BT GROUP PLC. Pension teach-in. 27 November 2013

Transcription

1 BT GROUP PLC Pension teach-in 27 November 2013 Company speakers: Alex Pike Investor Relations Paul Rogers Head of Pensions Risk Director Group Financial Control Mike Smedley Pensions Partner, KPMG Alex Pike Slide 1: Thank you very much for coming along today all of you and also for those of you on the phone today. For those of you that haven't been to one of these before this is a series of regular teach-ins that we do. They are designed to be educational as much as anything else and not to give any new financial information. Today of course we are looking at the pension. So we have 3 people that are going to be talking to you. is the Director of Group Financial Control and he leads on interacting with the BT Pension Scheme Trustee on behalf of the BT Pensions Committee. To the right over there is Paul Rogers who is an actuary working for BT. He supports Glyn on his interactions with the Trustee. Finally on the left there we have Mike Smedley who is a Pensions Partner at KPMG. He isn t involved directly with anything to do with the BT Pension Scheme so he is here to give an external perspective on the key themes in the pension industry. Slide 2: If we just have a quick look at the agenda. Paul is going to start by looking at the overview of the UK schemes and the various ways in which we value them. Glyn will then give you an overview of pensions accounting under IAS 19. Finally we will take a deeper look at the actuarial valuation first with Mike giving a view of the external landscape and then with Glyn giving BT s position. That should all take about an hour. If you could hold any questions to the end that would be great. We will have a Q&A session and Mike, Glyn and Paul will be more than happy to take questions then. We should wrap up shortly before 3pm I would have thought. With that I will hand over to Paul. Paul Rogers Slide 3: Thanks Alex, and good afternoon everyone. Slide 4: I am going to start with an introduction showing our two main UK pension schemes. The first is the BT Pension Scheme, or BTPS, and that is a defined benefit scheme which means member benefit is calculated looking at a predetermined formula. The benefit terms are defined and the contributions required for those benefits are uncertain. The eventual cost will depend on a number of factors such as investment returns, how long members will live, and therefore the investment and other risks are borne by BT and we are focusing on the majority of the session on the BTPS. The second scheme is the BT Retirement Saving Scheme, or the BTRSS, and that is a defined contribution scheme meaning the contributions payable are pre-determined and the eventual pension a member receives is uncertain depending on factors like investment returns and the rate 1

2 at which they can buy an annuity at retirement. For members, they bear the investment and other risks which leads to limited financial risk with BT. New entrants to BT join the BTRSS. The scheme is administered and run by Standard Life with each member having an individual contract with them. Slide 5: Now before going into the BTPS in some more detail I will just provide a short overview of the BTRSS. Now there are currently roughly 25,000 employee members contributing into the scheme and BT s contributions depend on the level of contributions made by the employee and are set out in the table shown on the slide. So for example if a member paid 5% of their pensionable pay then BT will make a corresponding payment of 8%. Members can then choose how they invest their contributions from a range of investment fund options or there is a default fund for those that prefer not to select their own funds. Contributions get paid into a member s individual account and they roll up with investment returns and are subsequently used by the member at retirement to purchase pension benefits. New entrants enter the BTRSS and this is a vehicle used to meet the government s new auto enrolment requirements. We have complied with those requirements and rules since they applied for us on 1 November 2012 when we were one of the first employers in the UK to be compliant. Slide 6: So I ll now move to focusing on the BTPS on the next slide. There are three main sections to the BTPS titled (a) (b) and (c). The section that a member is in will depend on when they joined the scheme and that is shown on the slide. Membership numbers are shown on the table at the bottom. One point to note here on the table is we group section (a) and (b) together on the slide and also in our annual disclosures and that s because section (a) members have the option to elect section (b) benefits before retirement and it's usually more beneficial for them to do so, so most members do convert to the section (b) terms. There are also less than 1,000 active section (a) members; the majority have now retired. In terms of numbers as of 31 March 2013 roughly 60% of the membership was pensioners. 25% were deferred members and by deferred member I am referring to former BT employees that are no longer earning benefits but have not yet retired. Then we have 15% that are active employee members. Liabilities are broadly split: 80% for sections (a) and (b) and 20% for section (c) using the IAS 19 basis. Slide 7: This slide is giving an outline of the benefits that we provide in the BTPS and I will spare you the finer details of exactly how all the scheme benefits are calculated and for those interested there are member booklets on the Trustee s websites that give you more details on the sections and the benefit formulas if you need them. So what we have concentrated on doing on this slide is show how the benefits increase from a financial perspective. So taking the first column of the table showing active members. We amended benefits in the BTPS in April When the changes were made the formula for benefits earned before that date was unchanged with the new formula used to calculate benefits for service after that date. For service before April 2009 the benefit formula is based on a member s salary at or close to retirement, referred to as the final salary benefit, and for service after April 2009 the benefit formula uses a career average approach. So here benefits at retirement are based on a member s average salary rather than the salary at retirement. In the BTPS, salaries from earlier years are indexed up each year by the lower of RPI and the member s actual salary growth. Now to meet the cost of these active member benefits being earned BT and the employee pay a combined rate of 13.5% of pensionable pay and that will be next reviewed at the June 2014 valuation. 2

3 Moving onto the second column showing deferred members. Once a member leaves the scheme, their benefits are fixed at that point and those benefits are then increased up to a member s retirement broadly in line with CPI inflation. The final column shows the pension increase for pensioners. Once members retire and benefits are in payment the benefits are increased annually on 1 April. For section (a) and (b) members the annual increase is based on CPI and for section (c) the increase is based on RPI with a cap of 5%. Some members have some smaller statutory and other elements of their pension that don't increase but the majority of benefits are inflation linked. Slide 8: In terms of governance for the scheme. In BT we have a Board sub Committee chaired by Patricia Hewitt which is responsible for overseeing the relationship with the Trustee of the BTPS. The Committee also considers pension policy and strategy for the group. It covers investment strategy and performance in the BTPS as well as reviewing and making a recommendation to our board on the actuarial valuation. It also reviews and approves our risk management activities relating to pensions. For example, this year we launched a pension increase exercise which offered pensioners the option to give up inflationary increases on part of their pension and receive an initially higher amount. This exercise was about giving members additional flexibility and choice but also providing increased certainty to the scheme and BT on future pension payment amounts. Slide 9: Governance on the Trustee side. The scheme itself is run and managed by an independent Trustee body. There are 9 Trustee directors appointed by BT. 4 of which are nominated by the trade unions and they are required by legislation to operate the scheme in accordance with the trust deed and rules and to act in the best interests of the scheme s beneficiaries. The day-to-day operations of the pension scheme are managed by the Trustee s executive arm and the BT Pension Scheme management. In common with other UK pension schemes, investment policies are set by the Trustee, the Trustee is required to consult with the company before making any changes on the investment strategy and we have a good two way dialogue with the Trustees in that area and BT regularly attends meetings with the Trustee s investment committee. So when investing in these assets the Trustees use a range of fund managers across different asset classes. Including Hermes which is a fund manager owned by the Trustee. The annual report published by the Trustee is on their website and again contains further information on governance, the assets, performance and experience over the year. The scheme s year end has been amended to 30 June and that will bring it in line with the valuation date of the scheme and that next set of accounts is expected to be issued later this year. Slide 10: I am now going to move onto discussing the assets and liabilities of the scheme starting with assets. As of 30 September 2013 the scheme assets totalled 39.3 million and the charts on the slide are showing the move in the target asset allocation over the last 10 years. There are a few points that I want to draw out from the charts: the reduction that can be seen in the level of equity allocation that has reduced the equity component in the scheme from 63% down to 29%. Correspondingly there has been an increase in diversification across the portfolio so there is now a more balanced portfolio that isn t overly reliant on returns from any one particular asset class. There is also an increased allocation to inflation linked assets that you can see up from 9% to 31% which is reflecting the nature of the scheme s inflation linked liabilities. The discount rate for the pension scheme is driven by that asset strategy and the actuarial funding assumptions to allow for that asset mix to gradually move towards lower risk assets over time. I will just give you a little bit more detail on the asset classes. The equities is a diversified mix across equities from around the world and around 75% of these are held in overseas equities. The fixed interest portion holds both government and corporate debt again in the UK and overseas. The inflation linked section is similar again and it is also invested in government and corporate 3

4 debt and it also has an allocation within that bucket to infrastructure assets. The property section is again diversified across countries but is typically mainly invested in the UK and the other assets categories include some other allocations including hedge funds, commodities and high yielding and emerging market debt. Slide 11: So looking at returns, the table on the slide shows the Trustee s expected returns on the different asset classes and that leads to an average expected return of 2.5% above RPI as of 31 December So as shown below the actual returns have exceeded that level in recent years with a 7.5% return in the year to 2012, 4.4% above RPI, and over a longer 10 year period returns have been 8.3%, so 5% above RPI. Slide 12: Now moving to liabilities. Before I come onto the different valuation approaches I just wanted to give a quick explanation of how the actuaries will approach valuations and essentially this is looked at in two parts. Firstly there will be a calculation of the value of the past service liabilities and this is looking at how much money is needed to provide the benefits promised based on service up to the valuation date. It uses a discounted cash flow approach in calculating those liabilities. That figure for the liabilities is then compared across to the assets to determine whether there is a surplus or deficit in the scheme. Secondly there is a separate calculation of the liability for future service benefits. This looks at the cost of the new benefits being earned and leads to the regular cash contribution amounts. Or for example determines the service costs under IAS 19. Slide 13: In terms of looking at those past service liabilities in more detail. The discounted cash flow approach will make a projection of the expected pension payments under the scheme for each year into the future. That is shown in the chart which shows the expected benefit outgoing for each year for the BTPS. What you can see is the benefit payments will increase up in the early years as members are retiring from the scheme before tailing off over a period and stretching out to around 80 years from now. In order to calculate those cash flows the actuaries will need to make a number of assumptions such as when members will retire, how long people will live and future inflation which affects how pensions will increase in the future. That will all be played back into the present value using a discount rate and that is one of the key assumptions in determining the liabilities. The chart is shown using the IAS 19 assumptions of our last year end when the present value of the liabilities was 47 billion. Slide 14: So this is a slide we showed at our half year results presentation outlining the three main approaches that we focus on and I will just spend a little bit more time on this now. The first column shows the IAS 19 valuation which is a prescribed measure that we are required to disclose every quarter. The second column is our median estimate which is taking out any prudent margins and that is our best estimate of the underlying financial positions. The third column is the actuarial valuation which we carry out every 3 years and used to set cash contributions. Moving on to IAS 19 and firstly the discount rate. The Accounting Standards require that the discount rate is based on high quality corporate bond yields for the appropriate duration for the liabilities. So for this we use a AA yield curve produced by our actuaries and the cash flow in each year as set out on the previous slide is then discounted at the relevant rate from the yield curve. So that allows accurately for the duration of the liabilities. The discount rate is then expressed as a weighted average of those rates. So projecting liabilities and changes in liabilities from one date to another using the movement in an index yield such as the iboxx might not always given an accurate result for example if you have a change in the shape of the yield curve over the period you can have a greater or lesser move than there has been in the index. For RPI inflation we look at different market indicators including the Bank of England data on implied inflation when setting out that function. For CPI inflation if there is no relevant market indicator for that we assess the long term expected difference between RPI and CPI and we look at differences in how those two indexes are constructed for coming up with the difference. Demographic assumptions such as life expectancy are typically reviewed following each actuarial valuation. 4

5 Moving onto the middle column the median. This is stripping out, as I said, the margins of prudence and determining what we believe are the most realistic assumptions for the future. Therefore the discount rates allowing for our expectations of returns for the current investment strategy but also how we expect it to develop over time. And, as I said, it s assumed both in the funding and median valuation that the scheme will continue to move to lower risk assets over time. On funding, the assumptions have to be agreed between BT and the Trustee. As required by legislation they have to be prudent overall and in particular the discount rate is set based on a prudent view of expected future returns. All those methodologies use the market value of assets. Slide 15: The next slide is again based on a slide we showed you in our half year results and the grey line is showing the movement in the IAS 19 deficit over recent periods which has been volatile and we have also added the green line on the chart which is showing the real discount rates and that is the discount rate above RPI inflation in this case on the right hand scale. What you can see is that the deficit moves are influenced by the real discount rates and clearly the other factor driving the deficit figure is the asset value. The other point that I will draw out from the chart is the fairly substantial fall in discount rates over the period of around 2.5% - starting at around 3.5% at the start of the period and falling to around 1% more recently. Slide 16: This chart is just bringing together all the three measures that we look at. The pink circles are median estimates. The green circles are showing the last two funding valuation deficits and the difference between those funding and median measures represent our view on the prudence incorporated at that time. A couple of points just to pull out from the chart. We commented at the half year the median remains in surplus and that stays relatively consistent in recent periods. Another point to note is that the funding deficit IAS 19 don't move in the same way and nor would you expect them to given the different approaches, particularly to discount rates. For example the 2008 funding valuation which is the first outlined box the IAS 19 deficit was just over 2 billion and the funding deficit was 9 billion. At the 2011 valuation in the second outlined box the IAS 19 deficit was at a similar level of the 2008 valuation level and the funding deficit had moved to 4 billion. So that is it from me and I will now hand you over to Glyn who will take you through the IAS 19 pension accounting in a bit more detail. Slide 17: Thank you Paul. So IAS 19 pension accounting. What you have all been waiting for! I guess it can be quite complex in terms of accounting because we have different schemes. We have DB [defined benefit] and DC [defined contribution] schemes and also the impact can be seen in various elements of our accounts and whether that be income statement, cash flow, balance sheet, statement of comprehensive income. It touches all those areas. It can be difficult to follow. I will try and demystify some of that for you. Slide 18: Starting with the overview. So this shows the last financial year the various elements from the impact on our accounts. So if you take the first column for defined contribution schemes, the income statement charge was 136 million and that was also reflected in a cash outflow of 136 million. As Paul said earlier with all the investment risk lying with the member there no further balance sheet exposure to the company and hence zero in the balance sheet row. On defined benefit schemes the operating charge or the service cost (i.e. the cost of that period s accrual of benefit under IAS19 assumptions) was 263 million and the net interest or the unwind of the discount on the deficit was 117 million. So that is what flows through the income statement. In terms of the balance sheet position at the end of March, net of tax, the deficit was 4.5 billion and from a cash perspective there were two elements of cash flows. There were the regular contributions - so Paul referred earlier to the 13.5% of pensionable pay that is set out in the last funding valuation - that amounted to 217 million and then there were the deficit contributions set out as part of the recovery plan from the 2011 valuation and that amounted to 5

6 325 million. So that in a nutshell is a very simple overview. What we are going to do on the following slide is take each of those areas in a bit more detail for the defined benefit schemes. Slide 19: So firstly on slide 19. In overview in terms of looking at how one reconciles from the deficit at the start of the year on the left hand side to the deficit at the end of the year there is a number of factors to take into account and that flow through there. Firstly on the left hand side you have got the service costs, annual accrual of benefits earned in the year, the net pension interest - the unwind of the discount. Those increase the deficit. Then the next two bars you have got the cash contributions going in the regular contributions and deficit contributions. Both of which are reducing the deficit and then on the right hand side you have got the actuarial movement in the assets and liabilities which can either increase or decrease the deficit depending upon the conditions. There are two constituent elements of those actuarial movements. Firstly there are differences between the actual experience in the period versus the assumption at the start of the year. So for example if investment returns are different to that assumed within the discount rate then that will give rise to an actuarial gain or loss on the assets and then the second element is actually as a result of changes in assumptions. So when a discount rate changes or there is a change in the inflation assumption that will give rise to an actuarial gain or loss. So with that as background, let's look at the BT position through last year in more detail. Slide 20: The chart shows from the left hand side the IAS 19 deficit at the start of the year which was 1.9 billion net of tax 2.4 billion gross of tax. Move across to the right where the closing deficit at 3 1 March 2013 was 4.5 billion net of tax and 5.9 billion gross of tax. What that chart shows is that the big driver of the change is all in the actuarial movement. It's those actuarial gains and losses that gives the volatility that you see in our quarterly IAS 19 numbers. It's not due to what is going through the income statement or the cash it's the actuarial movements that tends to drive that volatility. So on the next few slides I am going to break out each of those components in a bit more detail. Slide 21: So on slide 21 in terms of the income statement. As I said the service cost or operating charge, the cost of an additional years benefit, that s been earned by the active members of the scheme. The active members are those that are still employed by BT. It also includes within that the administration costs of the scheme and the levy that is paid across by the scheme to the pension protection fund. That was 263 million last year. Now when it comes to our disclosure that included within our staff costs and when you look at our quarterly KPI cost analysis it's within the staff cost line. The next component, the net interest, that is not a cash item it's a notional item and it's the unwind of the discount on the deficit. You can broadly estimate what that is by looking at the opening deficit at the start of the year and applying the discount rate to that deficit. Now there are small adjustments to that for the known cash flows during the year but that gives you a pretty good steer as to what the answer is. As you can see, the 2448m being the opening deficit and 4.95% being the discount rate would give you 121m. Now when it comes to our classification cost in the PNL account we treat that as a specific item and have always done. So that is the income statement. Slide 22: Moving on to the cash really the 13.5% of pensionable pay that is the regular contribution there and that was 217 million. You will see that there is a difference between that and what we charge through the P&L account as an operating charge. I referred earlier to 263 million as being the operating charge. The reason that there is a difference is because those two valuation bases use different assumptions and so that drives a different cost, for example the discount rate will be different. The difference between those two items when you look at our KPIs you will see in the Other category in our normalised free cash flow analysis. Then when it comes to the deficit contributions that is set out in the triennial funding valuation and so that gets revisited and reset every 3 years. We exclude that from our definition of normalised free cash flow. Slide 23: So onto slide 23 where we explode out the actuarial gains and losses. On the asset side of the equation there were actuarial gains in the year of 2.7 billion. Basically that reflects the difference between what was actually generated in terms of investment returns versus what was assumed within the discount rate. So the scheme generated a 12% return in the year to March 6

7 2013 and that compared to the discount rate, which is a double AA corporate yield based discount rate of 4.95% so clearly there is a considerable degree of outperformance there and you see that coming through the actuarial gain. This goes directly to reserves through the statement of other comprehensive income rather than going into our P&L account. The second element, the pink element, is an actuarial loss on the liabilities. That is made up of a number of components that I will explode on the next slide. Slide 24: So on slide 24 you will see the actuarial losses of 6.3 billion. The biggest component of that of 4.65 billion, is due to the change in the discount rates. So the discount rate changed from 4.95% to 4.2% reflecting the change in the AA corporate bond yield curve. So that 75 basis point fall gave rise to that increase in the liabilities. The next component up of 1.45 billion is due to changes in the assumption around future inflation. So during that period expectations of inflation increase by 25 basis points up to 3.3%. That 25 basis points change increased the projected liabilities. Then finally at the top of around 150 million you have the liability experience. This is where member experience has differed from the assumptions for things such as pay increases, mortality and leavers from the scheme. That is due to actual behavioural changes or experience changes. So what this chart really shows is that those actual changes are not generating much of that actuarial loss. The actuarial loss has been generated through changes in assumptions. The most significant in that year being the discount rate change. That is not a reflection of what we actually expect to generate in terms of return from the scheme but merely against the IAS 19 benchmark of the AA corporate bond yield. Slide 25: So moving onto slide 25. In term of some of those key assumptions I thought that it would be useful to summarise some of the sensitivities. So the impact of changes that some of those assumptions can have on the liabilities of the scheme. So starting with a 25 basis point increase to the discount rate. That would lead to a reduction in the liabilities of 1.7 billion. On the inflation side of things, if the assumption for future inflation was to increase by 25 basis points that would give rise to an increase in the liabilities of 1.5 billion. You will note the those two don't exactly contra each other out. That is because, as Paul said earlier on, some of the benefit streams under the scheme are not inflation linked, some of them are fixed. And so that differential reflects the fact that you have some fixed liabilities that are not therefore impacted by that inflation assumption. Then if we look at the assumption around salary increases. To the extent that we assume salary increases increase by 25 basis points above RPI which is the assumption made in the valuation. To the extent that they were 25 basis points higher that would give rise to 0.3 billion increase to the liabilities. Then finally on the slide in terms of mortality, if we were to assume that people lived for a year longer than we have assumed in the valuation, then that would generate an increase in the liabilities of 0.9 billion. The actual detailed assumptions around mortality, they re all set out in the accounts and in the glossary to the presentation, so I won't go through them individually here. But it gives you a degree of understanding of the magnitude of change that those assumptions can have on the overall liabilities of the scheme. In particular with the discount rate and inflation gives you a sense of why on a quarterly basis you can see big swings in that IAS 19 deficit position that gets reflected on the balance sheet. Slide 26: So hopefully that has given you a better understanding of some of the moving dynamics when it comes to the IAS 19 accounting for pensions. Hopefully it hasn't sent you to sleep. But having looked at the accounting, we are now going to move on and look at the actuarial funding valuation which is obviously the important measure that then determines the cash contributions into the scheme. So Mike, a pensions partner from KPMG, he s going to give you the landscape in terms of what s happening in terms of external developments in pension funding, and then I will come back and I ll talk more specifically about the BT position with regards to that. So I shall hand over now to Mike. 7

8 Mike Smedley Thanks Glyn and afternoon everyone. As you have heard a couple of times I don't advise BT. I suspect that I am probably the least knowledgeable person in the room about the BT Pension Scheme but I do know a lot about other pension schemes and the market generally and that is what I am going to give you a sense of here. Slide 27: I guess I wanted to start with the high level principles on page 27 about pensions funding and this is quite simplistic but it gives you the right picture. So there s a sort of circle here, if we start at the top, the starting features clearly, what are the benefits that have been promised to pensioners and future pensioners, and for most pension schemes these are now largely known. Most the promises have been made in the past and you can effectively, the cash flows that are due out to future pensioners and current pensioners are largely known. So that little piece is largely set down. If you move around clockwise on the circle to the assets and the investment returns this is, for most pension schemes, this is the most important element of the valuation. So what's in the bank already in terms of the assets of the pension scheme that the Trustees own, and critically what are the future investment returns going to be on those assets. Bear in mind for a pension scheme they are looking at investment returns up until the point ultimately - actuaries are good at being morbid - until the last member dies. So you are looking at a very long term time horizon. So you re looking at investment terms over 20, years potentially. The point at the bottom there around the covenant and prudence is slightly more subtle and I will just explain what I mean. When you are doing a funding valuation with the Trustees, there is a requirement in the law which we will come onto to be prudent. In a sense you could clearly argue that what is going on in the business and what is going on in the pension scheme are different things, and on one level they clearly are. The way that the pensions market work and the ways the Trustees think and the Pensions Regulator means that there is an interaction. How that typically works is in how much prudence is factored into the actual evaluation. So if you like, how much of a buffer is built in because, for example, the investment returns might not be delivered as hoped, and how big is that margin. The way that directionally it works is that if you have got a stronger company they can probably afford to bear more financial risk and therefore there is less need for prudence and a bigger buffer in the funding valuation. Conversely, a weaker company the Trustees might feel that they need to be more cautious about the financing of the scheme and have more prudence in the assumptions. There is no science to this, this is an area of judgment. But it does meant that there is an interaction there. Finally moving around on the left hand side the contributions from the company is one of the key outputs if you like from the actuarial valuation process. And typically expressed both as an ongoing rate at a % of pay for the staff that are still in the scheme plus some funding to meet the deficit. That is the typical way that the valuation is structured. Slide 28: So that is the basic framework and moving on to the next page, 28. It's worth spending a couple of minutes on the regulatory landscape of pensions funding because it is an area that is regulated, and the three legs which are put up there. So there is what is in the legislation, there s the pension funds trustees and then there s the Pensions Regulator. If we start with the legislation and what is actually in the legislation. There isn t a great deal specified in the legislation on pension funding. Basically it says that pension schemes have to be funded and that the actuarial valuation assumptions must be set prudently. So there must be some prudence but beyond that there s a lot of flexibility and the legislation is very clear that pension schemes should be funded on what s called the schemes specific basis i.e. reflecting their specific circumstances. There isn t a one size fits all for pensions funding. So the legislation is relatively flexible. There is one change in the legislation recently which is highlighted on the slide there. I will read it because it is quite important and this came in only 8

9 recently around the statutory objectives that are set out for the Pensions Regulator by government. And previously the Regulator s objectives have been primarily around protecting members of pension schemes and there was a sense that the Regulator needs to take a more balanced approach between members and employers and therefore there s a new objective which is the Regulator should seek to minimise any adverse impact on sustainable growth of an employer. Which effectively is government saying that if an employer is investing in their business the Regulator shouldn't stop them doing that for the sake of the pension fund. Because actually the government and the Pensions Regulator will all say the best thing for a pension scheme is a strong employer. The two should go hand in hand. The second leg that s in here, and this is probably the most important part, is the trustees, in the middle. The trustees have an important role in any pension scheme. They look after the money, look after the interest of members. They have to apply the legislation of pensions funding, they have to negotiate and usually agree with the sponsor how the actuarial valuation is done, the assumptions that are used and the outcome in terms of contributions. Finally, legally the trustees control the investment strategy. There is a requirement to consult the employer and typically there is a close working relationship, but ultimately the trustees control the assets. So that makes them quite key in any given situation. Then finally on the right hand side there we have got the Pensions Regulator, who in a sense is in the background. They like to describe themselves as an interested bystander in some way. But clearly they re still very influential either by intervening in individual circumstances or by way they say to the pensions industry and trustees generally. In terms of what the Regulator is looking for, I have put 3 points up there which I guess are some of the things that we see that the Regulator has asked at the moment. Clearly they want deficits to be met and all else being equal the Regulator likes more prudence rather than less. But equally they recognise that there is a balance between different stakeholders. One thing that they are particularly concerned about is the trustees going backwards if you like, particularly in a transaction or something like that, where the trustees are disadvantaged to the advantage of another creditor. So if there was a refinancing of the business and the bank took security over everything and leaving the trustee high and dry. They don't like that sort of thing and so they have got quite a strong look out for situations that could be a step change for the trustees. And finally they are moving towards an outlook that says pension funds should have an integrated plan looking at the funding of the scheme, the covenant of the employer that s supporting them and the investment of the assets. And that those should be more joined up than they might have been in the past for some schemes, and that trustees should start to have some contingency planning. So if the funding level improves for some reason what then would the trustees do? If something happened with the investments of the scheme what would the trustees do? So they have got more of a dynamic plan. So there s been some criticism that the trustees can be too disjointed in managing schemes. Slide 29: That s the regulatory framework. I just wanted to move finally onto the practical issues where you have got an actuarial valuation going on right now, what are the big issues and things that are coming out? I have put 4 on the page there. The first one and this is probably the most important one most of the time, is what s the discount rate used to value the liabilities? To discount those cash flows. And particularly when a lot of pension schemes would use gilt yields for example as one of their reference points. Gilt yields are looking very low at the moment, what does that mean for the discount rates and the valuation and clearly if you take a macro view, discount rates and actual valuations have been falling over the last few years which means the deficits have been rising. So there s that happening in terms of direction, but for any individual scheme there s a big debate to have and discussion around what s the right discount rate given the investments that the scheme holds, given the reference points in the market. And also given, and Paul touched on this earlier, given the investment strategy of the pension scheme not today, but what it will be in the future. Because the trustees are looking for long term returns, not just today. 9

10 And of course one of the big discussion points in that negotiations is how much prudence do we need and that s one of the big points of discussion that gets you to an agreement ultimately. The second point up there on mortality. I guess this is becoming less of an issue perhaps than it has been in the last 5 years or so, in that there seems to be less rapid improvement in mortality assumptions. People are still living longer, and I can say this as an actuary, but perhaps the actuaries are starting to get the assumptions right or at least changing their minds less frequently. So although there still is some movement in mortality assumptions it is becoming a less contentious, less significant point than it has done in the past few years. Cleary on mortality though there is still a big long term risk and uncertainty. And if there was, I know that the cure for cancer is one that normally gets quoted and those sorts of things could still have a big impact clearly. The third point up there on contributions and this is really company contributions. We are seeing more flexibility in terms of what trustees are prepared to agree in terms of funding of pension schemes. I think that there is a bit of an attitude that says we recognise that we are in a slightly funny environment and bond yields and gilt yields are very low. Some trustees will think, well the deficit we are not really sure it s as big as it's looking like it is and so there is more flexibility that we re seeing in funding plans. That can be longer term, so let s not fund the deficit over 7 or 10 years but let s fund it over longer, or let's have a different shape of contributions, or let's anticipate some more investment returns in arriving at our contribution plan. Generally we are seeing more flexibility than there s been recently. Then the final point up there that seems to be coming through quite strongly, is that trustees in looking at their valuations are saying it's not just about the valuation today but actually we d like to have a discussion about where it is going longer term. Where will the pension scheme be in 5, 10, 20 years time. Let's have half an eye for that and not just agree a valuation today and then go into a different room and come back in 3 years time. Actually let's try and have a discussion about the longer term too. And the Pension s Regulator is keen for that debate to happen as well. So that was really a picture of some of the things that we are seeing in the market. I will hand back to Glyn who is going to talk about the specifics of the BT scheme. Slide 30: Thanks Mike. So the BT Pension scheme. The last funding valuation was June 2011, showed a deficit of 3.9 billion gross of tax and in simple terms was based on a discount rate of 2.0% real relative to RPI. When thinking about the discount rate in terms of funding valuation, as Mike said, there is a duty on the Trustees to be prudent and to take a prudent view. And that s why in the past we have given you updates on what we think the median estimate is and that is based upon our view stripping out prudent margins in terms of returns. And so in June 2011 on the right hand side, the discount rate was 70 basis points higher than the actuarial funding valuation. That was the principle reason why in June 2011 our median estimate was up there with a surplus of circa 2.5 billion versus the funding deficit. I think that it's also worth being conscious of movement in index linked gilt yields when thinking about the discount rate as well. Because whilst the scheme actuary doesn't use a gilts plus basis to derive the discount rate, it is something that the Regulator looks at, to do something that trustees look at and think about, in terms of a benchmark in terms of assessing prudence. So it is something to consider. At June 2011 the gilt yield was 0.6%, so you can see that the actuarial funding discount rate was between those two. So between the median and between the gilt yield. At the previous valuation in December 2008 whilst the actuarial valuation discount rate was a bit higher at 2.5% the median estimate rate was also higher, as was the gilt yield as well. I guess it's also worth noting that since June 2011, as Mike said with QE etc, gilt yields have been very low at the flat to negative levels recently. 10

11 When it comes to the construction of the discount rates, Paul said earlier that that discount rate takes into account how it s assumed that the investment portfolio will develop over time. As the scheme matures and you have less active members and more pensioners drawing pension, the assumption within that strategy is that there is a de-risking of the investment strategy, such that there is a move into more gilts and bonds over time. That is built into the construct of the discount rate. We quote a single equivalent rate of 2.0% just to try and simplify that and taking into the account the duration of the liabilities. Slide 31: So if I look at what happened between December 2008 when the deficit was 9.0 billion and June 2011 when it had reduced to 3.9 billion, the big driver of the change was the actuarial gains on assets. The actual investment returns on the asset portfolio was substantially greater than had been assumed in the discount rate, and so on the previous slide I said that the December 2008 discount rate was 2.5% so the returns were materially above that and that is what gave rise to that 4 billion gain. But if I go across to the left [of the slide] and just work it through. The first pink bar, the 1.3 billion, that is the unwind of the discount on the liabilities during that two and a half year period. The next two boxes, the benefits earned and the regular contributions, offset each other which should be no surprise in as much as the monthly contributions that are being paid into the scheme for the accrual of benefits is offsetting the benefits earned by the members. So there is a plus 0.8 billion and a minus 0.8 billion. Then the next line is the 1.6 billion deficit contributions. Those were the contributions that were set under the 2008 valuation and were paid during that two and a half year period. As I said there was an actual gains of 4.2 billion. Then on the liability side of the equation there was a net actuarial gain of 0.7 billion. Now within that there was the gain that arose on the change to CPI when the government made their change in terms of the basis of state pension increases. That has an automatic flow through to some of the sections of the pension scheme. That effect is within that actuarial gain there. So that was what happened within that two and a half year period. Slide 32: So moving onto the recovery plan. So having agreed what the funding position is, the other part of the actuarial valuation process is to agree a recovery plan that makes good the deficit, and the plan that was agreed with the Trustee in 2011 was a 10 year recovery plan with a 2 billion lump sum payment in March 2012, two subsequent payments of 325 million in 2013 and 2014 followed by 7 annual payments of 295 million through to When you take those cash payments and whilst they add up in aggregate to 4.7 billion, the present value of those is 3.9 billion. That payment schedule made good the deficit and within that there was no assumed allowance for outperformance in terms of investment returns. So one of the things that Mike referred to earlier was flexibility in recovery plans where you might make an allowance for investment outperformance against the prudent measure. The valuation documentation gets signed off and certified by the Scheme Actuary and that was submitted to the Pensions Regulator and they informed us that their position would be informed by the final Court ruling on the Crown Guarantee case. So that is where the 2011 valuation stands. Slide 33: As part of the valuation and recovery plan we also pre-agreed, for clarity, what would happen at the 2014 valuation in the event that the remaining recovery plan was not sufficient to make good the position at So the table, if I go to the column under the 1.0 billion there, what that is saying is that if the deficit at the time exceeds the 2011 recovery plan by 1 billion, so taking all things into account and using the 2011 discount rate we would expect the 2011 recovery plan to have a remaining value of 1.7 billion at So if the deficit is greater than 1 billion over that then these three payments in 2015, 2016 and 2017 kick in. So 199 million in 2015, 205 million in 2016 and 211 million in If I move across to the right hand side, if it exceeds the recovery plan by 2.9 billion, so if it s 4.6 billion, then these three payments kick in. So 360 million in 2015, 371 million in 2016 and

12 million in So this provides a starting point for the 2014 position and if those 3 payments are not sufficient then that would get incorporated into a revised recovery plan. We also agreed a similar mechanism that would apply at the 2017 valuation as well. Slide 34: In terms of other considerations, we also had a number of other aspects of the agreement with the Trustees to provide them with protection in certain circumstances. The first of which is with regards to shareholder distributions. The agreement here was that if the distribution to shareholders exceeded the billion of deficit contributions in the period from 1 March 2012 to 30 June 2015, so if we were to pay distribution to shareholders in excess of that, then there would be a matching contribution into the scheme to level the playing field there. Now that distribution definition excluded any buyback of shares to meet any employee awards and so excludes our current share buyback programme. There was also a commitment around disposals and acquisitions to the extent that if we generated more than 1 billion from disposals net of acquisitions in any one year, then a third of that net proceeds would get paid into the pension scheme. There was also a negative pledge to Mike s point earlier around Trustees wanting to make sure that they are not disadvantaged in terms of the pecking order. If we were to raise any secured financing greater than 1.5 billion then we would be required to give equivalent security to the pension scheme so that they kept to a level peg there. It's worth pointing out that our practice is to raise unsecured debt. In that regard, these were part of the agreement and as part of the next valuation they would all be subject to renegotiation. Slide 35: So looking forward to 2014 amazing how quickly it s come around actually! In terms of the key actions that need to be taken and the schedule of work that needs to be undertaken. I thought that it was worth giving you some context and background on what needs to go on. So if you take the membership data, for example, that all has to be collated and checked as at 30 June. You can imagine with that sort of level of membership, we have got 320,000 members, there are numerous data points that feed into that valuation calculation. It is not an insignificant piece of work just to undertake that. There is also a requirement for the Trustee and for us to consider any changes from the last valuation in terms of legal, demographic or economic position and assess what impact that may have on the assumptions or approach. Clearly as Mike said the sponsor covenant is an important part of the dynamic for the Trustees to consider. All of those things need to be taken into account when considering the level of prudence that is appropriate to incorporate into the valuation. On the asset side of the equation the assets have to be audited. They have to be marked to market as at 30 June. So when we have got all of those pieces the actuary can plug all that into their models, update their calculations and give us an updated position. Clearly there are some negotiations between us and the Trustees around the assumptions, around what we consider is an appropriate level of prudence. When we have got a position on the funding deficit or surplus then we enter into a negotiation around what is an appropriate recovery plan and schedule contributions to make that good and whether any other protections are appropriate. At that point when there is an agreement reached between us and the Trustees, the Scheme Actuary is required to certify all of the valuation documentation and then that gets submitted to the Pensions Regulator. From a regulatory perspective that is required to be submitted within 15 months of the valuation date, so by 30 September Clearly we will be seeking to undertake it quicker than that and in the past we have done so. That gives you an idea of the steps that need to be worked through post 30 June. Slide 36: I just wanted to pull together some of the key factors that we need to work through and need to be considered in the context of the 2014 valuation. So firstly market conditions. The market conditions are clearly a key factor to the assets, because the assets have to be marked to market. Also market conditions will affect views on future returns and will affect views on future inflation expectations and clearly those two have an important influence on the liabilities because 12