Annual report 2020 - KLP Boligkreditt ASKLP BOLIGKREDITT AS

The financial statements for KLP Boligkreditt AS have been prepared in accordance with international accounting standards (IFRS) and interpretations from the IFRS interpretations committee (IFRIC), as adopted by the EU. The Norwegian Accounting Act and the Regulations concerning annual accounts for banks, mortgage firms and finance companies (the Accounting Regulations) contain individual requirements for additional information, which is not required under IFRS. These supplementary information requirements have been incorporated into the notes to the financial statements.

The annual accounts have been prepared on the principle of historic cost, with the following exceptions:

• Financial assets and liabilities (including financial derivatives) are valued at fair value through profit or loss
• Financial assets and liabilities are valued in accordance with the rules on fair value hedging

To prepare the accounts in accordance with IFRS, management has to make accounting estimates and approximate valuations. This will affect the value of the Company’s assets and liabilities, income and expenses recognised in the financial statements. Actual figures may deviate from estimates used. Areas in which discretionary valuations and estimates of material significance to the Company have been used are described in Note 3.

All amounts are presented in NOK thousands without decimals unless stated otherwise.

The financial statements have been prepared in accordance with the going concern assumption.

The most important accounting policies relating to financial instruments are described below.

Cash and cash equivalents are defined as receivables from credit institutions without termination date. The amount does not include receivables from credit institutions that are linked to purchase and sale of securities in the management of the securities portfolios. The statement of cash flows has been prepared in accordance with the direct method.

The Company’s financial liabilities comprise liabilities to credit institutions and covered bonds issued.

The owners’ equity in the Company comprises owners’ equity contributed and retained earnings.

Tax costs in the income statement comprise tax payable and changes in deferred tax. Tax is charged to the income statement, apart from tax relating to items reported under “other comprehensive income”. Deferred tax and tax assets are calculated as differences between the accounting and taxation value of assets and liabilities. Deferred tax assets are capitalized to the extent it can be shown probable that the Company will have sufficient taxable profit to exploit the tax asset.

The company is a part of a financial services group and a tax group. Except for the limitations pursuant to the Financial Institutions Act, any tax-related surplus may be passes in its entirety to the parent company and subsidiaries as a group contribution with tax effect.

The company pays no benefits to employees and is not covered by the rules on capital activity tax. The Company's nominal income tax rate in 2020 is 22 per cent.

Bloomberg is used as a source for pricing Norwegian government bonds. The prices are compared with the prices from Nordic Bond Pricing to revaeal any errors.

Norwegian fixed-income securities except government are generally priced using prices from Nordic Bond Pricing. Those securities that are not included in Nordic Bond Pricing are priced theoretically. In theoretical price is based on the assumed present value on the sale of the position. A zero-coupon curve is used for discounting. The sero-coupon curve is asjusted upwards by means of a credit spread, which is to take account of the risk the bond entails. The credit spread is calculated on the basis of a spread curve taking account of the duration of the bond. Nordic Bond Pricing is the mainsource of spread curves. They provide company-specific curves for Norwegian saving banks, municipalities and energy. Saving banks have various spread curves based on total assets. For companies where Nordic Bond Pricing do not deliver spread curved, the Group use spread curves from three Norwegian banks. When spread curves are available from more than one of these banks, an equal-weighted average is used. If a bond lacks an appropriate spread curve, spread from a comparable bond from the same issuer is used.

Fair value of lending without fixed interest rates is considered virtually the same as book value since the contract terms are continuously changed in step with market interest rates. Fair value of fixedrate loans is calculated by discounting contractual cash flows by the marked rate including a relevant risk margin on the reporting date. This is valued in Level 2 in the valuation hierarchy, cf. note 7.

All receivables from credit institutions (bank deposits) are at variable interest rates. Fair value of these is considered virtually the same as book value since the contract terms are continuously changed in step with change in market interest rates. This is valued in Level 2 in the valuation hierarchy, cf. note 7.

These transactions are valued using a valuation model, including relevant credit spread adjustments obtained from the market. This is valued in Level 2 in the valuation hierarchy, cf. note 7.

Fair value in this category is determined on the basis of internal valuation models based on external observable data. This is valued in Level 2 in the valuation hierarchy, cf. note 7.

Instruments at this level obtain fair value from listed prices in an active market for identical assets or liabilities to which the entity has access at the reporting date. Examples of instruments in Level 1 are stock market listed securities.

Instruments at this level obtain fair value from observable market data. This includes prices based on identical instruments, but where the instrument does not maintain a high enough trading frequency and is not therefore considered to be traded in an active market, as well as prices based on corresponding assets and price-leading indicators that can be confirmed from market information. Example instruments at Level 2 are fixed-income securities priced on the basis of interest rate paths.

Instruments at Level 3 contain non-observable market data or are traded in markets considered to be inactive. The price is based generally on discrete calculations where the actual fair value may deviate if the instrument were to be traded.

Note 6 discloses the fair value of financial assets and financial liabilities that are recognized at amortized cost. Financial assets measured at amortized cost comprise lending to and due to credit institutions, and retail customers. The stated fair value of these assets is determined on terms qualifying for level 2. Financial liabilities recognized at amortized cost consist of debt securities issued. The stated fair value of these liabilities is determined by methods qualifying for level 2.

There have been no transfers between Level 1 and Level 2.

KLP Boligkreditt AS is a wholly owned subsidiary of KLP Banken AS. The Board of Directors of the Bank has established a risk management framework aimed at ensuring that risks are identified, analysed and managed based on policies, limits, procedures and instructions. The Board has adopted risk policies covering the key individual risks as well as an overarching risk policy that covers principles, organisation, limits etc. for the Bank’s total risk. The risk policies are of an overarching nature and are complemented by procedures, guidelines and instructions laid down at the senior management level. The policies state which departments are responsible for handling the various risks and also cover the establishment of a separate risk control function. One purpose of the risk control function is to check that the risk policies and other guidelines for risk management are being followed. This function is carried out by the head of the Risk Analysis and Control Department, which is responsible for preparing periodic risk reports to senior management and the Board as well as reporting on any breaches of policies or guidelines. The Department, which has an independent role in relation to other departments, also has other tasks associated with the Bank’s risk management. The responsibility for the operational direction of the Bank’s liquidity risk, exchange rate risk and interest rate risk lies with the Finance Department.

The Board has adopted a policy for credit risk which contains overarching guidelines, requirements and limits associated with credit risk. The policy states that KLP Boligkreditt AS should have a low credit risk profile and includes limits on types of lending and principles for the organisation and operation of the Bank’s lending activity. The policy also includes an overarching mandate structure for lending and other counterparty exposure.

Credit risk associated with issuers of securities, derivative counterparties and other counterparties in the financial area is also limited by Board-determined limits on individual counterparties. These limits are based on the counterparty’s solvency and other assessments of counterparties’ creditworthiness.

KLP Boligkreditt only provides loans mortgaged in residential property within 75% of the market value of the mortgaged object. In processing loan applications the borrower’s servicing ability and the value of the mortgage object is assessed and loans are provided only within set limits and authorisations.

The market value of the mortage assets is updated quarterly against Eiendomsverdis of market value of housing in Norway.

KLP Boligkreditt uses a risk classification system to classify retail customers with loans or credits. Customers are classified from A to K, where A indicates very low risk while K is for customers on which the bank has incurred losses. Below is a distribution table with the volume of loans divided into low, medium and high risk, where low risk is defined as lending to customers in class A or B, medium risk is defined as lending to customers in class C or D, and high risk is defined as lending to customers in classes E to K.

The table below shows the total book value of the various risk classes and per stage in the impairment model. Stage 1 is all healthy loans, which must be written down by the estimated losses for 12 months. Stage 2 indicates that the exposure has a substantially increased credit risk since its initial recognition on the balance sheet, and means that the loan must be written down by the estimated losses throughout the entire term. Stage 3 is all loans in default (over 90 days past due) or with individual loss write-downs, which must be written down by the estimated losses throughout the entire term.

The Company also invests in securities issued by municipalities and county administrations and will in addition have credit risk exposure in the form of “additional collateral”. The additional collateral can amount up to 20 percent of the cover. In accordance with the Company’s internal guidelines the additional collateral may be in the form of deposits in banks satisfying minimum rating requirements as well as covered bonds issued by Norwegian credit enterprises.

The Company may also be exposed to credit risk as a result of derivative agreements struck. The purpose of such agreements is to reduce risks arising as a result of the Company’s borrowing and lending activities. The Company’s internal guidelines specify creditworthiness requirements for derivative counterparties. As of 31 December 2020, KLP Boligkreditt AS had no derivative agreements with any counterparties.

KLP Boligkreditt AS measures maximum exposure as principal and accrued interest. Security in cash or securities is not exchanged, nor are other credit improvements carried out. The table below shows the maximum exposure for KLP Boligkreditt AS.

The Company has not incurred losses on lending. The company considers all receivables to be satisfactorily secured. The loans are secured with mortgages within 75% of market value, and losses will only arise when the value of the mortgaged object falls below the residual amount of the loan.

The Company's lending is in its entirety linked to financing of real estate where the loans are secured with mortgages within 75% of the market value. All borrowers are Norwegian and the collateral is in Norwegian. The company has a risk consentration where it is exposed to a general impairment in the Norwegian housing market.

The concentration against individual borrowers is limited by individual Board-set limits.

KLP Boligkreditt AS’s largest exposure as at 31 December 2020 was about 0.1 per cent of the Company's total lending.

Interest rate risk is measured as change in value on a one percentage point change in all interest rates.

The market risk policy is the Company’s overarching guidelines, requirements and limits associated with market risk. The policy dictates that the market risk should be minimized so that the total market risk is low. It further states that the Company should not actively take positions that expose the Company to market risk. The policy also sets limits for interest rate risk both for the total interest rate risk for the infinite future and for rolling 12-month periods. The risk limits are set to ensure that the low market risk profile policy is complied with. The operational responsibility for managing the Company’s market risk lies with the Finance Department in KLP Banken. The Risk Analysis and Control Department reports the Company's actual exposure in relation to limits in accordance with guidelines set by the Board.

Interest rate risk arises because the fixed interest periods for the bank’s assets and liabilities are not the same. The table below shows repricing dates for the Company’s interest-bearing assets and liabilities, and the gap shows the difference between assets and liabilities that can be interest-adjusted within the given time intervals. Lending at variable rates, and cash and receivables with credit institutions, are assumed to be able to be repriced within a 1-month horizon. The debt falls into the time interval for which interest adjustment has been agreed.

The management of KLP Boligkreditts’s liquidity risk must be viewed in the context of the management of the liquidity risk in the KLP Banken Group. A liquidity policy is established for the Group containing principles, guidelines, requirements and limits that apply to the management of the liquidity risk. The policy states that the Group is to have a moderate liquidity risk profile and various requirements and limits have been set to achieve this, including targets for deposit cover, limits for refinancing requirements for various time periods and requirements for liquidity buffers. The Board has further adopted an emergency plan for liquidity crises as part of the liquidity policy. In addition to the requirements at Group level, separate specific requirements have been established for KLP Boligkreditt, including requirements for continually positive cash flows, limits for refinancing requirements and requirements for liquidity reserves and drawing rights. The operational responsibility for managing the Company’s liquidity risk lies with the Finance Department. The Risk Analysis and Control Department reports the Company's actual exposure in relation to limits in accordance with guidelines set by the Board.

The tables below show the maturity analysis of the Company’s assets and liabilities including stipulated interest rates.

The new accounting standard IFRS 9 was introduced on 1 January 2018 and changed the methodology for provisions for losses on financial instruments in the accounts. The measurement of the provision for expected credit losses on financial assets depends on whether the credit risk has increased significantly since initial recognition. At initial recognition and if the credit risk has not increased significantly, the provision should equal 12-month expected credit losses (stage 1). If the credit risk has increased significantly from the initial recognition (stage 2) or if the asset is classified as impaired (stage 3), the provision should equal lifetime expected credit losses.

Expected credit loss (ECL) is calculated as the exposure at default (EAD) multiplied by the probability of default (PD) multiplied by the loss given default (LGD).

Probability of Default (PD) is a calculated probability based on statistical models to estimate the probability of an exposure going into default during the following 12-month period (12-month PD). In addition to calculating 12 months PD, the bank has developed PD graphs used for calculating marginal PD for the exposure’s remaining lifetime (Lifetime PD).

Loss given default (LGD) is what the bank expects to lose given that an exposure goes into default. The calculation is based on how probable it is that a defaulted exposure is cured and expected credit loss if the exposure is not cured.

Exposure at default (EAD) is expected exposure at the moment of a future default.

In KLP Boligkreditt AS, the assessment of what is considered to be a significant change in credit risk for retail mortgage loans is based on a combination of quantitative and qualitative indicators and ‘backstops’. The most important driver for a significant change in credit risk for retail mortgage loan is a change in the probability of default (PD) from the initial recognition up to the reporting date. A relative change in 12 month PD of more than 2.5 is considered a significant change in credit risk. In addition, the change in 12 month PD must also be at least 0.6 percentage points for the change to be considered significant. Exposures that are more than 30 days past due will automatically be placed in Stage 2, and exposures more than 90 days past due will be placed in Stage 3. The loans go back to Stage 2 and Stage 1 when the criteria for significant change in credit risk and default are no longer fulfilled.

Default is defined as "a claim that is over 90 days past due, or an account that has been continuously overdrawn for a minimum of 90 days (minimum amount NOK 1,000). Furthermore, a commitment is considered as default if for various reasons it has been written off, e.g. through debt negotiations, established debt settlement and/or bankruptcy.

Mortgages in arrears are handled by a special commitments department in the bank, Banken Group currently uses its own collection process up to and including legally enforced recovery and execution of sale/foreclosure. If a repayment agreement is not reached, any residual debt after realisation of the collateral is transferred to a collection agency for further follow-up.

Mortgages over 90 days past due are reviewed and followed up regularly. In addition, exposures are also reviewed when the bank receives information about debt negotiations or other conditions that would indicate increased risk. A loss assessment is carried out for all such exposures. The collateral is assessed on the basis of previously determined value, in addition to new information about the bank's collateral in the case, for example from a broker if a sale/foreclosure has already been initiated. If the realisation value proves to be lower than the residual debt of the commitment, a loss write-down of the exposure is carried out.

Exposures with individual loss write-downs are followed up with a view to the realisation of the collateral. This can be undertaken by agreement on an ordinary sale or legally by means of foreclosure. In some cases, a payment agreement to repay the full amount of residual debt is reached. In these cases, the loss write-down will be maintained for a minimum of 1 year after the loan has been satisfactorily served, before the commitment is considered cured.

For mortgages, the determination of loss will only occur after the security has been realised and further legal proceedings have not succeeded, that is after an application for distraint has not yielded a result. The case is then monitored by a debt collection agency and followed up on a regular basis.

In connection with the transition to IFRS 9 and new methods for loss calculation, KLP Banken has developed PD and LGD models for the bank's/group's mortgage loan portfolio. A PD model has been developed for new mortgage customers and a PD model for existing mortgage customers. The first model uses data that is available at the time of application and is valid for 3 months after the mortgage is granted. The second model begins after 3 months, and also include data that depends on the customer's behaviour (for example the number of days in arrears). Explanatory variables are age, income, number of reminders sent in the last 12 months, total number of days in arrears in the last 12 months, loan-to-value ratio, co-borrower, defaults in the last 12 months and product type.

Logistical regression was used to create the PD model. This method is considered an industry standard for PD models, it is easy to interpret and analyse the output from the model and it can provide high coefficient of determination given that certain assumptions are met. The method also makes it possible to combine pure quantitative analyses with expert assessments, which was useful when the data base was somewhat limited. A thorough manual analysis of a relatively small sample of potential variables (due to limited data basis) was carried out to arrive at an optimal combination of variables.

The most important measure for a PD model is the model's ability to discriminate, i.e. the ability to distinguish bad customers from good customers. The ability to discriminate is measured using ROC (Receiver Operating Characteristic), which provides some information about the proportion of predictions that are correct. The model is validated or recalibrated at least yearly and the coefficient values can then be updated and the prediction level adjusted.

The lifetime probability of default (Lifetime PD) is used for all retail mortgage loans in KLP Banken Groups excluding senior loans. The lifetime probability of default (LTPD) of an exposure is calculated based on aggregated figures for historically observed default rates for each year of all exposures and each exposure's probability of default 12 months after start. The results from model development show that the default rate increases slightly in year 2 before then decreasing, so that the PD in year 2 is higher than in year 1. This is in line with the expected result, since it is expected that it will take some time before a newly granted mortgage loan experiences problems. A customer will typically seek to avoid default on the mortgage loan, and will typically default on other debts before he goes into default on the mortgage loan. The reduction in PD after year 2 can be explained by a "survivalship effect", i.e. the contracts that have not defaulted in the first 2 years are typically of better credit quality, and as the loans are repaid the risk becomes lower. Experience from the industry is that contracts that have existed for a certain period of time converge towards a stable observed default rate. For KLP Banken/Group's mortgage loan portfolio, 3 years has been set as the parameter for when the default level converges towards a long-term PD level. The long-term PD level is set at 0.3 per cent, which corresponds to the average PD for the best contracts in the portfolio.

Exposure at default (EAD) is used for all mortgages in KLP Banken/Group excluding senior loans. The EAD model has the same data sample as the LTPD model. If an exposure is in default, the exposure's balance at the time will be the bank's/group's exposure exposure at default. EAD can be expressed for an exposure as a function of the likelihood that the contract will not be repaid within the time t. For repayment loans, EAD at time t is estimated as the exposure's balance at the time pursuant to the repayment schedule multiplied by the likelihood of the contract not being repaid within time t. The probability of a contract being terminated early within the year t is calculated as a percentage for each year in the future from 1 to 7 years.

When estimating future credit loss it is important to look at the proportion of customers in default whose accounts become cured. The bank/group has examined at all historical defaults over 90 days and analysed the outcomes of these defaults. The results of the analysis show a very high level of defaults becoming cured. KLP Banken/Group has, since its inception, handled defaults and debt collection internally within the bank/group, and has one dedicated employee who handles exposures in default. The cases are followed closely, and there has been a limited number of defaults since the bank's inception. The analysis shows that the bank has had minimal losses, and most defaults have been reported as cured.

Cured default is defined as the account returning to ongoing status (no longer 90 days past due/90 days in arrears over the bank's significant amount), or that the account is terminated without loss (typically through voluntary sale of collateral or refinancing in another bank). Non-cured default is defined as where the recovery process has resulted in the account having an established loss, or that an application for distraint has been made against the customer (foreclosure of the property or recovery of guarantee). Customers with status "nothing for distraint" also belong in this category). If the customer has entered into debt negotiations, this is also defined as non-cured default. One last possibility is that we do not know the final outcome of the default due to a short time horizon between the default date and modelling date. The figure below illustrates the various outcomes for a default.

The observed cure rate is calculated and validated at least yearly in the same way as it during model development. If the observed cure rate deviates by more than 10 percentage points from the estimate used in the IFRS 9 model, an assessment shall be made of whether measures are needed, e.g. a re-estimation of the model.

A part of the assessment of future losses is the assessment of how the future will look regarding to the macroeconomic conditions that affect the bank's credit losses, e.g. interest rates, housing prices, unemployment rates etc. To calculate the expected credit loss (ECL), the bank has assumed three different scenarios, which are weighted for probability based on an assessment of the probability of each of the three outlined scenarios occurring. The scenarios used by the bank are one expected outcome, one pessimistic outcome and one optimistic outcome for expected credit loss, where the three scenarios have a factor for outcome and a probability that the scenario occurs. The sum of the weighted scenarios constitutes the expected credit loss, and the probability that each scenario will occur will thus affect the expected credit loss. In the negative scenario, a house price fall of 10 per cent and an increase in average PD of 32 per cent are assumed, while the number of cured defaults decrease with 5 per cent. This scenario is given a 30 per cent weight. In the positive scenario it is assumed that house prices increase with 5 per cent and that the PD is halved. The positive scenario is given a 10 per cent weight. The expected scenario is thereby given a 60 per cent weight.

KLP Banken's risk forum assesses these scenarios and their weighting on a quarterly basis, based on changes in macroeconomic factors or other factors that may affect expected credit loss in the bank.