Regulatory insight. Analysis of government-related private loans under FTK, Solvency II and Basel III. 6 November 2017

Transcription

1 Regulatory insight Analysis of government-related private loans under FTK, Solvency II and Basel III 6 November

2 Analysis of government-related private loans under FTK, Solvency II and Basel III In this article we give an overview of the treatment of government-related private loans under different regulatory frameworks. Our focus is on capital requirements for institutional investors who invest in these loans and who are subject to FTK, Solvency II or Basel III regulations. We discuss the stand-alone capital requirements, but also the impact of these requirements on the overall balance sheet and solvency position. This provides more insight into the relative attractiveness of government-related private loans, compared with other asset classes, for institutional investors. We can conclude that this asset class has potentially a very low capital requirement under the different regulatory frameworks. Under FTK, the capital requirement is limited due to the underlying government guarantees, which leads to low expected losses. Under Solvency II and Basel III, almost all government-guaranteed loans are exempted from capital charges. This makes this asset class attractive from a capital point of view. In addition, the more illiquid nature of these loans is also translated into additional returns compared to liquid sovereign or credit bonds with a similar credit quality. Government-related private loans: an introduction We here consider highly-rated private placements and bonds of predominantly European public entities. A private placement is a debt agreement between a borrower and a lender, where the loan is not stock exchange listed. Through a private contract, the debt obligation is acknowledged by the debtor. Private placements are less liquid than traditional fixed income instruments, but compensate for this with attractive additional returns. We focus in this article on highly-rated private placements to, or guaranteed by, public agencies, both in the Netherlands and abroad. Borrowing entities typically include multinational institutions, social housing associations, hospitals, municipalities and regional authorities. In the Netherlands, the largest lenders are the NWB Bank 1 (which specializes in arranging loans for the Dutch public sector), the Dutch Municipal Bank (BNG) 2 and a few large institutional investors. More recently, private placements via an export credit agency (ECA) have also experienced an upswing. Examples are Atradius in the Netherlands, Euler-Hermes in Germany or Coface in France. These ECAs provide trade financing to domestic companies for their international activities and are fully guaranteed by their respective central governments. FTK, Solvency II, Basel III Various regulatory frameworks exist for different institutional investors. For example: banks face Basel III regulations, European insurance companies face Solvency II and Dutch pension funds FTK. Similar to Basel, three distinct pillars are used under FTK and Solvency II in order to structure the regulatory process: (i) capital requirements, (ii) governance & supervision and (iii) reporting & disclosure. This article focuses on the first pillar (i.e., capital requirements). Capital requirements can be determined using a standard model approach or using an internal model. We use the standard model approach in this article. In practice, capital requirements will vary for different institutional investors due to: 2

3 Another risk tolerance (e.g., a 1 in 200 year probability of a negative surplus under Solvency II versus 1 in 40 years under FTK); Another balance sheet structure (e.g., short funding for property & casualty or health insurers versus long funding for pension funds and life insurers); Another historic development over time of the various regulatory frameworks. We start with an overview of the expected returns used in this article. We then give an overview of the treatment of government-related private loans for (i) Dutch pension funds (under FTK), (ii) European insurance companies (under Solvency II) and (iii) banks (under Basel III). We start these sections with a global overview of the required capital calculation for the entire balance sheet. We then give more information about the stand-alone required capital calculation for government-related private loans. We conclude with the impact of investing in private loans on the overall required capital and expected return. Expected returns The expected returns in this article are according to the latest long-term economic forecasts by Aegon Asset Management in The Netherlands. 3 This forecast is for the period We use the average expected return over this four-year period. We consider three scenarios: a base scenario, a positive scenario, and a negative scenario: In the base scenario, to which we attach a probability of 60%, we expect that economic growth in the US leads to a tighter job market, resulting in higher wage inflation. This translates into higher price inflation and gradually increasing interest rates. Higher interest rates will put pressure on corporate profitability due to higher funding costs. It will also lead to a decline in asset values and a fall in consumer confidence and spending power. As a result, we expect a minor slowdown of the US economy and a short-lived recession at some point in the coming years. It seems unlikely that the Eurozone will be able to withstand our predicted US slowdown. A major recession is, however, not likely, so emergence monetary measures are not expected from the ECB in the coming years. In the negative scenario, with a probability of 15%, the economy continues to grow aggressively, leading to overheating and inflation exceeding the central bank targets. We expect that the Federal Reserve and the ECB will then increase interest rates to counter high inflation. This fast and abrupt tightening of monetary conditions will reduce investments, and cause consumer confidence to fall. Also, the combination of higher corporate leverage and increased interest rates will have a strong effect on the profitability of corporations which will in turn cause an economic recession around 2020, much more severe than the recession we predict in our basis scenario. In the positive scenario, with a probability of 25%, we expect economic momentum to continue until the end of 2021 due to productivity gains. In this case, we do not foresee a recession during the forecasting period. Also, inflation will increase but will remain at acceptable levels. This allows central banks to normalize monetary policy cautiously without imposing negative effects on the economy. We assume in these economic forecasts that government-related private loans generate an additional return of 0.8% compared to core Eurozone government bonds. 3

4 Capital requirements for Dutch pension funds: FTK FTK: general capital requirements The set of regulatory rules for Dutch pension funds is commonly referred to as the Financieel Toetsingskader (Financial Assessment Framework) or FTK. Under FTK rules, the solvency capital of a pension fund should be sufficient to avoid a nominal funding ratio of less than 100% with a probability of 97.5% (i.e., 1 in 40 years) over a one-year horizon. This risk tolerance has been translated into the following capital requirements. Table 1: Overview of current FTK stress test parameters 4 FTK capital charges Risk factor Risk event FTK shock Interest rate risk (S 1) 5 Downward/upward interest rate shocks 0.75 (for a 15-year rate decrease) Equity risk (S 2) 6 Downward shock for developed equity markets 30% Downward shock for emerging equity markets 40% Downward shock for private equity 40% Downward shock for non-listed real estate 15% Currency risk (S 3) 7 Downward currency shock 15% Commodity risk (S 4) Downward commodity shock 35% Credit risk (S 5) Upward spread shock for AAA rating (Euro gov.) 0% Upward spread shock for AAA rating 0.6% Upward spread shock for AA rating 0.8% Upward spread shock for A rating 1.3% Upward spread shock for BBB rating 1.8% Upward spread shock for BB rating 5.3% Actuarial risk (S 6) Shock for actuarial risk of the liabilities No standard model Liquidity risk (S 7) 8 Shock for liquidity risk (by default 0) No standard model Concentration risk (S 8) Shock for concentration risk (by default 0) No standard model Operational risk (S 9) Shock for operational risk (by default 0) No standard model Active equity risk (S 10) 9 Shock for active equity risk Formula Dutch Central Bank Table 1 summarizes the FTK stress test parameters for each risk factor. It should be noted that there is no prescribed recipe to determine the required capital for actuarial risk, liquidity risk, concentration risk and operational risk (S6 - S9), but pension funds should hold additional capital for these risks if they are material. Active equity risk (S10) only applies for active equity portfolios with an ex-ante tracking error larger than 1%. The Dutch Central Bank has indicated how the S10 shock can be calculated in this case, using a formula which depends on the tracking error and the total expense ratio. 10 The correlations between the risk scenarios are shown below. These correlations are used to aggregate the capital requirements for the risk scenarios (S1 through S10) and determine the overall capital requirement. 4

5 Table 2: Overview of stress test correlations under FTK 11 FTK correlations Correlation Risk factor Risk factor Current FTK ρ 1 2 Interest rate risk (S 1) Equity risk (S 2) 0.4 ρ 1 5 Interest rate risk (S 1) Credit risk (S 5) 0.4 ρ 2 5 Equity risk (S 2) Credit risk (S 5) 0.5 All correlations which are not mentioned in Table 2 are equal to zero. Note, however, that correlations within the credit (S5) module are assumed to be 1 (e.g., the correlation between AAA and BBB credit bonds). The overall required capital can then be calculated, using the strategic weights of the assets. The following S-formula is used for this purpose to determine the total capital requirement: 12 As a final step, the available assets in the calculation are increased or decreased until the available capital becomes equal to the required capital. 13 This situation is typically reached in a few iterations. The advantage of this approach is that the required capital is not depending anymore on the actual funding ratio of the pension fund. The required capital thus becomes a measure for the strategic risk profile of the pension fund, when the available capital equals the required capital. FTK: capital requirements for government-related private loans Government-related private loans must be treated using the credit risk module (the S5 stress test in Table 1) under FTK. In this case, the applied spread shock depends on the rating. Note that we cannot apply a spread charge of zero for the AAA government-guaranteed loans. A zero spread charge is only applicable for AAA euro sovereign debt, not for debt guaranteed by AAA-rated euro member states. An AAA-rated private loan thus receives an upward spread shock of 0.6%, whereas a BBB-rated private loan will receive a spread shock of 1.8%. An example is given in Table 3 for a typical mandate for government-related private loans. 14 This table shows that most private loans for this example have an AAA rating, followed by AA. No investments in ratings below BBB are allowed in this mandate. This is an attractive feature from an FTK perspective, due to the severe stress test for ratings below BBB (a 530 basis points spread increase). Table 3: Breakdown of the FTK capital requirement for an example mandate of government-rated private loans Rating Weight Spread shock Spread duration Capital charge AAA 83.9% 0.6% % AA 16.1% 0.8% % Total 100% % In the last column of Table 4 we shown the contribution of each rating class to the total capital requirement for this example mandate. The securities with an AAA rating contribute most to the overall capital charge, due to the high weight for this rating class. 5

6 Note that the interest rate sensitivity should also be taken into account under FTK, using the S1 stress test in Table 1. On the total balance sheet of a pension fund, the interest rate sensitivity of these loans will, however, offset part of the interest rate sensitivity of the liabilities. The overall effect on the capital requirements is thus positive. FTK: impact of government-related private loans on overall required capital As an example, consider a pension fund with 10bn assets and liabilities. The stylized asset mix is as follows: 15 50% in equities (developed): required capital (equity risk) is 30% 35% in AAA/AA euro sovereign bonds (duration of 7.5 years): required capital (spread risk) is 1% 16 15% in investment grade euro credit bonds (duration of 4.5 years): required capital (spread risk) is 9% We assume for simplicity that interest rate risk and currency risk are hedged on the balance sheet; these risks therefore do not lead to additional capital charges in this example. 17,18 Note that we only have two (relevant) market risks in this example: equity risk and spread risk. Equity and spread risk have a correlation of 0.5, see Table 3. The required capital for market risk then becomes 1.592bn. 19 For this pension fund, we can now easily calculate the overall required capital using the S-formula. When we assume that the required capital for actuarial risk (the S6 component) is equal to 0.25bn, the overall required capital becomes equal to 1.611bn. 20 As the final step, we change the total amount of assets in the calculation until the available capital becomes equal to the required capital. In the first iteration, we set the available capital equal to the required capital by increasing the total asset value from 10bn to bn. The asset mix is kept constant (50% equities, 35% sovereigns and 15% credits). If we recalculate the required capital, we then arrive at 1.865bn. We subsequently increase the total asset value to bn, which leads to a required capital of 1.905bn. By repeating this approach a few more times, the required capital converges to 1.913bn. We can now investigate the effect of adding government-related private loans to the asset mix. As mentioned above, the required capital for this asset class is calculated in the spread risk (S5) module. As an example, we consider the example government-related private loan mandate that has been discussed in the previous section. The required capital for spread risk is thus 4.7%. We allocate 10% of the assets to government-related private loans and study the effect on capital requirements and expected return. The results are shown in Table 4. 6

7 Table 4: FTK - Impact on required capital, required funding ratio and expected return when allocating assets to government-related private loans Table 4 shows that the required capital increases only slightly (from to 1.943bn, so by 2%) when we sell sovereigns and buy government-related private loans. When we sell credits and buy government-related private loans, the required capital decreases by 2%. Funding from all other asset categories also leads to a substantial (10%) decrease in the required capital. An even stronger effect is visible by substituting equities with government-related private loans. This is due to the low risk charge for this asset class (4.7%, instead of 30% for developed market equities). Similar effects are visible in terms of the required funding ratio. Comparing the expected return for the different portfolios, we see a positive effect (of 0.04% %) in the base scenario when substituting sovereigns or credits with government-related private loans. Substituting equities with government-related private loans leads to a decreasing expected return in the base or positive scenario, but also a much lower required capital, as mentioned above. Funding from all asset categories leads to results falling somewhere in between. FTK: conclusions The required capital for spread risk is limited for government-related private loans under FTK ( 4.7%). The main reason is that these loans are backed up by the government, leading to low expected losses. The overall required capital under FTK therefore decreases if government-related private loans are funded from credits or other (risky) assets. At the same time, the expected return will typically increase when we invest in government-related private loans and use sovereigns or credits as funding, making an investment in this asset class also attractive from a return perspective. Table 5 gives a schematic overview of these findings. Government-related private loans are thus clearly an attractive alternative for sovereign bonds or credits. 7

8 Table 5: Stylized overview of FTK results 21 We now turn our attention to the capital requirements for government-related private loans for European insurance companies under Solvency II. Capital requirements for European insurance companies: Solvency II Solvency II: general capital requirements 22,23 As of January 1, 2016, Solvency II is the supervisory framework for European insurance companies. Under Solvency II, the required capital is the amount of capital that an insurance company should hold to be able to withstand a severe stress scenario (occurring once every 200 years). The required capital is equal to the basic solvency capital requirement (BSCR) plus the required capital for operational risk (Op) and an adjustment for the loss absorbing effect of technical provisions and deferred taxes (Adj), see Figure 1. Figure 1: Structure of the overall required capital calculation 24 The BSCR is the solvency capital requirement before any adjustments and combines the capital requirements for six major risk categories. Note that these risk categories are typically broken down into more categories (e.g. market risk is sub-divided into interest rate risk, equity risk, property risk, etc.). 8

9 The capital charges associated with each of these risks are combined using a correlation matrix, which is prescribed under Solvency II. 25 The purpose of this correlation matrix is to produce an adequate overall capital charge for a oneyear time horizon. In practice, the correlation between the different components of the BSCR is assumed to be small. For example, the probability of a (1 in 200 year) tail event occurring simultaneously for market risk and life risk is assumed to be quite low (a correlation of only 0.25). Due to the low correlation between the major risk categories it is possible to achieve a significantly lower required capital by diversifying risk. Mortgages are in fact a very good required capital diversifier, because this asset class falls under the counterparty default risk module instead of the market risk module. Market risk typically contributes heavily to the overall required capital, whereas the contribution of default risk is much smaller. 26 It therefore pays to transfer risk from market risk to default risk in order to better exploit the low correlation (of 0.25) between these modules. An example of this approach is given in Van Bragt (2017). Solvency II: Capital charge for private loans without a government guarantee The Solvency II capital charge for spread risk for private loans without a government guarantee is depending on the rating and the (spread) duration, see Table 6 below. 27,28 Table 6: Spread risk factors for bonds under Solvency II. N.B. A credit quality step of 0 corresponds to an S&P AAA rating; 1 corresponds to AA; 2 to A; 3 to BBB; 4 to BB; 5 to B and 6 to CCC or lower. As an example, we consider a private loan portfolio with 70% AAA-rated loans and 30% AA-rated loans. When available, the instrument s rating is based on the assessment of a credit rating agency. When an instrument does not have a credit rating, a rating will be determined using an internal rating methodology. 29 The (spread) duration of the loans in this example portfolio is approximately 7. The capital charge (for spread risk) for AAA-rated loans is thus 4.5% + 0.5%*(7-5) = 5.5%, see Figure 2 (use a credit quality step of 0 for AAA). For the AA-rated loans the capital charge is 5.5% + 0.6%*(7-5) = 6.7% (use a credit quality step of 1 for AA). This implies a capital charge of the entire portfolio of 70%*5.5% + 30%*6.7% = 5.9%. 9

10 Solvency II: Capital charge for private loans with a government guarantee Private loans with a government guarantee are typically exempted from a (spread) capital charge. To be precise, no capital requirement for spread risk applies to loans by or demonstrably guaranteed by: The national government of a state of the European Economic Area (EEA), issued in the currency of the government; 2. A multilateral development bank, such as the European Investment Bank or the European Investment Fund; An international organization, such as the European Union, the International Monetary Fund, the Bank for International Settlements, the European Financial Stability Facility, or the European Stability Mechanism; The European Central Bank. To qualify for a zero spread shock, a guarantee by one of the above counterparties should be fully, unconditionally and irrevocably. Government-backed loans by credit export agencies are also exempted from a spread capital charge. 33 In addition, EIOPA has published a list of regional governments and local authorities, exposures to whom are to be treated as exposures to the central government of the jurisdiction in which they are established. 34 For example, for the Netherlands, any province, water board or municipality is on this list. 35 These direct exposures are capital exempted. However, loans guaranteed by these regional governments and local authorities, so indirect exposures, are not exempted from a capital charge. 36 Loans that are guaranteed by so-called guarantee funds are capital exempted when loans are fully, unconditionally and irrevocably guaranteed by the central government. 37 Two important guarantee funds in the Netherlands are the Stichting Waarborgfonds voor de Zorgsector ( WFZ ) and the Stichting Waarborgfonds Sociale Woningbouw ( WSW ). The WFZ fund provides loans for the health sector and is fully guaranteed by the central Dutch government. WFZ loans can thus be treated as solvency free (in the spread risk module) under Solvency II. The WSW fund provides loans for the social housing sector and is partly guaranteed by the central Dutch government and partly by the decentralized government (municipalities). In the Netherlands, this has led to an interesting discussion related to the particular guarantee structure of the WSW fund, see the boxed text below. 10

11 Which guaranty matters most? The case of WSW. In the Netherlands, the Stichting Waarborgfonds Sociale Woningbouw ("WSW") backs up private loans to social housing corporations. This entity is guaranteed for 50% by the central government and for 50% by municipalities. Since only the guarantee of the central governments counts under Solvency II, and this guarantee is not fully, one could argue that these loans are not exempted from a spread charge under Solvency II. However, the central government of the Netherlands (hereafter: the State ) will ultimately (i.e. in the course of time) have to act as effectively the complete back-stop for the loans guaranteed by WSW. After all, the State will have to continue issuing interest-free loans each month for 50% of WSW's liquidity deficit, until the deficit is removed at WSW, and irrespective of whether the municipalities have (also) issued interest-free loans to WSW in that regard. So, even if the municipalities were not issuing any interest-free loans to WSW - whether or not in breach of their contractual obligations - the State would still be under an ongoing obligation to keep issuing interest-free loans for half of WSW's liquidity deficit each month, until the deficit has been halved so often (on a monthly basis) by the State that the liquidity problem has actually been solved. WSW states that, based on this line of reasoning, DNB has decided that WSW loans can be treated as solvency free (in the spread risk module) under Solvency II (see WSW, 2016). Note that, in principle, the State can almost, but never entirely, cancel a liquidity deficit; after all, a monetary amount can never be halved to zero. For example, let assume the WSW runs a deficit of 10 mln. Euro. The State and the municipalities should then both issue 5 mln. Euro of interest-free loans. If the municipalities do not provide this funding, the remaining deficit will then be 5 mln. Euro one month later. The State then again funds 50% (2.5 mln. Euro). If the municipalities again do no provide funding, the remaining deficit will become 2.5 mln. Euro, and so on. Source: Legal opinion by Van Doorne (2015) on this matter, as requested by WSW. Solvency II: impact of government-related private loans on overall required capital As an example, let s consider a life insurer who has invested 10 bln. The asset mix is as follows: 10% in Dutch mortgages: SCR (default risk) is 5% 38 20% in equities (developed): required capital (equity risk) is 39% % in core euro sovereign bonds: required capital (spread risk) is 0% 41 40% in A/BBB-rated euro credit bonds (duration of 4.5 years): required capital (spread risk) is 9% 42 We now study the effect of adding government-related private loans to the asset mix of this insurer. Our example private loan portfolio has an SCR of 0.5%. 43 The slightly positive SCR is be caused by an investment in a loan guaranteed by a municipality. Such an indirect exposure to a local government is not solvency free, as we discussed above. Note that without guarantees the capital charge for our example portfolio would be much higher ( 6%). The underlying guarantees thus dramatically reduce the SCR. 11

12 In the analysis, interest rate risk is assumed to be fully hedged with an interest rate swap overlay, so the SCR for interest rate risk is equal to 0. All assets and liabilities are assumed to be denominated in euros, so the SCR for currency risk is 0. The assets are assumed to be well spread over different counterparties, so the SCR for concentration risk is also 0. We can now calculate the total SCR for market risk, using the correlation matrix in Table 7. Table 7: Correlation matrix for the market risk SCR calculation 44,45 Note that we only have two (relevant) market risks in this example: equity risk and spread risk. Within the market risk module, equity risk and spread risk have a correlation of The SCR for market risk thus becomes bln. 46 The SCR for default risk, due to the exposure to mortgages, is equal to 10 bln.*0.1*0.05= 0.05 bln. Assume also that the SCR for life underwriting risk is equal to 0.35 bln. 47 Neglecting operational risk and the loss absorbing capacity of technical provisions and deferred taxes, the overall required capital then becomes bln. As an example, we allocate 10% of the assets to government-related private loans and study the effect on capital requirements and expected return. The results are shown in Table 8. 12

13 Table 8: Solvency II - Impact on required capital, solvency ratio and expected return when allocating assets to government-related private loans 48 Table 8 shows that when we sell sovereigns and buy government-related private loans, the required capital remains almost the same. We have only a slight increase due to the solvency charge of 0.5% for government-related private loans. When funding government-related private loans with mortgages, the required capital decreases slightly (with 1%). When we sell credits and buy private loans, the required capital decreases more significantly (with 6%). This is due to the low risk charge for private loans in our example (0.5%, instead of 9% for credits). Funding private loans from all asset categories also leads to a substantial (8%) decrease. An even stronger effect is visible by substituting equities with private loans. Comparing the expected return for the different portfolios in Table 8, we see a positive effect (of 0.08%) when funding private loans from sovereigns. When we substitute credits by private loans, a positive effect is visible in the baseline and negative scenario, but a negative effect for the positive scenario. When we substitute mortgages by private loans, the expected return decreases, however. Funding private loans from all assets also leads to a slight decrease of the expected return (but this also leads to a much lower required capital, as mentioned above). When funding from equities, both the required capital and the expected return decrease significantly. We can thus conclude that government-related private loans are particularly attractive as an alternative for sovereign or credit portfolios. Solvency II: conclusions Investing in government-related private loans can lead to a significantly lower Solvency II capital charge, due to the underlying guarantees. In our example, the overall solvency charge only increases (very slightly) if private loans are funded from euro sovereign bonds, which are not subject to a Solvency II credit risk charge. When private loans are funded from other asset classes, the solvency charge will typically decrease substantially. From a return point of view, government-related private loans are particularly attractive as an alternative for sovereign or credit portfolios. Table 9 sets out a stylized overview of these results. 13

14 Table 9: Stylized overview of Solvency II results 49 We now turn our attention to the capital requirements for government-related private loans for banks under Basel III. Capital requirements for banks: Basel III Basel III: general capital requirements 50 Basel III is the third of the Basel Accords. Basel III is currently being phased in as the leading global regulatory framework for banks. In response to the last financial crisis, Basel III has stepped up the capital requirements and introduced additional limits on leverage and liquidity ratios. 51 Full implementation of the Basel III accords has been delayed several times and is currently expected in In the meantime, work is already underway on the successor to Basel III, Basel IV. We focus here on the current EU capital requirements under Basel III. 53 Note also that the capital requirements for banks are sometimes adopted in other regulatory frameworks as well. For example, the Basel II credit risk approach has also been incorporated into the Swiss Solvency Test (SST) for Swiss insurance companies. 54 The capital adequacy ratio (CAR) of a bank is equal to the ratio of the total capital and the total risk-weighted assets (RWA). The CAR must be higher than 10.5% according to the Pillar 1 requirements when Basel III is fully implemented in Total capital is defined as Tier 1 plus Tier 2 capital. Tier 1 capital is the core capital of a bank, which includes equity capital and disclosed reserves. Tier 2 capital consists of supplementary capital, like subordinated debt, and is used to absorb losses in a case of a bankruptcy. A bank can also be required to hold more capital as a result of the Supervisory Review and Evaluation Process (SREP), as conducted by supervisor (the ECB). The SREP is tailored to the individual bank. In the SREP decision, the supervisor may ask the bank to hold additional capital and/or set qualitative requirements (usually referred to as Pillar 2 ). The latter could refer to the bank s governance structure or its management. 56 A bank can, of course, also hold an additional buffer to ensure that the regulatory limit will not be breached. In the remainder of this section, we neglect the impact of market and operational risk and focus on the impact of credit risk on total RWA. The definition of the risk weights for credit risk for various assets is given in CRD IV / CRR (2013). 57 A short overview for the main asset classes is provided in the table below. 14

15 Table 10: Overview of risk weights for credit risk for various assets under Basel III 58 Note that the above risk weights apply for exposures to rated counterparties. The risk weights for credit risk for government-related private loans are discussed in more detail in the next section. Basel III: capital requirements for government-related private loans We first consider direct exposures to government-related entities. The following government-related counterparties are treated favorably under Basel III regulations: Exposures to EU member states' central governments and central banks These exposures are assigned a risk weight of 0% if they are denominated and funded in their domestic currency. 59 Exposures to multilateral development banks Exposures to the following counterparties are assigned a risk weight of 0%: the World Bank Group - comprised of the International Bank for Reconstruction and Development (IBRD) and the International Finance Corporation - the Inter-American Development Bank, the Asian Development Bank, the African Development Bank, the Council of Europe Development Bank, the Nordic Investment Bank, the Caribbean Development Bank, the European Bank for Reconstruction and Development (EBRD), the European Investment Bank (EIB), the European Investment Fund, the Multilateral Investment Guarantee Agency, the International Finance Facility for Immunisation and the Islamic Development Bank. 60 Exposures to international organizations Exposures to the following international organizations are also assigned a risk weight of 0%: the European Union, the International Monetary Fund, the Bank for International Settlements, the European Financial Stability Facility and the European Stability Mechanism. 61 Exposures to regional governments and local authorities For exposures to regional governments and local authorities, the standard risk weight is 20%. 62 But, if there is no difference in exposure between local/regional and central government exposure, the local/regional exposure can benefit from the 0% risk weight of the member state. The European Banking Association (EBA) has produced a list disclosing this possibility for all member states. A short summary of this list is given in the table below. 15

16 Table 11: Overview of regional governments and local authorities (per country) which qualify for a 0% risk weight under Basel III 63 Note that not all EU member states are on this list. For example, regional governments or local authorities in France do not qualify for this list. We now consider indirect exposures (i.e., exposures guaranteed by a third party). Under Basel III, risk mitigation techniques fall into different categories, and each has its special requirements. The main distinction is between funded and unfunded credit protection. Funded risk protection describes the case of transfer of property on certain assets or collateral after a credit event. The unfunded risk protection comes from the existence of a third party s obligation to pay an amount in case of a specified credit event. We here consider exposures of the unfunded type. To effectively reduce risk, a guaranty must be eligible and meet some requirements before being taken into account. Eligible protection providers include, among others, central and regional governments, local authorities, multilateral development banks, international organizations (with a risk weight of 0%) and public sector entities. 64 The guaranty must however meet a number of conditions to be eligible. The main conditions are: 65 the guarantee provides a direct protection; the extent of credit protection is clearly defined and incontrovertible; the guarantee does not contains any clause to reduce the extent of the guaranty (cancel protection, increase cost of protection in case of a deterioration of credit, reduction of guaranty maturity); the guarantee is legally effective and enforceable in all relevant jurisdictions. To reduce the risk weight, the guarantor needs to have a reduced risk weight, for instance 0%. For the guaranteed part of the loan, the risk weight of the guarantor applies, for the remaining part the risk weight of the obligor applies. Note that partial guarantees are thus also accommodated under Basel III. This is another approach than under Solvency II, where only full guarantees are acknowledged. Table 12 gives some example of government-guaranteed loans which qualify for a 0% risk weight under Basel III. 16

17 Table 12: Overview of government-guaranteed loans which qualify for a 0% risk weight under Basel III 66 Basel III: impact of government-related private loans on overall required capital As an example, consider a bank with 10bn assets and a total (Tier 1 plus Tier 2) capital of 0.6bn. The asset mix is as follows: 10% in cash: risk weight is 0%; required capital (credit risk) is 0% 30% in mortgages: risk weight is 20%; required capital (credit risk) is 2.1% 67 20% in short-term lending to investment grade banks: risk weight is 20%; required capital (credit risk) is 2.1% 10% in core euro sovereign bonds (rating AAA/AA): risk weight is 0%; required capital (credit risk) is 0% 30% in commercial loans to midsized companies (rating BBB/BB): risk weight is 100%; required capital (credit risk) is 10.5% We assume for simplicity that market risk and operational risk do not lead to additional capital charges. We can then easily calculate the overall required capital and RWA by adding up the individual required capital and RWA numbers for all assets. The overall required capital then becomes 0.42bn and the overall RWA becomes 4.0bn. With a total capital of 0.6bn, this means that the capital adequacy ratio is equal to 15%. We can now investigate the effect of adding government-related private loans to the balance sheet. All private loans in our example mandate match the characteristics in Table The risk weight of this mandate is thus equal to 0%. As an example, we allocate 10% to government-related private loans and study the effect on the capital requirements and expected return. Table 13 shows the results. 17

18 Table 13: Basel III - Impact on required capital, risk-weighted assets, capital adequacy ratio and expected return when allocating assets to government-related private loans 69 This table shows that the required capital decreases significantly (by 5-25%) when funding government-related private loans from risky asset classes. Similar effects are visible in terms of the capital adequacy ratio. Government-related loans can thus be used to free up capital on the balance sheet (if needed). Comparing the expected return for the different portfolios, the added value of government-related private loans is mixed. We mainly see a positive effect when adding government-related private loans at the expense of sovereigns. In practice, banks will, however, probably hold on to their sovereign portfolios to keep enough liquid funds, so a major switch from sovereigns to less liquid private loans is unlikely to be practical. Switching to private loans would also have a negative effect of the net stable funding ratio. In the negative scenario, the government-related loans work well, by giving the protection of government guarantees and providing an additional spread over sovereigns. Basel III: conclusions Government-related private loans do not require capital under Basel III and are thus an attractive asset class if capital needs to be freed up. Looking at expected returns, the picture is more mixed. The limited liquidity of these loans is also a point of attention under Basel III. See the stylized results in Table 14 for an overview. 18

19 Table 14: Stylized overview of Basel III results 70 Overall conclusions The required capital for spread risk is limited for government-related private loans under FTK. The main reason is that these loans are backed up by the government, leading to low expected losses. The overall required capital under FTK therefore decreases if government-related private loans are funded from credits or other (risky) assets. At the same time, the expected return will typically increase when we invest in government-related private loans and use sovereigns or credits as funding, making an investment in this asset class also attractive from a return perspective. Investing in government-related private loans can lead to a Solvency II capital charge close to zero, due to the underlying guarantees. When private loans are funded from more risky classes, the Solvency II charge will therefore also decrease substantially. From a return point of view, government-related private loans are again particularly attractive as an alternative for sovereign or credit portfolios. Government-related private loans typically do not require capital under Basel III and are thus an attractive asset class if capital needs to be freed up. Looking at expected returns, the picture is more mixed. The limited liquidity of these loans is also a point of attention under Basel III. 19

20 References Basel (2013), Basel Committee on Banking Supervision, Basel III Phase-In Arrangements. Available at Basel (2015), Basel Committee on Banking Supervision, Revisions to the Standardised Approach for Credit Risk - Second Consultative Document, December Available at CRD IV / CRR (2013), Capital Requirement Directive (CRD IV) and Capital Requirements Regulation (CRR), Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26 June 2013 on Prudential Requirements for Credit Institutions and Investment Firms and Amending Regulation (EU) No 648/2012, Official Journal of the European Union, L 176, Volume 56, 27 June Available at EIOPA (2011), EIOPA Report on the Fifth Quantitative Impact Study (QIS5) for Solvency II, Report EIOPA-TFQIS5-11/001, 14 March 2011, Graph 35, p. 67. Available at EIOPA (2014), Technical Specification for the Preparatory Phase (Part I), Report EIOPA-14/209, 30 April 2014, p Available at _Technical_Specification_for_the_Preparatory_Phase Part_I_.pdf. EU (2015), Commission Delegated Regulation (EU) 2015/35 of 10 October 2014 supplementing Directive 2009/138/EC of the European Parliament and of the Council on the taking-up and pursuit of the business of Insurance and Reinsurance (Solvency II), Official Journal of the European Union, January Available at EY (2016), CRD V/CRR II - Revisions to the Capital Requirement Directive and the Capital Requirement Regulation, December Available at ved%20for%20external%20distribution% pdf. Van Bragt, D. (2017), Capital Requirements for Mortgage Loan Investments under FTK, Solvency II and Basel III, Regulatory Insight, Aegon Asset Management. Available at 20

21 About the authors This article is written by David van Bragt of Aegon Asset Management and Rémi Lamaud of La Banque Postale Asset Management (LBPAM). Aegon Asset Management and LBPAM have a strategic, long-term partnership to jointly develop and sell investment products. LBPAM is France's fifth largest asset manager. David van Bragt is a senior consultant in the Investment Solutions team at Aegon Asset Management. Rémi Lamaud is Head of Regulation and ALM at LBPAM. The authors advise institutional investors about ALM, LDI, risk management and regulatory developments. Acknowledgements The author would like to thank Gerard Moerman, Eddo van den Bogaard, Hendrik Tuch and Niek Swagers for their useful suggestions when preparing this article. More information Sander van der Wel, Head of Financial Institutions Aegon Asset Management Netherlands T (0) E. sander.vanderwel@aegonassetmanagement.com Frank Drukker, Sr. Business Development Director Aegon Asset Management Netherlands T (0) E. frank.drukker@aegonassetmanagement.com Disclaimer Aegon Investment Management B.V. is registered with the Netherlands Authority for the Financial Markets as a licensed fund management company. On the basis of its fund management license Aegon Investment Management B.V. is also authorized to provide individual portfolio management and advisory services. The content of this document is for information purposes only and should not be considered as a commercial offer, business proposal or recommendation to perform investments in securities, funds or other products. All prices, market indications or financial data are for illustration purposes only. Although this information is composed with great care and although we always strive to ensure accuracy, completeness and correctness of the information, imperfections due to human errors may occur, as a result of which presented data and calculations may differ. Therefore, no rights may be derived from the provided data and calculations. 21

22 1 In Dutch, NWB stands for Nederlandse Waterschapsbank. 2 In Dutch, BNG stands for Bank Nederlandse Gemeenten. 3 See for more information. 4 See for a detailed specification of the standard required capital calculation under FTK. 5 The S1 capital charge is calculated according to a (multiplication) factor for each maturity, e.g. the multiplication factor for a 15-year interest rate decrease is The S2 capital charge is calculated according to a correlation matrix for all (4) types of equity. 7 The S3 capital charge is calculated according to a correlation matrix for all open currency positions. 8 The S7, S8 and S9 capital charges are by default 0 but should be added when these risks are substantial. 9 The S10 capital charge for active equity portfolios can be calculated using the tracking error and total expense ratio. 10 See 11 See 12 See 13 See 14 We use the AEAM Government Related Investment Fund of Aegon Asset Management as our reference portfolio here. 15 This example is based on data for all Dutch pension funds. See for the underlying data. 16 With 85% AAA-rated sovereigns (with a capital charge of 0) and 15% AA-rated sovereigns. 17 Most Dutch pension funds only partially hedge the interest rate risk on their balance sheet. The overall required capital will thus typically be higher in practice. Replacing credits or equities with private loans could also require additional interest rate hedging, due to the higher interest rate sensitivity of our example private loan portfolio. 18 For simplicity, transaction costs, management fees and additional hedging costs are not considered in this article, although they may be important in practice. 19 The required capital for market risk is, in this example, equal to 2 2 (SCR equity + SCR spread + 2 correlation equity,spread SCR equity SCR spread ). SCR equity is here 10bn*0.5*0.3= 1.5bn; SCR spread is 10bn*(0.35* *0.09)= 0.17bn and correlation equity,spread is The required capital is, in this example, equal to (SCR market + SCR life + 2 correlation market,life SCR market SCR life ). SCR market is here 1.592bn; SCR life is 0.25bn and correlation market,life is This table is a highly stylized version of the results in Table 4. Please refer to Table 4 for the quantitative results. 22 See EIOPA (2014) and EU (2015) for a detailed specification of the standard required capital calculation under Solvency II. 23 This section has been published earlier in Van Bragt (2017). 24 See EIOPA (2014), p See EIOPA (2014), p See for example typical (average) numbers for life insurers from the QIS-5 impact study for Solvency II. See EIOPA (2011), graph 35, p EU (2015a), Article 176/2. 28 We do not consider the impact of interest rate risk under Solvency II in this article. In general, private loans offset part of the interest rate risk of the liabilities for a life insurer, leading to a lower capital charge compared to a cash investment. The opposite effect occurs for an insurer with shortdated liabilities (like a health or property & casualty insurer), leading to a higher capital charge compared to a cash investment. 29 Insurers that use internal models can use internal credit ratings to assess the capital requirements of unrated instruments under Solvency II. The regulator should, however, give approval of the internal credit rating processes and the governance surrounding them. 30 EU (2015a), Article 180/2. 31 EU (2013), Article 117/2. 32 EU (2013), Article EU (2015a), Article EU (2015b), Article Dutch water boards (in Dutch: waterschappen or hoogheemraadschappen ) are regional government bodies charged with managing water barriers, waterways, water levels, water quality and sewage treatment in their respective regions. 36 See for more background information (in Dutch). 37 See for more information (in Dutch). 38 For See Van Bragt (2017). 39 For simplicity, we here ignore the symmetric adjustment mechanism. This mechanism increases the magnitude of the equity shock in case of an upward equity market (and vice versa). 40 The average allocation to equities is 5-10% for a typical European insurer, so this is an example of a more aggressive asset allocation. 41 EU (2015), Article 169(1). 22