kpmg's swiss pensions accounting survey 2017 · accounting survey 2017 kpmg.ch. ... 11%...

Swiss Pensions Accounting Survey 2017

kpmg.ch

KPMG‘s Swiss Pensions AccountingKPMG’s Swiss pensions accounting survey 2017 looks at trends in accounting assumptions based on the experience of 77 Swiss companies reporting under IFRS or US GAAP as at 31 December 2016. The survey covers companies advised by a range of actuarial consultancies and includes both Swiss listed companies and Swiss subsidiaries of overseas parents.

2 | KPMG‘s Swiss Pensions Accounting Survey 2017

Key trends for FY16 reporting

There were three main themes in international pensions accounting during FY16:

1. Falling discount rates, with rates becoming negative for a period near the middle of the year.

2. The introduction of a more sophisticated way of projecting expected improvements in mortality.

3. The introduction of “risk sharing” approaches within liability calculations under IFRS (but not US GAAP).

Discount rate

Market uncertainty driven by Brexit and the US presidential elections resulted in falls in corporate bond yields (and thus accounting discount rates) during the first three quarters of the year. This occurred both globally and specifically in Switzerland.

The most extreme position was reached shortly after the conclusion of the Brexit referendum when Swiss accounting discount rates briefly became negative.

Subsequently rates recovered such that, on average, rates fell by around 20-30bps over the year (increasing measured liabilities by around 5%). Over the same period typical pension fund asset returns were around 0.63% (Source: Credit Suisse Pension Fund Index).

Thus, most companies will have seen an increase in deficits recorded on their balance sheet as at 31 December 2016.

Our survey showed a wide range of rates being used (from 0.40% to 0.85%) as well as a large range of movements in rates (from an increase of 5bps to a reduction of 80bps).

While these differences will partially reflect duration differences (though the data in our survey does not suggest a strong correlation between duration and discount rate), they suggest a range of methodologies being applied by different firms and a number of changes to methodologies during the year.

11%

Source: 64 companies

Increase / (decrease) in discount rates from 31 December 2015 to 31 December 2016

0.60 - 0.69%

47%

34%

6%

0.70 - 0.79% 0.80 - 0.89% 0.90 - 0.99% 1.00 - 1.09%

Source: 77 companies; 2015 and 2016 indicators are median

Discount rates

0.40 - 0.49%

25%

45%

22%

5%

0.50 - 0.59% 0.60 - 0.69% 0.70 - 0.79% 0.80 - 0.89%

▼2016

2015: 0.90% – 0.99%

2%

3%

KPMG‘s Swiss Pensions Accounting Survey 2017 | 3

Interest credit rate

For 2017 the legal minimum interest credit rate (which applies only to the legal minimum part of employees’ savings account balances) was reduced to 1.00% (2016: 1.25%). We have seen a similar reduction in the interest credit rate assumptions used by companies for accounting purposes.

Around one third of firms used the same interest credit rate and discount rate at 31 December 2016. The difficulties in using an approach where these two rates are always equal became particularly apparent when discount rates became negative as it is not permitted to credit negative interest.

16%


0.30 - 0.60%

31% 30%

15%

0.61 - 0.90% 0.91 - 1.20% 1.21 - 1.50% 1.51 - 1.80% 1.81% +

7%

▼2016

▼2015

2%

Pension increases

Pension increases are not mandatory in Switzerland and are provided at the discretion of a plan’s foundation board.

The majority of companies surveyed assume that no pension increases will be granted in future. Where an allowance is made, this tends to reflect a foundation board which has a stated policy regarding increases.

91%

0.00 - 0.19% 0.20 - 0.39% 0.60 - 0.79% 0.80 +

2%5%

2%

0.40 - 0.59%

▼2015 & 2016

2%


Interest credit rates

Pension increases

Mortality

Mortality assumptions comprise of two elements:

– a “base table” which determines the probability of an individual dying at the time the underlying mortality data was observed; and

– an allowance for improvements beyond the effective date of the base table (either up to a specified future date or on an open ended basis).

The BVG/LPP 2015 base tables were used by all of the surveyed companies. The majority of companies surveyed (over 90%) used the standard “generational” projections which are provided with the BVG/LPP 2015 tables.

However, using work carried out by AON Hewitt, certain companies used an alternative future improvement allowance based on a model published by the UK Continuous Mortality Investigation Bureau (“the CMI model”) calibrated with parameters based on Swiss population mortality experience.

This was done to overcome certain perceived shortcomings in how the standard generational projections allow for the continuation of the “cohort effect” (observation that groups of individuals born in certain years have experienced significantly different improvements in longevity compared to individuals born in other years).

The CMI model uses a considerably larger number of parameters in its future projections and therefore is better able to allow for the continuation of the cohort effect into the future than the standard generational model.

Our observation is that the CMI model results in liabilities somewhere between those resulting from the use of the BVG/LPP 2015 base table (with no allowance for future improvements) and the standard generational projections.In some isolated cases we observed companies using the standard generational projections up to a specific point in the future, with no allowance made for improvements after this point. However, this approach is not in widespread use.

70 | INSP IRAT ION

Schweizer Personalvorsorge | Prévoyance Professionnelle Suisse | 01·17

Deutlich sichtbar sind die diagonalen Musterverläufe, die Veränderungen der Sterblichkeitsrate bei in bestimmten Jah-ren geborenen Kohorten von Individuen anzeigen. Dieses weithin anerkannte Merkmal wird in der Regel als Kohorten-effekt bezeichnet. Die Grafiken 1 und 2 veranschaulichen, dass der Kohortenef-fekt in der historischen Verbesserung der Sterblichkeitsrate in der Schweiz und Grossbritannien auftritt.

Ein alternatives Modell: Continuous Mortality Investigation (CMI)

Das Continuous-Mortality-Investi-gation-Modell (CMI1) wurde in Gross-britannien vom Institute and Faculty of Actuaries entwickelt, um den beobach-teten Kohorteneffekt zu berücksichti-gen. Es findet bei britischen Pensions-kassen und Versicherungen häufig Ver-

1 Details: https://www.actuaries.org.uk/learn-and-develop/continuous-mortality-investiga-tion

wendung bei der Prognose zur Veränderung der Sterblichkeitsraten. Auch Langlebigkeits-Rückversicherer in Grossbritannien und anderen Ländern greifen das Modell zunehmend häufiger auf.

Menthonnex und CMI im VergleichIm Gegensatz zum Menthonnex-

Prognosemodell können Sterblichkeits-veränderungen beim CMI-Modell auf dem Geburtsjahr («Kohorte») sowie dem Alter und dem Kalenderjahr («Periode») basieren. Wir haben das Menthonnex-Modell mit dem CMI-Modell vergli-chen, kalibriert mit Bevölkerungsdaten aus der Human Mortality Database (HMD).

In den Grafiken 3 und 4 zeigt die Darstellung auf der linken Seite der schwarzen Linie die auf den HMD-Da-ten basierenden historischen Sterblich-keitsveränderungen nach Alter und Jahr. Auf der rechten Seite der Linie ist hin-gegen die vom jeweiligen Modell er-

zeugte Prognose der Sterblichkeitsver-änderung nach Alter und Jahr zu sehen. Es ist deutlich zu erkennen, dass das Menthonnex-Modell den Kohortenef-fekt nicht abbildet.

Zur Veranschaulichung der Unter-schiede zwischen den beiden Modellen gibt die Tabelle «Unterschiede zwischen den Modellen» einen Überblick über die Lebenserwartung bei bestimmten Alters-gruppen. Beim CMI-Modell2 kommt

2 Die in diesem Artikel verwendeten Grafiken und Analysen basieren auf dem CMI-Modell «Proposed», wie es im Juni 2016 als Entwurf veröffentlicht wurde, finale Veröffentlichung durch CMI geplant für März 2017. Auch die mit dem noch gültigen CMI-Modell durchge-führten Analysen bestätigen, dass sich die Schlussfolgerungen durch die Verwendung des «Proposed»-Modells nicht ändern. Nach der Abgabe des Artikels ist Mitte Dezember 2016 das neue CMI-Modell «Revised Proposed» diskutiert worden (noch unveröffentlicht). Wir konnten dieses neue Modell noch nicht in die Arbeit einbeziehen.

(bis 2014), gefolgt von BVG 2015 (ab 2015)

(bis 2014), gefolgt von CMI mit LTR 1.25% (ab 2015)

+4.75%+4.50% ±0.25%+4.00% ±0.25%+3.50% ±0.25%+3.00% ±0.25%+2.50% ±0.25%+2.00% ±0.25%+1.50% ±0.25%+1.00% ±0.25%+0.50% ±0.25% Nil ±0.25% –0.50% ±0.25%–1.00% ±0.25% –1.50% ±0.25%

–1.75%

Alte

r

Jahr

Schweiz100

90

80

70

60

50

401940 1950 1960 1970 1980 1990 2000 2010

Alte

r

Jahr

GB100

90

80

70

60

50

401940 1950 1960 1970 1980 1990 2000 2010

Alte

r

Jahr

HMD-Daten (Männer) CMI-Prognose90

80

70

60

50

401960 1970 1980 1990 2000 2010 2020 2030 2040 2050

Alte

r

Jahr

HMD-Daten (Männer) Menthonnex-Prognose90

80

70

60

50

401960 1970 1980 1990 2000 2010 2020 2030 2040 2050

Quelle: Aon Hewitt

New demographic tables

The most recent version of the Swiss standard private sector actuarial tables are the BVG/LPP 2015 tables, issued in December 2015 (previously being the BVG/LPP 2010 tables), though very few companies used them for FY2015 reporting. These new tables reflect:

Human mortality database Generational projection Human mortality database CMI projection

Mortality changes, men (until 2014), followed by BVG 2015 Mortality changes, men (until 2014), followed by CMI 2015

Year Year

Age

Age

Sour

ce:

AO

N H

ewitt

≥ +4.75% +4.00% ±0.25% +3.00% ±0.25% +2.00% ±0.25% +1.00% ±0.25% NIL ±0.25%

KPMG‘s Swiss Pensions Accounting Survey 2017 | 5

– relatively little change in turnover rates; – increased life expectancy; and – reduced disability rates.

The impact of the new tables on measured liabilities was generally low, reflecting increases due to longer life expectancies having been offset by falls in disability probabilities.

+4.50% ±0.25% +3.50% ±0.25% +2.50% ±0.25% +1.50% ±0.25% +0.50% ±0.25% -1.00% ±0.25%

Employee turnover

The majority (85%) of companies in our survey used the standard BVG/LPP 2015 employee turnover scale. While only the larger foundations have sufficient data to produce their own employee turnover scale, smaller companies may apply an adjustment factor to the BVG/LPP 2015 tables such that they better reflect actual turnover experience (e.g. using 150% of the BVG/LPP 2015 rates at all ages). Modified approaches such as this are generally seen as positive as they meet the requirement for assumptions to be “best estimate”.

Making an allowance for higher (lower) employee turnover generally reduces (increases) measured liabilities as, when an employee leaves, their accumulated account balance transfers to another arrangement and the requirement to provide interest credits and conversion guarantees is removed.

Disability

The disability rates in the BVG/LPP 2015 tables are around 50% lower than those in the BVG/LPP 2010 tables on average. Disability benefits tend to be relatively generous in most Swiss plans such that this change resulted in a reduction in measured liabilities.

Lump sum proportion

The median assumption in our survey was that employees would take 20-30% of their account balance as a lump sum on reaching retirement age, although some companies used a significantly higher assumption (as high as 60%), while others made no allowance.

Making an allowance for employees to take part of their benefit as a lump sum at retirement, as opposed to taking it as pension, reduces the IFRS/US GAAP measured liability as conversion rates are determined using a local technical interest rate which is in most cases (significantly) higher than a current accounting discount rate.

If the current low interest rate environment persists, we anticipate that individuals will express a greater preference for pension over cash, reflecting the relative generosity of conversion rates and that limited returns can (currently) be generated by investing amounts taken as a lump sum.

Changes to behavior may materialise quickly and companies should monitor their experience closely.

6%

15%


17%

38%

13%9%

▼2015 & 2016

2%

0 - 9% 10 - 19% 20 - 19% 30 - 39% 40 - 49% 50 - 59% 60% +

Proportion of benefits taken as lump sum

Salary increases

0.50 - 0.99%

51%

29%

11%7%

1.00 - 1.49% 1.50 - 1.99% 2.00 - 2.49% 2.50% +

Inflation is not generally a significant assumption for the measurement of Swiss pension liabilities. However, it is often used as a basis for setting the salary increase assumption (though this is also not particularly significant for typical Swiss cash balance plans).

As there are no Swiss inflation linked bonds it is not possible to set the inflation assumption using a “mutually compatible” (i.e. market neutral) approach (IAS19.75). Hence

Inflation and salary increases

this assumption is set as “best estimate” and we see a wide range of assumptions being used from year to year. One public source of future inflation expectations is the Swiss National Bank quarterly bulletin. However, this is only a relatively short term forecast.

Our survey shows that there has been only minimal movement in typical inflation and salary increase assumptions during the last year.

0.00 - 0.30%

32%

48%

7%

0.31 - 0.61% 0.62 - 0.92% 0.93 - 1.23% 1.24% +

5

1

1 1

Swiss National Bank leaves expansionary monetary policy unchanged The Swiss National Bank (SNB) is maintaining its expansionary monetary policy. Interest on sight deposits at the SNB is to remain at –0.75% and the target range for the three-month Libor is unchanged at between –1.25% and –0.25%. At the same time, the SNB will remain active in the foreign exchange market as necessary, while taking the overall currency situation into consideration. The SNB’s expansionary monetary policy is aimed at stabilising price developments and supporting economic activity. The negative interest rate and the SNB’s willingness to intervene in the foreign exchange market are intended to make Swiss franc investments less attractive, thereby easing pressure on the currency. The Swiss franc is still

forecast has been revised slightly downwards in the short

expected. The SNB nevertheless anticipates an unchanged

however, the forecast is down to 0.1%, from 0.2% in the

that the three-month Libor remains at –0.75% over the entire forecast horizon.

The global economy has continued to recover in line with the SNB’s expectations. In the US, third-quarter

to improve. The other major economic areas also recorded favourable economic activity in the third quarter. The euro area and Japan continued to register moderate growth,while in China, growth remained strong thanks to a variety

of the Brexit decision has so far proved less pronounced than originally feared.

The SNB expects the moderate pace of global growth to continue in 2017. The baseline scenario for the world economy is still subject to considerable risks, however. Structural problems in a number of advanced economies could negatively affect the outlook. Added to this are a multitude of political uncertainties which are particularly associated with the future course of economic policy in the US, upcoming elections in several countries in the euro area as well as the complex and arduous exit negotiations

According to an initial quarterly estimate, GDP in Switzerland grew at an annualised rate of 0.2% in the third quarter. This small increase must also be seen in the context of high growth in the second quarter. Year-on-year,

indicators point to a continuation of the moderate economic recovery in Switzerland and are thus consistent with our previous GDP growth forecast of around 1.5%

support this view. Up until November, the seasonally

conditional inflation forecast of december 2016Year-on-year change in Swiss consumer price index in percent

%

–1.5

–1.0

–0.5

0.0

0.5

1.0

1.5

2.0

2013 2014 2015 2016 2017 2018 2019

Inflation Forecast December 2016 with Libor at –0.75% Forecast September 2016 with Libor at –0.75%

Source: SNB

Inflation

3%

Source: 76 companies; 2015 and 2016 indicators are median Source: 44 companies; 2015 and 2016 indicators are median

▼2014 & 2015

▼2015 & 2016

Conditional inflation forecast of December 2016

7% 7%


Risk sharingPrior to 2016, the consensus interpretation of IFRS effectively required that the liability on the balance sheet in respect of a Swiss pension plan reflected the benefits set out in the formal terms of the plan with any deficit allocated entirely to the sponsoring employer.

However, since discount rates started to fall, it was generally understood in the market that an IFRS liability was not able to fully capture how Swiss pension plans operated due to risk sharing mechanisms which existed in the Swiss pensions framework and which allowed (amongst other things) annuity conversion rates to be reduced and up to half of deficit contributions to be paid by employees.


Towards the end of 2016 this consensus approach was challenged and it is now accepted amongst auditors and actuaries that, if the specific facts and circumstances of a plan permit, it is possible to adjust the IFRS liability to reflect the risk sharing mechanisms which are available.

Factors to consider when determining whether it might be appropriate to allow for risk sharing mechanisms in an IFRS calculation are:

– a history of restructuring measures to secure a pension fund’s financial stability;

– communication to/expectations by employees of future reduced benefits, and a company’s willingness

to publicly communicate shared deficit financing responsibilities;

– discussions and communication by the foundation board and/or employer about required restructuring/ derisking measures; and

– the existence of structural deficit - i.e. benefits promised in the plan regulations that are not fully financed by recurring contributions, and expected return on assets.

The wording in IFRS (IAS19) which allows risk sharing to be implemented does not exist under US GAAP.At the time of writing, only a handful of companies are applying risk sharing adjustments and it is therefore possible that market practice regarding how it is implemented will develop in coming years.

This survey is based on data gathered from publicly listed companies, KPMG clients and other firms where KPMG has received access to pensions accounting information.

Although relatively few companies surveyed reported under US GAAP (18% of the total), making a reliable separate analysis impossible, we did not observe any substantive difference between the assumptions used by IFRS reporters and US GAAP reporters.

Similarly, we did not observe significant differences in the assumptions used by different types of companies (listed, non-listed, etc.).

Companies surveyed

Lump sums

KPMG’s pensions accounting survey 2017 looks at trends in accounting assumptions based on the experience of 77 Swiss companies reporting under IFRS or US GAAP as at 31 December 2016. The survey covers companies advised by a range of actuarial consultancies and provides a detailed insight into market-wide practice.

We are a network of KPMG pensions consultants who work together to service the needs of our national and multinational clients.

Our global pension practice has over 400 professional staff based in territories where defined benefit pensions and other long term employee benefits are common, with our team in Zürich being a centre of excellence for international pensions.

As part of a wide professional services firm, KPMG Pensions frequently works alongside other parts of the business (Audit, Tax and Advisory) on multi-disciplinary projects. On pensions-led projects we are able to leverage this wider familiarity with the commercial environment to bring a more holistic business perspective to pensions issues.

This understanding of corporate processes is particularly key in times of change e.g. around legislative and regulatory developments, in commercial transaction situations; in internal reorganisations or restructurings, hence this is a key pillar of our service lines.

About our survey KPMG’s Pensions practice

Overview of companies included in survey

IFRS US GAAP Total

SIX listed 30 5 35

Swiss subsidiaries 15 9 24

Others 18 0 18

Total 63 14 77

Source: KPMG Analysis


Contacts

Graham MiddletonPensionsPartner+41 58 249 34 [email protected]

Daniel TonksPensionsSenior Manager+41 58 249 33 [email protected]

The information contained herein is of a general nature and is not intended to address the circumstances of any particular individual or entity. Although we endeavor to provide accurate and timely information, there can be no guarantee that such information is accurate as of the date it is received, or that it will continue to be accurate in the future. No one should act on such information without appropriate professional advice after a thorough examination of the particular situation. The scope of any potential collaboration with audit clients is defined by regulatory requirements governing auditor independence.

© 2017 KPMG AG is a subsidiary of KPMG Holding AG, which is a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (“KPMG International”), a Swiss legal entity. All rights reserved.

kpmg.com/socialmedia kpmg.com/app

For further information, please contact:

kpmg's swiss pensions accounting survey 2017 · accounting survey 2017 kpmg.ch. ... 11%...

Documents