries, which are jurisdictions under which we have already removed uplift for our expectation of the likelihood of extraordinary government support from our ratings (see “U.S. Global Systemically Important Bank Holding Companies Downgraded Based On Uncertain Likelihood Of Government Support,” and “Most European Bank Ratings Affirmed Following Government Support And ALAC Review,” both published Dec. 2, on RatingsDirect). We will take this factor into consideration as we continue to evaluate our view on the likelihood of extraordinary government support in Canada relative to not only the U.S. and Europe, but also other jurisdictions where we maintain a government support assessment of “supportive” or “highly supportive” under our criteria (such as for many countries in Latin America and Asia-Pacific; see “Banking Industry Country Risk Assessment Update: November 2015,” published Nov. 27).

We now believe the procedural hurdles to passing legislation and related regulations (the latter after passage of the former) for a bail-in regime will alone require a minimum of one-to-two years, after the new government decides on a final legislative framework to propose to Parliament. Considering all of this, we now expect the eventual date for initial implementation of the bail-in regime (that is, banks issuing bail-inable debt) could be in 2018 or later.

In addition, and in contrast to bail-in frameworks outlined by U.S. authorities or in European countries like Germany, Canadian officials’ statements have made clear that only debt issued or renegotiated after an initial implementation date would be subject to conversion. It will take some time for the banks to issue or renegotiate bail-inable debt. We believe this means it could take several years after the initial implementation date before we would consider a Canadian bail-in regime effective, so as to provide a viable alternative to the direct provision of extraordinary government support.

As well, and again in contrast to the U.S. and EU jurisdictions, Canadian governments have made no attempt to limit their ability to provide direct extraordinary support to their banks, if needed. We expect bailing in senior creditors to be the first Canadian policy response in the face of a crisis. At the same time, we believe Canadian governments would be likely to consider all policy options, in such a circumstance. It is therefore not certain that the introduction of a bail-in regime would of itself result in our revising our government support assessment on Canada to “uncertain” from the current “supportive” and the removal of rating uplift for our view on the likelihood of extraordinary government support from our ICRs on systemically important Canadian banks. Rather, our decision would depend, among other factors, on the details of the eventual bail-in regime, including the extent to which bail-inable and unbail-inable senior debt is distinguishable.

Partly to honor G-20 and other international commitments, the Canadian government will, we expect, present a finalized legislative proposal for the bail-in framework in 2016 or 2017. However, we expect an implementation date that could be in 2018 or later, and we think it could take at least one and possibly several years more for substantial bail-in eligible debt to be in place. With a runway that long, the potential ratings impact from removing uplift for the likelihood of extraordinary government support is beyond our standard two-year outlook horizon for investment-grade ratings, and could by then be more meaningfully affected by either ALAC uplift (from the bail-inable debt, assuming our related criteria are met) or SACP changes, than under the previously contemplated timetable.

When the government presents the detailed provisions of the framework, along with a more specific timeframe, we will review the applicable notching for various bank liabilities, taking into account the framework’s implications on different instruments. We expect that issue ratings on new bail-inable instruments will be at a level that is notched in reference to banks’ SACPs, while ratings on non-bailinable senior debt may continue to incorporate rating uplift above the banks’ SACPs, based on our expectation of the likelihood of extraordinary government support, or ALAC.

OUTLOOK
Our outlooks on the systemically important Canadian banks are stable, based on our reassessment of the likelihood, degree, and timeframe with respect to which the default risk of systemically important Canadian banks may change as a result of the government’s progress toward introducing a bank bail-in framework. We believe that the likelihood of extraordinary government support will continue to be a factor in systemically important Canadian bank ratings throughout the current outlook period.

Moreover, we believe these banks will continue to exhibit broad revenue diversification, conservative underwriting standards, and strong overall asset quality. Our current view is that the impact of low oil prices on their profitability and credit quality will be contained, given the modest direct exposure of the banks to the oil and gas sector, and the limited knock-on impact so far on consumer credit in regional economies affected by low oil prices.

On the other hand, we continue to monitor a number of key downside risks to our ratings on these banks, including low margins, high Canadian consumer leverage, residential real estate prices we believe are at least somewhat inflated in some parts of Canada, a Canadian macroeconomic outlook that is very tentative, and the higher-risk nature of certain recent foreign acquisitions.

The August 2014 imposition of Outlook-Negative was reported on PrefBlog, as was the federal consultation on the recapitalization regime. As far as I can tell, the comments received on the consultation paper have not been published; I believe this is because Canadians are too stupid to understand smart stuff like legislation and parliament and all that – if given a pile of comments to work through, we’d probably try to eat them.

Issues affected are:

BMO.PR.K, BMO.PR.L, BMO.PR.M, BMO.PR.Q, BMO.PR.R, BMO.PR.S, BMO.PR.T, BMO.PR.W, BMO.PR.Y and BMO.PR.Z

BNS.PR.A, BNS.PR.B, BNS.PR.C, BNS.PR.D, BNS.PR.L, BNS.PR.M, BNS.PR.N, BNS.PR.O, BNS.PR.P, BNS.PR.Q, BNS.PR.R, BNS.PR.Y and BNS.PR.Z

CM.PR.O, CM.PR.P and CM.PR.Q

NA.PR.Q, NA.PR.S and NA.PR.W

RY.PR.A, RY.PR.B, RY.PR.C, RY.PR.D, RY.PR.E, RY.PR.F, RY.PR.G, RY.PR.H, RY.PR.I, RY.PR.J, RY.PR.K, RY.PR.L, RY.PR.M, RY.PR.N, RY.PR.O, RY.PR.P RY.PR.W and RY.PR.Z

TD.PF.A, TD.PF.B, TD.PF.C, TD.PF.D, TD.PF.E, TD.PF.F, TD.PR.S, TD.PR.T, TD.PR.Y and TD.PR.Z

Contingent Capital

Update On OSFI Insurer Regulation

OSFI Assistant Superintendent Neville Henderson gave a speech to the 2015 Life Insurance Invitational Forum:

Domestic Insurance Capital Standards

On the domestic front, we are still on track to implement OSFI’s new life insurance regulatory capital framework in 2018. The new capital framework will provide a superior risk based assessment methodology for determining capital requirements. The new test will make use of more current analysis and methodologies as well as explicitly taking into account mitigating actions and diversification benefits. It will allow our capital requirements to remain state of the art compared to those of other jurisdictions.

The capital changes in the new framework are explicitly calibrated to a consistent level of conditional tail expectation (CTE) across the various risks. Actuarial valuation of insurance company liabilities are explicitly intended to include conservative margins with the degree of conservatism varying across risks.

To help ensure that this approach results in consistent capital measures across companies, OSFI has asked the Canadian Institute of Actuaries and the Actuarial Standards Board to consider certain issues with a view to updating actuarial standards and /or guidelines if required.

To avoid double counting and inconsistent treatment of different risks, this new framework will include margins for adverse deviations as an available capital resource.

While we are awaiting the results of Quantitative Impact Study (QIS)7, we are in the process of planning to conduct two framework runs, one in 2016 followed by another one in 2017. These “test drives” will allow us to validate the new capital test and help insurers gear up for the updated regulatory compliance requirements under the new framework.

We should also have a final guideline ready for issue in July 2016, following input from the industry on the draft. Any anomalies uncovered in the testing will be taken into consideration prior to implementation. This will allow time for industry feedback and enable insurers to plan and prepare their systems for implementation of the framework in early 2018.

Global Insurance Capital Developments

While work continues on the domestic front, there are also developments in standards for internationally active insurers.

The International Association of Insurance Supervisors (IAIS) is refining the Basic Capital Requirement (BCR) and Higher Loss Absorbency (HLA) requirements for Global Systemically Important Insurers (GSIIs) for implementation in 2019. Work in this area is aimed at mitigating or avoiding risks to the global financial system.

To eventually replace the BCR, the IAIS is developing an internationally agreed upon risk based capital test. The Insurance capital standard (ICS 1.0) for the broader list of Internationally Active Insurance Groups (IAIG) will be ready by the end of 2016, for implementation in 2019.

OSFI looks carefully at the Canadian marketplace and Canadian requirements before deciding whether to adopt international standards. We will take ICS into consideration as we fine tune our current capital tests. The work we do on the OSFI life insurance framework already includes many of the changes stemming from these international standards and we don’t expect ICS 1.0 to be as sophisticated as our current Minimum Continuing Capital and Surplus Requirements (MCCSR) capital test. Consequently, we do not foresee a need to implement any significant changes.

The significant changes will likely come as ICS 2.0 is finalized. It may bring sufficient worldwide convergence for OSFI to start thinking about implementation.

The important thing about ICS is that this is what will determine whether or not preferred shares must be convertible into equity (or have other pre-bankruptcy capital loss absorption features) in order to be counted as Tier 1 capital. This proposal is outlined in the Consultation Paper “Risk-based Global Insurance Capital Standard” which is available in a ludicrously inconvenient manner, paragraph 92 with associated question 25:

92. The IAIS is considering a requirement for a principal loss absorbency mechanism to apply to Tier 1 instruments for which there is a limit. This principal loss absorbency mechanism would provide a means for such instruments to absorb losses on a going-concern basis through reductions in the principal amount in addition to cancellation of distributions.

Question 25. Should Tier 1 instruments for which there is a limit be required to include a principal loss absorbency mechanism that absorbs losses on a going-concern basis by means of the principal amount in addition to actions with respect to distributions (e.g. coupon cancellation)? If so, how would such a mechanism operate in practice and at what point should such a mechanism be triggered?

OSFI’s response to this question is available in the document “Compiled Responses to ICS Consultation 17 Dec 2014 – 16 Feb 2015”, which is also available in a ludicrously inconvenient manner:

No, OSFI does not support the inclusion of a principal loss absorbency mechanism on Tier 1 instruments for which there is a limit. Tier 1 instruments must be able to absorb losses on a going concern basis, which these instruments do through coupon cancellation.

Despite this, I expect that OSFI will adopt whatever ends up being in ICS, as in this way any future criticism will be deflected to the international body and they will be able to keep their jobs and continue angling for future employment with those whom they currently regulate.

OSFI’s response to this – and other – questions has never been explained to the Canadian public as far as I know, because we’re disgusting taxpayer and investor scum, not worth the dirt underneath our own fingernails.

Further discussion of the capital standard and my reasons for believing that the NVCC rule will be applied to insurers and insurance holding companies are provided in every edition of PrefLetter.

Regulation

OSFI: Ineffectual, Uninformed Grandstanding on D-SIBs

The Office of the Superintendent of Financial Institutions has announced:

Canada’s six largest banks have been identified as being of domestic systemic importance, and will be subject to continued supervisory intensity, enhanced disclosure, and a one per cent risk weighted capital surcharge by January 1, 2016.

Grant Robertson of the Globe claims:

The move is designed to avert a liquidity crisis in the sector, and comes on top of the 7 per cent of capital that the Office of the Superintendent of Financial Institutions (OSFI) requires them to hold, which can be easily liquidated by the banks during a time of financial pressure to stabilize operations.

This shows a common confusion between “liquidity” and “solvency”. If you own a house worth a million with no mortgage, but can’t pay for groceries, you are solvent, but illiquid. If you pay for the groceries with all that’s left of the 1.5-million mortage you took on the place five years ago, you are liquid, but insolvent. There was a time when reporters were familiar with their subjects and had the names and ‘phone numbers of experts available to explain arcane elements of business news. Imagine that!

The adjustment to the capital rules under discussion here addresses expectations of solvency but do nothing directly to address liquidity.

Be that as it may, OSFI provided some charts with its cover letter to the banks:

Click for Big

As is OSFI’s habit, the Advisory giving effect to the decision, Domestic Systemic Importance and Capital Targets – DTIs, makes only the slightest possible effort to explain the decision:

The common equity surcharge associated with D-SIB status in Canada will be 1% Risk Weighted Assets (RWA).This surcharge takes into account the structure of the Canadian financial system, the importance of large banks to this financial architecture, and the expanded regulatory toolkit to resolve a troubled financial institution. This means that banks designated as a D-SIB will be required to meet an all-in Pillar 1 target common equity Tier 1 (CET1) of 8% RWA commencing January 1, 2016. The 1% capital surcharge will be periodically reviewed in light of national and international developments. This is consistent with the levels and timing set out in the BCBS D-SIB framework.

The BCBS D-SIB framework provides for national discretion to accommodate characteristics of the domestic financial system, and other local features, including the domestic policy framework. The additional capital surcharge for banks designated as systemically important provides credible additional loss absorbency given:

Extreme loss events as a percentage of RWA among this peer group over the past 25 years would be less than the combination of the CET1 (2.5%) capital conservation buffer and an additional 1%; and
Current business models of the six largest banks are generally less exposed to the fat tailed risks associated with investment banking than some international peers, and the six largest banks have a greater reliance on retail funding models compared to wholesale funding than some international peers – features that proved beneficial in light of the experience of the last financial crisis.
From a forward looking perspective:

o Canadian banks that hold capital at current targets plus a 1% surcharge (i.e. 8%) should be able to weather a wide range of severe but plausible shocks without becoming non-viable; and
o The higher loss absorbency in a crisis scenario (conversion to common equity or permanent write downs) of the 2% to 3% non-common equity capital in Tier 1 and subordinated debt in total capital required by Basel III also adds to the resiliency of banks.

It gives me a warm feeling inside knowing that OSFI has looked at the past twenty-five years of history to gauge extreme loss events; the Basel II guidelines supposedly calibrated more stringently:

The confidence level is fixed at 99.9%, i.e. an institution is expected to suffer losses that exceed its level of tier 1 and tier 2 capital on average once in a thousand years.

OSFI’s document has a few references, but only to other OSFI documents and a few Basel Committee on Banking Supervision hymn books; nothing of any meat, nothing that would provide any comfort that these guys have thought things through and know what they’re doing – but OSFI’s institutional intellectual dishonesty is well known.

Their efforts may be compared – just for starters – with a paper titled Australia: Addressing Systemic Risk Through Higher Loss Absorbency—Technical Note, published by the IMF and reposted by the Australian Prudential Regulation Authority. One of the useful features of this report is “Table 4. Cross-Country Comparison of Approaches to D-SIBs”, which – although one can hardly credit it – looks at what other countries are doing! Here’s an extract:

Country	HLA
Singapore	2 percent additional by 2015
Sweden	Accelerated adoption of Basel III; plus 3 percent by 2013; 5 percent by 2015
Switzerland	19 percent of RWA total capital, of which up to 9 percent cocos, by 2016
United Kingdom	Proposal: 3 percent additional to Basel III and up to 17 percent of RWA loss absorbency for the largest institutions and ring-fenced entities
United States	Supplementary Tier 1 of 3 percent of RWA for complex institutions

Now it may very well be that OSFI is taking a prudent route in being so much more lenient with the banks than their international counterparts – but you’d never know it from reading OSFI material. Canadians are forced to take it on trust that the banking regulator knows what it’s doing; and OSFI’s arrogance makes such trust an awfully scarce commodity.

One highly recommended example of how a prudential regulator should operate is the UK’s Independent Commission on Banking – Final Report Recommendations – September 2011:

The Independent Commission on Banking (the Commission) was established by the Government in June 2010 to consider structural and related non-structural reforms to the UK banking sector to promote financial stability and competition. The Commission was asked to report to the Cabinet Committee on Banking Reform by the end of September 2011. Its members are Sir John Vickers (Chair), Clare Spottiswoode, Martin Taylor, Bill Winters and Martin Wolf.

This report has one of its recommendations highlighted in the table extracted above:

As to that, the Commission recommends that the retail and other activities of large UK banking groups should both have primary loss-absorbing capacity of at least 17%-20%. Equity and other capital would be part of that (or all if a bank so wished). Primary loss absorbing capacity also includes long-term unsecured debt that regulators could require to bear losses in resolution (bail-in bonds). If market participants chose, and regulators were satisfied that the instruments were appropriate, primary loss-absorbing capacity could also include contingent capital (‘cocos’) that (like equity) takes losses before resolution. Including properly loss-absorbing debt alongside equity in this way offers the benefit that debt holders have a particular interest, in a way that equity holders do not, in guarding against downside risk. If primary loss-absorbing capacity is wiped out, regulators should also have the power to impose losses on other creditors in resolution, if necessary.

Assiduous Readers will recognize that I have a fundamental distaste for the trashing of five hundred years of bankruptcy law implied by the last sentence, but at least the rationale is spelt out in credible format – far different from the Canadian model.

Regulation

OSFI: Life Insurance Regulatory Framework

On September 5 the Office of the Superintendant of Financial Institutions announced:

released a Life Insurance Regulatory Framework to provide life insurance companies and industry stakeholders with an overview of regulatory initiatives that OSFI will be focusing on over the period ending 2016. It outlines how the regulatory framework will evolve to ensure Canadians continue to benefit from a strong life insurance industry.

“In laying out OSFI’s initiatives, we hope to encourage discussion and strong participation by industry stakeholders in our regulatory development process,” said Mark Zelmer, Assistant Superintendent, Regulation Sector. “Canadians have benefited from a strong life insurance industry and a flexible, effective regulatory framework. Our initiatives aim to ensure this continues.”

The Framework outlines OSFI’s priorities and addresses issues such as corporate governance and risk management, evolving regulatory capital requirements, and promoting transparent information on the financial condition of life insurance companies to support the regulatory framework.

“This framework addresses OSFI’s key regulatory objectives and its approach to refining regulatory oversight and guidance that is already robust,” continued Mr. Zelmer. “By issuing the regulatory framework at this time, OSFI hopes it will help life insurers and industry stakeholders in their planning processes.”

I missed this at the time, but it was brought to my attention by Assuiduous Reader dudsy in the comments to another thread.

The document is titled Life Insurance Regulatory Framework. Naturally, my main concern is to parse the text for any hints about the application of the NVCC rule to insurers and insurance holding companies:

OSFI recognizes that life insurance companies are in many ways significantly different than banks, particularly due to the long-term nature of traditional life insurance business. Therefore, in considering these developments, OSFI will not indiscriminately implement any of them (e.g., Basel III) into the life insurance regulatory framework.
…
To achieve these objectives, OSFI will introduce enhancements to the regulatory framework for life insurance companies through:…Revised regulatory capital requirements guidance that:…Links risk measures to the quality of capital
available to absorb losses
…
OSFI is approaching the review of the regulatory capital requirements with the belief that, in aggregate, the industry currently has adequate financial resources (total assets) for its current risks.
…
Capital will improve in terms of its ability to absorb losses, from the perspective of both a “going concern” and a “gone concern” basis.

The last seems quite encouraging, as far as NVCC is concerned. More important to OSFI, however is plausible justification for mission creep and increased employment at OSFI:

OSFI may need more specialized resources as these initiatives are incorporated into our regulatory and supervisory frameworks.

They’re going to introduce something called ORSA, which does not, surprisingly, stand for OSFI Retirees Superannuation Arrangement, but Own Risk and Solvency Assessment:

The minimum capital requirements set in OSFI’s regulatory framework may not be adequate to address this institution-specific risk-taking, as the regulatory capital requirements are based on industry averages which, at any point in time, may not fully capture new risk exposures or product developments. For this reason, institutions should have their ORSA process. Life insurance companies should not simply rely on minimum regulatory capital requirements as a proxy or as a starting point for measuring their own risk profile.
…
ORSA should not be seen as an OSFI compliance requirement but as a sound business practice. This will be reflected in the principles-based approach OSFI will outline in the ORSA Guideline. The ORSA Guideline will build on existing industry practice and OSFI guidance while considering international practices, in addition to seeking input and perspective from Canadian industry stakeholders.

In the section titled “Evolving Regulatory Capital Requirements”, they say:

The objective of this review is to improve our regulatory capital requirements by:…Improving Risk Measurement…Recognize the quality of capital available to absorb losses on both a “going concern” and a “gone concern” basis
…
The evolution of regulatory capital requirements into a more risk sensitive framework may result in more volatile regulatory capital requirements (capital available and/or capital required). OSFI will consult with industry to assess whether that volatility provides a true reflection of the evolution of the risk and is thus “appropriate” for purposes of setting regulatory capital requirements, or whether the volatility in capital requirements amplifies the variations in risk and is thus “inappropriate.”

Of great interest is their commentary on accounting standards:

Where necessary, OSFI will consider measures to address inappropriate volatility. For example, we will investigate options to moderate the impact of volatility on regulatory capital requirements, when:
1. For remaining long duration liabilities, markets for matching purposes do not exist, and
2. For solvency purposes, accounting/actuarial rules do not appropriately reflect the long-term characteristics of these portfolios.

That might be code for “Don’t worry about the IFRS Insurance Contracts Exposure Draft, guys!” The Insurance Contracts issue is actually mentioned explicitly in the concluding section of the paper:

The IASB insurance contracts project (IFRS 4 Phase II) will have a significant impact. While the extent of the impact is not fully known (and will not be until the final standard is set), OSFI is committed to consulting with industry stakeholders on how the final standard should be incorporated into the regulatory framework. Ideally, our initiatives would incorporate a final IFRS 4 Phase II standard. However, should a significant delay occur in the IASB work, OSFI will continue to move its work forward using current international financial reporting standards.

The section titled “Capital and Risk Measurement”:

The level of protection being tested by OSFI in QIS 3 is for each risk separately to cover a 1-in-200 year event (a rare, but plausible scenario) over a one-year time horizon. OSFI believes this level of protection would be equivalent to the low end of the investment grade range. An adequate provision after one year is defined as the amount of assets required for the insurer to either fulfil its policyholders’ and senior creditors’ obligations over the remaining lifetime of the obligations or to transfer them to another company.

Of great importance is their admission that:

The current approach to determining liability and regulatory capital requirements for financial guarantees embedded in segregated fund products has the following drawbacks:
• It can produce values that are materially lower than the cost of hedging.

The closes that they get to addressing the NVCC issue with respect to preferred shares is:

OSFI believes that high quality capital instruments should form a substantial part of the capital resources of an insurer when times are good. This provides the company, and OSFI, with the flexibility to respond in a constructive way in times of stress.

OSFI will consider these elements in developing guidance for the level and quality of available capital in the revised regulatory capital requirements.
…
The review of the definition of capital component is necessary to incorporate lessons learned during the recent financial crisis. These relate to the quality of certain capital instruments during periods of stress, the appropriateness of deductions and adjustments made to regulatory capital. The review provides an opportunity to consider each available capital element and assess its contribution to two goals: financial strength and protection of policyholders and creditors.

Revisions will provide increased transparency with respect to the meaning and purpose of both total (protection of policyholders and senior creditors) and tier 1 (financial strength) capital elements.

OSFI believes going concern capital (tier 1) should be largely comprised of equity (common and perpetual preferred shares). Items not considered to be readily available to absorb losses in a stress scenario (i.e., not fungible, not permanent, introduce an element of double-counting) should be deducted from it. Going concern capital is important to support ongoing insurer viability over the longer term given the longer-term nature of the life insurance business.

Gone concern capital (total) helps ensure that policyholders and senior creditors can be paid when the insurer is in winding-up mode. Gone concern capital may include forms of lower-quality “additional” capital components, such as hybrids and subordinated debt instruments that meet minimum quality criteria.

OSFI plans to issue a draft Definition of Capital paper for public consultation in late 2012 or early 2013.

The phrase “lessons learned during the recent financial crisis” might – might! – be taken as a reference to hybrid capital not defaulting when financial institutions were bailed out, which is the justification for the NVCC rule.

However, the last sentence quoted implies that we’ll start getting some meat in the sandwich sometime around year-end … roughly TWO FREAKING YEARS after the NVCC rule was applied to banks.

The timeline section at the back gives the following estimates for “Definition of Capital”:
Project Initiation: 2011Q1
Quantitative Impact Study: 2013Q3
Public Consultation: 2013Q4
Final Guideline issued: 2014
Implementation Milestone: 2015

At this point, I see no reason to change my views regarding the potential for the eventual imposition of the NVCC rules on insurers and insurance holding companies in a similar manner to banks. The draft consultation to be issued around year-end may help firm up the matter; readers and investors should be aware that I may well change the “Deemed Maturity” date for insurers and insurance holding companies.

Regulatory Capital

OSFI Releases New Draft Capital Guidelines

The Office of the Superintendent of Financial Institutions has released a new draft version of the capital guidelines for Deposit Taking Institutions. Comments are being accepted to November 19, but I won’t bother – OSFI has never shown any good-faith interest in encouraging public debate.

2.1.1.4. Examples of acceptable features

Outlined below are examples of certain preferred share features that may be acceptable in tier 1
capital instruments:

a simple call feature that allows the issuer to call the instrument, provided the issue cannot be redeemed in the first five years and, after that, only with prior supervisory approval
a dividend that floats at some fixed relationship to an index or the highest of several indices, as long as the index or indices are linked to general market rates and not to the financial condition of the borrower
a dividend rate that is fixed for a period of years and then shifts to a rate that floats over an index, plus an additional amount tied to the increase in common share dividends if the index is not based on the institution’s financial condition and the increase is not automatic, not a step-up, nor of an exploding rate nature
conversion of preferred shares to common shares where the minimum conversion value or the way it is to be calculated is established at the date of issue. Examples of conversion prices are: a specific dollar price; a ratio of common to preferred share prices; and a value related to the common share price at time of conversion.

2.1.1.5. Examples of unacceptable features

Examples of preferred share features that will not be acceptable in tier 1 capital are:

an exploding rate preferred share, where the dividend rate is fixed or floating for a period and then sharply increases to an uneconomically high level
an auction rate preferred share or other dividend reset mechanism in which the dividend is reset periodically based, in whole or part, on the issuer’s credit rating or financial condition
a dividend-reset mechanism that does not specify a cap, consistent with the institution’s credit quality at the original date of issue

These examples have not changed since the November 2007 edition.

Regulatory Capital

SLA SLEECS: 5.863%+340

Sun Life has announced:

that Sun Life Capital Trust II, a trust established by Sun Life Assurance Company of Canada (“SLA”), will issue in Canada $500 million principal amount of Sun Life ExchangEable Capital Securities Series 2009-1 due December 31, 2108 (“SLEECS Series 2009-1”) under a final prospectus that it intends to file with the Canadian securities regulators as soon as possible. The SLEECS Series 2009-1 are expected to qualify as regulatory capital of SLA. The net proceeds of the issue will be used by SLA for general corporate purposes, including investments in subsidiaries.

Interest on the SLEECS Series 2009-1 will be payable semi-annually. The interest rate on the SLEECS Series 2009-1 from the date of issue to but excluding December 31, 2019 will be 5.863% per annum. On that date and thereafter on each fifth anniversary of that date, the interest rate on the SLEECS Series 2009-1 for the ensuing five years will be reset at a rate equal to 3.40% above the then-yield on a Government of Canada bond having a term to maturity of five years.

As described further in the prospectus, the SLEECS Series 2009-1 may in certain circumstances be automatically exchanged for a series of SLA preferred shares, and in certain other circumstances a series of SLA preferred shares may be issued in lieu of interest payable on the SLEECS Series 2009-1.

On or after December 31, 2014, subject to certain conditions, the SLEECS Series 2009-1 may be redeemed in whole or in part at the option of Sun Life Capital Trust II.

The issue of SLEECS Series 2009-1 is underwritten by a syndicate co-led by Scotia Capital Inc. and RBC Dominion Securities Inc., and is expected to close on November 20, 2009.

Note that these are issued at the level of SLA, the OpCo, which is intrinsically a better credit than SLF (the holdco), and in times of trouble they will convert to SLA prefs. DBRS rates existing SLEECS at A(high), the same as SLF sub-debt.

The following is from the preliminary prospectus:

From the Closing Date to but excluding •, 2019, the interest rate on the SLEECS will be fixed at •% per annum. Assuming the SLEECS are issued on •, 2009, the first interest payment due on the SLEECS on •, 2009 will be $• per $1,000 principal amount of SLEECS. Each interest payment on the SLEECS after the first interest payment (subject to the reset of the interest rate from and after •, 2019) will be in the amount of $• per $1,000 principal amount of SLEECS. Starting on •, 2019, and on every 5th anniversary of such date thereafter until •, 2104 (each such date, an “Interest Reset Date”), the interest rate on the SLEECS will be reset at an interest rate per annum equal to the Government of Canada Yield plus •%. The SLEECS will mature on •, 2108

On or after •, 2014, the Trust may, at its option, with the prior approval of the Superintendent, on giving not more than 60 nor less than 30 days’ notice to the holders of the SLEECS, redeem the SLEECS, in whole or in part. The redemption price per $1,000 principal amount of SLEECS redeemed on any day that is not an Interest Reset Date will be equal to the greater of par and the Canada Yield Price, and the redemption price per $1,000 principal amount of SLEECS redeemed on any Interest Reset Date will be par, together in either case with accrued and unpaid interest to but excluding the date fixed for redemption, subject to any applicable withholding tax. The redemption price payable by the Trust will be paid in cash. See “Description of the Trust Securities — SLEECS—Trust Redemption Right”.

Interest Reset Date means •, 2019, and every fifth anniversary of such date thereafter until •, 2104, on which dates the interest rate on the SLEECS and the SLA Debenture will be reset as described in this prospectus.

In bad times, the SLEECS convert to SLA preferred shares paying 30-Year Canadas + •

The SLF sub-debt, 5.4 of pretend-2037, are quoted by Perimeter to yield 6.24% (which will almost certainly assume a call at par in 2037, but I haven’t checked that), while the ENB 7.22 of 2030 are quoted at 5.73. … so the SLEECS seem kind of expensive to me. However, they will be quoted, traded and indexed as pretend-ten-years, and if anything goes wrong, who cares?

Regulatory Capital

GWO Issuing 30-Year Debs at 5.998%

They’re busy little beavers over at the Great-West Lifeco treasury! Hard on the heels of the GWO.PR.X redemption call, they have unveiled two financing announcements today. The first is the announcement of a 30-year debenture issue:

Great-West Lifeco Inc. (Lifeco) announced earlier today that it had entered into an agreement with a syndicate of agents co-led by RBC Capital Markets and BMO Capital Markets for the sale on an agency basis of $200 million aggregate principal amount of debentures maturing November 16, 2039 (the “Debentures”).

The Debentures will be dated November 16, 2009, will be issued at par and will mature on November 16, 2039. Interest on the Debentures at the rate of 5.998% per annum will be payable semi-annually in arrears on May 16 and November 16 in each year, commencing May 16, 2010, until November 16, 2039. The Debentures are redeemable in whole or in part at the greater of the Canada Yield Price and par, together in each case with accrued and unpaid interest. The Debenture offering is expected to close on or about November 16, 2009. The net proceeds will be used by the Company for general corporate purposes and to augment the Company’s current liquidity position.

The syndicate of agents will include RBC Capital Markets, BMO Capital Markets, CIBC World Markets Inc., Scotia Capital Inc., TD Securities Inc., Merrill Lynch Canada Inc., National Bank Financial Inc., Casgrain & Company Limited, and Desjardins Securities Inc.

They have also announced another issue of long debs – at least, I think it’s a different issue, the press release is not entirely clear – in an exchange offer for Innovative Tier 1 Capital:

Great-West Lifeco Inc. (Lifeco) announced today that it is making an offer to acquire (the “Offer”) up to 170,000 of the outstanding Great-West Life Trust Securities – Series A (“GREATs”) of Great-West Life Capital Trust and up to 180,000 of the outstanding Canada Life Capital Securities – Series A (“CLiCS”) of Canada Life Capital Trust. If more than 170,000 GREATs or 180,000 CLiCS are tendered to the Offer, the tenders will be subject to pro ration.

Lifeco also announced today that it has entered into an agreement with a syndicate of agents co-led by RBC Capital Markets and BMO Capital Markets for the sale of $200 million aggregate principal amount of debentures on an agency basis (the “Debentures”).

Pursuant to the Offer, holders of GREATs and CLiCS will have the opportunity to tender all or a portion of their GREATs and/or CLiCS, as applicable, for the applicable GREATs or CLiCS purchase price payable, at the election of the depositing securityholder, either in (a) cash, or (b) debentures with a term to maturity of approximately 30 years plus cash equal to the amount, if any, by which the GREATs purchase price (described below) or CLiCS purchase price (described below), as the case may be, exceeds the debenture price (described below).

The purchase price for the GREATs will provide a yield on each GREATs to December 31, 2012 equal to the yield of the 2% Government of Canada bond due September 1, 2012, as determined one business day prior to the expiration of the Offer, plus a spread of 1.20%. The purchase price for the CLiCS will provide a yield on each CLiCS to June 30, 2012 equal to the yield of the 3.75% Government of Canada bond due June 1, 2012, as determined one business day prior to the expiration of the Offer, plus a spread of 1.05%. In addition, the debentures to be issued under the Offer will provide a yield to maturity equal to the yield to maturity of a 5% Government of Canada due June 1, 2037 plus an equivalent credit spread to the Debentures to be determined and included in the Offer to Purchase and Circular to be mailed to all holders of the GREATs and CLiCS shortly.

The Company will publicly announce the determination of the purchase prices for the GREATs and the CLiCS as well as other details regarding the debentures offered as consideration payable for the GREATs and CLiCS by way of a news release and will post such release on the Company’s website not later than 5:00 p.m. (Toronto time) on the business day immediately prior to the expiration date of the Offer (such date currently expected to be December 15, 2009).

According to the 2008 Annual Report (which is copy-protected because it’s SECRET), there were $350-million GREATs outstanding, which are Tier 1 Capital of GWL. There was a total of $450-million CLiCS outstanding, $300-million of which were Series A. They disclose that subsidiaries of GWO held $167-million of the total, but do not provide a breakdown of his holding into Series A & B. CLiCS are Tier 1 Capital of Canada Life.

CLiCS were issued in February 2002 with a 6.679% coupon and the Series A were due to mature 2012-6-30 according to the prospectus on SEDAR.

The GREATs were issued in December 2002 and have a pretend-maturity of 2012-12-31 according to SEDAR.

GWO seems to be rejigging its capital structure somewhat! We will see if this is a normal term-extension type of refinancing (the GWO debs will be worse credits than the CLiCs & GREATs), or whether there’s something else cooking….

Regulatory Capital

MFC 3Q09 Results

Manulife Financial has released its 3Q09 results and – as they warned in the 2Q09 release – results were severely impacted by changes in actuarial assumptions:

the Company completed its annual review of all actuarial assumptions in the third quarter. This resulted in a charge to earnings of $783 million, including $469 million due to changes in assumptions of policyholder behaviour for segregated fund guarantee products (a charge that was within the Company’s previously communicated expectations of less than $500 million). The remainder of the charge included assumption changes related to morbidity and other policyholder behaviour, partially offset by assumption changes related to mortality, expenses and investment related items.

The morbidity charge comes as something of a shock, and details are a little skanty:

Driven by increases due to impact from higher projected net long-term care claims costs. Partially offsetting these increases were reductions from mortality releases in Japan and the Reinsurance Division.

They make particular note of the potential for being regulated at the holdco level:

In Canada, OSFI has announced that it (i) will be proposing a method for evaluating stand-alone capital adequacy for operating life insurance companies, such as MLI, (ii) is considering updating its regulatory guidance for non-operating insurance companies acting as holding companies, such as MFC, and (iii) is reviewing the use of internally-modeled capital requirements for segregated fund guarantees. The outcome of these initiatives is uncertain and could have a material adverse impact on the Company or on its position relative to that of other Canadian and international financial institutions with which it competes for business and capital.

They disclose their market risk sensitivity as:

The interest scenario we have adopted uses the structure of the prescribed scenario that currently produces the highest policy liability, which is a gradual decline in market interest rates from current market levels to lower assumed ultimate reinvestment rates over 20 years, with additional prudence introduced through use of lower ultimate reinvestment rates than the maximum levels permitted. The decrease in sensitivity to public equity market values reflects the impact of significantly improved equity markets in 2009, which has both reduced the liability for existing segregated fund guarantees and reduced the sensitivity of this liability to changes in equity market levels. Additional sensitivity reduction resulted from the increase in the amount of business that is hedged. Sensitivity to other non fixed income assets has increased from 2008 due to additional acquisitions of non fixed income assets in 2009 in support of the Company’s long-term investment strategy and the inclusion of the impact of future income taxes.

“Non Fixed Income Assets” are described as:

Other non fixed income assets include commercial real estate, timber and agricultural real estate, oil and gas, and private equities

Private equity, I’m convinced, is a way to dress up equities as bonds, valuing them on the basis of discounted cash flows since they don’t have a publicly quoted market price. Somewhere in the world, for some company, somehow, that masquerade is going to blow up some day. However, MFC is less than forthcoming on just how these investments – and their risks – are valued.

The fact that they will experience a loss due to interest decreases implies that their assets have lower duration than their liabilities.

Various leverage factors may be calculated as:

MFC Leverage
Item	3Q09	2Q09	4Q08
Tangible Common Equity	15,275	16,575	16,482
Bond Exposure	147,056	149,353	149,733
Bond Leverage	963%	978%	908%
Reported Bond Sensitivity	2,000	*	1,300
Bond Sensitivity / Equity	13.1%	*	7.9%
Equity Exposure	10,437	9,688	8,240
Equity Leverage	68%	58%	50%
Reported Equity Sensitivity	1,300	*	1,500
Equity Sensitivity / Equity	8.5%	*	9.1%
“Non-Fixed Income” Exposure	11,510	12,181	12,259
“Non-Fixed Income” Leverage	75.3%	73.5%	74.4
Reported “Non-Fixed Income” Sensitivity	700	*	600
“Non-Fixed Income” Sensitivity / Equity	4.6%	*	3.6%
Tangible Common Equity is Common Shareholders’ Equity including all elements of Other Comprehensive Income less goodwill less intangibles
Bond Exposure is Securities-Bonds plus all elements of Loans
Bond Leverage is Bond Exposure divided by Tangible Common Equity
Reported Bond Sensitivity is the midpoint of the reported effect on earnings of an adverse 100bp move in interest rates, for AFS and HFT bonds taken together.
Equity Exposure is Securities-Stocks
“Non-Fixed Income” Exposure is Real Estate plus Other Investments

I confess that I’m a bit perplexed at their sensitivity reporting. I have taken the sensitivities above from the table “Sensitivity of Policy Liabilities to Changes in Asset Related Assumptions” but in the section headed “Net Income Sensitivity to Interest Rate and Market Price Risk” they state:

The potential impact on net income attributed to shareholders as a result of a change in policy liabilities for a one per cent increase in government, swap and corporate rates at all maturities across all markets was estimated to be a gain of approximately $1,600 million as at September 30, 2009 (December 31, 2008 – approximately $1,100 million) and for a one per cent decrease in government, swap and corporate rates at all maturities across all markets was estimated to be a charge of approximately $2,000 million as at September 30, 2009 (December 31, 2008 –approximately $1,300 million).

… which are the same numbers. Taken literally, this would mean that changes in policy liabilities flow straight through to the bottom line, which would make sense only if their assets were comprised of 100% cash.

The earnings release quotes Chief Financial Officer Michael W. Bell as saying:

As a result of the decline in interest rates and changes in lapse assumptions, our interest rate sensitivity has increased.

Later on, just after the table showing the sensitivities, the release states:

The increase in the sensitivity to changes in market interest rates is primarily due to the impact of the current lower market interest rates on liabilities with minimum interest guarantees and changes in lapse assumptions.

I get the “minimum interest guarantee” part, but am a little fogged by the “changes in lapse assumptions”. I can only assume that they are assuming that this means they are assuming they will get fewer gifts in future from policyholders terminating agreements that are in the policyholders’ favour, but this is not spelt out – and neither is the breakdown between the two major components of the sensitivity.

Their presentation slides include the remark:

Changes in interest rates impact the actuarial valuation of in-force policies by changing the future returns assumed on the investment of net future cash flows

By and large, I’d guess they’re making long-term guarantees backed by short-term investments … the banks’ “maturity transformation” in reverse – and, what’s more, making this a big bet.

Regulatory Capital

SLF 3Q09 Results

Sun Life Financial has released its 3Q09 results and – as they warned in the 2Q09 release – results were severely impacted by changes in actuarial assumptions:

Sun Life Financial Inc.2 reported a net loss attributable to common shareholders of $140 million for the quarter ended September 30, 2009, compared with a net loss of $396 million in the third quarter of 2008. Net losses in the third quarter of 2009 were impacted by the implementation of equity- and interest rate-related actuarial assumption updates of $513 million and reserve increases of $194 million for downgrades on the Company’s investment portfolio. These decreases were partially offset by reserve releases of $161 million as a result of favourable equity markets. Results in the third quarter of 2008 were impacted primarily by asset impairments and credit-related losses and a steep decline in equity markets. Results last year also included earnings of $31 million or $0.06 per share from the Company’s 37% ownership interest in CI Financial, which the Company sold in the fourth quarter of 2008.

They make particular note of the potential for being regulated at the holdco level:

In Canada, OSFI has proposed a method for evaluating stand-alone capital adequacy and is considering updating its current regulatory guidance for insurance holding companies. While the impacts on the life insurance sector are not known, it remains probable that increased regulation (including at the holding company level) will lead to higher levels of required capital and liquidity and limits on levels of financial leverage, which could result in lower returns on capital for shareholders.

They disclose their market risk sensitivity as:

the impact of an immediate 10% drop across all equity markets would be an estimated decrease in net income in the range of $125 million to $175 million.
…
an immediate 1% parallel decrease in interest rates would result in an estimated decrease in net income in the range of $325 million to $400 million. The increase in sensitivity to a downward movement in interest rates from the second quarter of 2009 is primarily due to the implementation of equity- and interest rate-related assumption updates.

The fact that they will experience a loss due to interest decreases implies that their assets have lower duration than their liabilities.

Various leverage factors may be calculated as:

SLF Leverage
Item	3Q09	2Q09	4Q08
Tangible Common Equity	8,272	8,678	8,332
Bond Exposure	81,188	81,565	81,376
Bond Leverage	981%	940%	977%
Reported Bond Sensitivity	???	???	???
Bond Sensitivity / Equity	???	???	???
Equity Exposure	4,710	4,612	4,458
Equity Leverage	57%	53%	54%
Reported Equity Sensitivity	150	237.5	312.5
Equity Sensitivity / Equity	2%	3%	4%
Tangible Common Equity is Common Shareholders’ Equity less goodwill less intangibles
Bond Exposure is Bonds-Held-for-Trading plus Bonds-Available-for-Sale plus Mortgages and Corporate Loans
Bond Leverage is Bond Exposure divided by Tangible Common Equity
Reported Bond Sensitivity is the midpoint of the reported effect on earnings of an adverse 100bp move in interest rates, for AFS and HFT bonds taken together.
Equity Exposure is Stocks-Held-For-Trading plus Stocks-Available-for-Sale

I don’t understand their interest rate senstivity figure for 3Q09. The 3Q09 Earnings Release states:

The estimated impact of an immediate parallel increase of 1% in interest rates as at September 30, 2009, across the yield curve in all markets, would be an increase in net income in the range of $150 million to $200 million. Conversely, an immediate 1% parallel decrease in interest rates would result in an estimated decrease in net income in the range of $325 million to $400 million. The increase in sensitivity to a downward movement in interest rates from the second quarter of 2009 is primarily due to the implementation of equity- and interest rate-related assumption updates.

While the 2Q09 Report to Shareholders states:

For held-for-trading assets and other financial assets supporting actuarial liabilities, the Company is exposed to interest rate risk when the cash flows from assets and the policy obligations they support are significantly mismatched, as this may result in the need to either sell assets to meet policy payments and expenses or reinvest excess asset cash flows under unfavourable interest environments. The estimated impact on the Company’s policyholder obligations of an immediate parallel increase of 1% in interest rates as at June 30, 2009, across the yield curve in all markets, would be an increase in net income in the range of $100 to $150. Conversely, an immediate 1% parallel decrease in interest rates would result in an estimated decrease in net income in the range of $200 to $275.

Bonds designated as available-for-sale generally do not support actuarial liabilities. Changes in fair value of available-for-sale bonds are recorded to OCI. For the Company’s available-for-sale bonds, an immediate 1% parallel increase in interest rates at June 30, 2009, across the yield curve in all markets, would result in an estimated after-tax decrease in OCI in the range of $325 to $375. Conversely, an immediate 1% parallel decrease in interest rates would result in an estimated after-tax increase in OCI in the range of $325 to $375.

Adding the AFS and HFT bond figures for 2Q09 results in an estimate of $525 to $650, which is greater than the estimate in the 3Q09 release, whereas the commentary implies it should be less. It is probable that the 3Q09 figure reflects only AFS bonds, but I’ll wait until the 3Q09 report is available before updating the table.

Just to confuse matters, the 4Q08 earnings release states:

The estimated impact from these obligations of an immediate parallel increase of 1% in interest rates as at December 31, 2008, across the yield curve in all markets, would be an increase in net income in the range of $100 to $150 million. Conversely, an immediate 1% parallel decrease in interest rates would result in an estimated decrease in net income in the range of $150 to $200 million.