1 00:00:04,200 --> 00:00:11,760 So welcome, everyone. My name is Julian Savelescu. I'm the Uehiro chair in practical ethics and welcome to you here. 2 00:00:11,760 --> 00:00:20,280 Our lectures for 2020. The first you hear our lecture is jointly held with the Moral Philosophy Seminar. 3 00:00:20,280 --> 00:00:27,690 And I'll be passing over to Ed and Jeremy in a moment to formally introduce this year's speaker, 4 00:00:27,690 --> 00:00:32,190 Michael Otsuka, who is professor of philosophy, the London School of Economics. 5 00:00:32,190 --> 00:00:36,780 The title of his series is How to Pool Risks Across Generations. 6 00:00:36,780 --> 00:00:42,690 The Case for Collective Pensions. Michael Otsuka the 17th Uehiro lecturer. 7 00:00:42,690 --> 00:00:49,140 Past lecturers have included Geoff McMann, Peter Singer, Frances Kam, Philip Pettit, 8 00:00:49,140 --> 00:00:56,550 Janet Radcliffe Richards, Tim Scanln, Christine Koorsgard, Elizabeth Anderson and Samuel Scheffler. 9 00:00:56,550 --> 00:01:03,870 The lectures were founded in 2004 and have been funded by the Uehrio Foundation on Ethics and Education. 10 00:01:03,870 --> 00:01:11,250 And they're accompanied by the Uehiro series. In Practical Ethics, published by Oxford University Press. 11 00:01:11,250 --> 00:01:22,750 The twenty twenty one lectures will be given by Professor Howard McGarity at Rutgers University, and the timing will depend on the COVID 19 pandemic. 12 00:01:22,750 --> 00:01:33,930 So without further ado, let me hand over to Ed Lamb and Jeremy Fix to introduce today's lecturer and the speaker in more detail. 13 00:01:33,930 --> 00:01:38,440 Thanks very much I'm Ed Lamb the co convenor. 14 00:01:38,440 --> 00:01:45,540 Quote, The Moral Philosophy Seminar, it gives me great pleasure today to introduce Michael Otsuka, 15 00:01:45,540 --> 00:01:54,960 professor of philosophy at the London School of Economics, and he does Mike's work exemplify the virtues of clarity, simplicity and precision. 16 00:01:54,960 --> 00:02:01,980 He has played a significant role in many of the major debates of moral and political philosophy the last 30 years. 17 00:02:01,980 --> 00:02:09,360 More recently, his work has taken a more practical turn, focussing on issues such as risk pooling, insurance and pensions. 18 00:02:09,360 --> 00:02:19,100 His lectures will make the case for collective pensions beginning today with the case for a funded pension with a defined benefit. 19 00:02:19,100 --> 00:02:25,100 Thanks very much, Ed and Julian. So let me. 20 00:02:25,100 --> 00:02:41,840 Put up a slide. OK, so. 21 00:02:41,840 --> 00:02:48,740 Lot to gain by taking us back in time to 1990 or actually the mid 1990s. 22 00:02:48,740 --> 00:02:55,280 So chosen that decade because it's the last of a golden age for defined benefit pension schemes in the United Kingdom. 23 00:02:55,280 --> 00:03:00,880 And we're going to have a look at the books of the valuation of one such scheme in 1996. 24 00:03:00,880 --> 00:03:05,390 And I'm sure that you can guess the scheme will be well known to many people. 25 00:03:05,390 --> 00:03:10,430 It's called USAC University's Superannuation Scheme. 26 00:03:10,430 --> 00:03:12,110 For those of you who aren't completely familiar, 27 00:03:12,110 --> 00:03:21,290 USS consists of the roughly 70 UK universities that were founded from mediaeval times through to the late 1960s, 28 00:03:21,290 --> 00:03:30,540 plus the roughly 70 Oxford and Cambridge colleges. Now, back in 1996, and actually for most of USSR existence, 29 00:03:30,540 --> 00:03:42,930 promised members a pension equal to one 80th of one's final salary for every year of employment in exchange for employer and member contributions. 30 00:03:42,930 --> 00:03:50,340 There is also a guaranteed re-evaluation of one's pension each year by RPI increases in inflation. 31 00:03:50,340 --> 00:03:53,790 Now, that was how the benefit was defined, defined benefit. 32 00:03:53,790 --> 00:04:00,570 The promise was modelled on the pensions provided in the public sector to civil servants and school teachers at the time. 33 00:04:00,570 --> 00:04:11,550 Now the method employed to value the scheme back then was roughly along the lines that funded public sector schemes are valid in the United States. 34 00:04:11,550 --> 00:04:15,540 To this day, and in certain key respects, 35 00:04:15,540 --> 00:04:21,990 this is the method evaluation to which the union now representing the interests of scheme members and their actual 36 00:04:21,990 --> 00:04:32,390 actuarial adviser first actuarial would like you SS to return to at least insofar as current regulations permit. 37 00:04:32,390 --> 00:04:41,240 Now, they've argued that by returning to what they regard as sound actuarial and investment practises of the past. 38 00:04:41,240 --> 00:04:51,170 Can be shown to be possible. Continue to fund the current level of defined benefit promise at a contribution rate far lower than 39 00:04:51,170 --> 00:04:56,600 what both the pensions regulator and the trustee that administers the pension scheme are demanding. 40 00:04:56,600 --> 00:05:04,640 So question, should occupational pension schemes return to these practises of valuing and funding a pension scheme? 41 00:05:04,640 --> 00:05:10,550 That's the question I mean, to set out to address today. I want to leave you in too much suspense. 42 00:05:10,550 --> 00:05:13,310 My answer to the question will be at least a qualified yes. 43 00:05:13,310 --> 00:05:20,690 We should return to these historical practises because I need to say a few words about how a pension scheme is valued. 44 00:05:20,690 --> 00:05:25,850 I'm going to be examining the method of building a pension scheme today. 45 00:05:25,850 --> 00:05:32,560 Well. Going back to 1996, very different time for valuation. 46 00:05:32,560 --> 00:05:42,850 As of 1996, the assets in the pension fund were valued at one hundred and eight percent of the liabilities that had been accrued to date. 47 00:05:42,850 --> 00:05:49,630 Now, this valuation of the assets is greater than the liabilities implied that the scheme was in surplus. 48 00:05:49,630 --> 00:05:54,790 It was also projected that the scheme would be able to continue to make that same one 80th final salary 49 00:05:54,790 --> 00:06:03,020 defined benefit pension promise and each future year at the cost of only 20 percent of salaries. 50 00:06:03,020 --> 00:06:12,810 Well, the question where these reasonable and prudent assumptions grounded in sound methods of valuing pension schemes generally and if they were, 51 00:06:12,810 --> 00:06:17,210 why is USS now proposing that nothing less than an overall contribution, 52 00:06:17,210 --> 00:06:23,280 twice as high 40 percent of salaries will now be sufficient to provide what is actually 53 00:06:23,280 --> 00:06:32,100 a much Janner and A than the one area final salary that was promised in 1996? 54 00:06:32,100 --> 00:06:37,030 Now. To value pensions, liabilities or promises. 55 00:06:37,030 --> 00:06:43,000 The scheme actuary needs to predict what people's final salary will be under the final salary scheme. 56 00:06:43,000 --> 00:06:48,310 How long people will live in retirement. Looking at mortality tables on average. 57 00:06:48,310 --> 00:06:55,960 And what the future rate of inflation would be because from pensions go up by RPI inflation and then with these assumptions, 58 00:06:55,960 --> 00:07:00,850 then the actuary projects the stream of cash flows that would be paid out of the 59 00:07:00,850 --> 00:07:07,850 fund as pensions to scheme members in retirement decades out into the future. 60 00:07:07,850 --> 00:07:17,990 Now. In order to convert the sum of all these future cash flows into a single figure that captures their present value, 61 00:07:17,990 --> 00:07:26,450 what they actually did and and does to this day is discount this sum by the expected rate of return on the assets in the scheme. 62 00:07:26,450 --> 00:07:36,800 This is the so-called discount rate, which is by far the main source of controversy regarding the proper valuation of the pension scheme. 63 00:07:36,800 --> 00:07:42,290 So the higher the expected rate of return on the assets in future years, 64 00:07:42,290 --> 00:07:49,700 the lower the level of assets and pension contributions now needed to fully fund pension promises. 65 00:07:49,700 --> 00:07:59,570 Let's take us back in 1996, screenshot of the photograph of the cover of the valuation document in 1996. 66 00:07:59,570 --> 00:08:08,910 Now, back then. As you can see, the discount rate, the assumed rate of return on investments with eight point five percent per annum. 67 00:08:08,910 --> 00:08:13,650 That was their assessment of the returns learnt estimates that will be achieved 68 00:08:13,650 --> 00:08:18,380 on the scheme's assets per annum on average over the next several decades. 69 00:08:18,380 --> 00:08:24,500 It's also soon that inflation RPI. Back then, re five percent per annum. 70 00:08:24,500 --> 00:08:32,360 Now. Though this will strike you, as I think, very high assumptions, especially the investment return assumption, 71 00:08:32,360 --> 00:08:42,370 these assumptions were actually in line with an actual little bit more true, the more general actuarial practises at the time. 72 00:08:42,370 --> 00:08:48,490 Now, since for a scheme such as USFS, the liabilities pretty much rose and fell with inflation. 73 00:08:48,490 --> 00:08:54,630 What was most relevant was the margin between investment assumption and the inflation assumption. 74 00:08:54,630 --> 00:08:59,470 No, you assessed calculated, smarter margin with three point five percent, 75 00:08:59,470 --> 00:09:07,900 which is to say they assume that the assets would grow by three point five percent each year in real RPI inflation adjusted terms. 76 00:09:07,900 --> 00:09:14,810 RPI plus three point five question. Well, how sound were these assumptions? 77 00:09:14,810 --> 00:09:16,760 Now, in order to answer this question, 78 00:09:16,760 --> 00:09:25,880 we can actually draw now on almost a quarter of a century of actual returns on various types of assets since nineteen ninety six. 79 00:09:25,880 --> 00:09:33,830 Now, on the basis of U.S. assets modelling or simulating or the historical portfolio, it's actually move now to a different portfolio. 80 00:09:33,830 --> 00:09:35,180 But on the basis of their modelling, 81 00:09:35,180 --> 00:09:43,160 it appears that the scheme would in fact have achieved the assumed annualised real return of over three point five percent. 82 00:09:43,160 --> 00:09:54,000 I mean, not quite the nominal return, but they would have got. Three point five percent over RPI, at least, even up to thirty one July 20 20. 83 00:09:54,000 --> 00:09:59,180 If the scheme had remained invested as it was invested back then. 84 00:09:59,180 --> 00:10:05,530 So why then is the scheme now perceived in such financial difficulty? 85 00:10:05,530 --> 00:10:16,090 Well, it mainly comes down to the fact that for its latest valuation just issued for consultation process is proposing a discount rate, 86 00:10:16,090 --> 00:10:26,200 which assumes that the assets in the scheme will actually shrink by two tenths of a percent against RPI each year for the next 30 years, 87 00:10:26,200 --> 00:10:30,030 rather than growing by RPI plus three point five percent. 88 00:10:30,030 --> 00:10:39,210 Now, largely owing to such more pessimistic assumptions, USSA is now calling for a rise in contributions to least 40 percent of salaries, 89 00:10:39,210 --> 00:10:48,870 double what the contributions were in 1996 to find what is, as I mentioned now, a much less generous pension than what was on offer in 1996. 90 00:10:48,870 --> 00:10:59,970 So why is it that the current discount rate is now so much lower than three point five percent over RPI that was assumed in 1996? 91 00:10:59,970 --> 00:11:10,610 Well, there are three reasons, I think. Expected real long term returns on investments are now lower than they were in 1996. 92 00:11:10,610 --> 00:11:19,590 So on USS his most recent best estimates of expected returns on assets, the portfolio into which the scheme was invested, 93 00:11:19,590 --> 00:11:28,100 1996, is expected to grow by a bit less than three percent per annum above RPI over the next 30 years. 94 00:11:28,100 --> 00:11:35,420 Now note that you assess his current best estimates of real returns on equities, property and bonds. 95 00:11:35,420 --> 00:11:46,220 These estimates are therefore not that much less optimistic than they were back in 1996 and prior to the dot com bubble bursting in the early 2000s, 96 00:11:46,220 --> 00:11:55,030 prior to the global financial crisis of 2007 to 2008, and obviously prior to the current coronavirus pandemic, 97 00:11:55,030 --> 00:12:00,080 you would've thought that maybe they would've adjusted more so. But. But the. 98 00:12:00,080 --> 00:12:06,500 Thinking that returns won't be as high as as was thought in 1996 isn't the main reason why the discount rate is now so low. 99 00:12:06,500 --> 00:12:13,280 It just brings it down to three percent above RPI. Well, the second reason is much more significant for the lower discount rate. 100 00:12:13,280 --> 00:12:21,050 Second reason is that the portfolio out of which you assess will fund its defined benefit promises. 101 00:12:21,050 --> 00:12:28,200 Is now much more weighted towards bonds than the historical 1996 portfolio. 102 00:12:28,200 --> 00:12:36,220 Now, like many defined benefit portfolios, back in the 90s, USS was about 80 percent invested in equities. 103 00:12:36,220 --> 00:12:42,520 And that was sort of the investment pattern up until around the global financial crisis. 104 00:12:42,520 --> 00:12:50,750 Now, the portfolio into which U.S. US now plans to invest over the next several decades will actually end up averaging out. 105 00:12:50,750 --> 00:12:59,370 By contrast, two thirds bonds. So USS is shifting out of equities and into bonds. 106 00:12:59,370 --> 00:13:09,440 In spite of the fact that it actually expects equities to outperform UK government bonds by a huge amount, about 5.5 to six percent per annum. 107 00:13:09,440 --> 00:13:19,310 And USS also expects equities stocks and bonds to outperform corporate bonds by about two point seventy five percent over the next 30 years. 108 00:13:19,310 --> 00:13:28,510 And so this shift from equities to bonds significantly reduces the expected returns on the portfolio. 109 00:13:28,510 --> 00:13:32,080 Down from three point oh above RPI. 110 00:13:32,080 --> 00:13:36,770 That remain invested 80 percent in equities to one point to five percent above RPI. 111 00:13:36,770 --> 00:13:45,730 Now, these returns I've been quoting are those that one can expect on average per annum over the long term. 112 00:13:45,730 --> 00:13:55,990 Back in the nineteen nineties, contributions were set on the basis of such just pure best estimate expected returns on the assets. 113 00:13:55,990 --> 00:14:00,820 Now, by contrast, you assess this current valuation and in fact, 114 00:14:00,820 --> 00:14:08,150 all of its valuations from 2008 onwards have set contributions on the basis of more conservative assumptions. 115 00:14:08,150 --> 00:14:12,430 That one will have a significantly greater than 50 percent chance of achieving. 116 00:14:12,430 --> 00:14:18,230 There's a third reason why the discount rate is now so much lower than it was in nineteen ninety six. 117 00:14:18,230 --> 00:14:25,950 Now for the most recent valuation. Contributions have been proposed on the basis of returns that one can expect to 118 00:14:25,950 --> 00:14:29,940 achieve about three quarters of the time instead of roughly 50 percent of the time, 119 00:14:29,940 --> 00:14:35,820 and moving all the way up to no assumptions on the basis of what can be achieved three quarters of the time. 120 00:14:35,820 --> 00:14:42,720 And this is, in fact, how you assess interprets requirement of current regulations from 2008. 121 00:14:42,720 --> 00:14:50,010 Applying from 2008. The 2008 valuation. That the discount rate must, quote, be chosen prudently. 122 00:14:50,010 --> 00:14:52,840 And once we prudently adjust the discount rate, 123 00:14:52,840 --> 00:15:02,340 it soon returns on assets drops from one point to five above RPI to the two tenths of a percent below RPI. 124 00:15:02,340 --> 00:15:06,440 Or CPI plus five percent. 125 00:15:06,440 --> 00:15:15,060 Now, in pensions, contributions are set on the assumption that long term returns on the assets under which they'll be invested will fall short of RPI, 126 00:15:15,060 --> 00:15:20,550 as opposed to this assumption that prevailed in the 90s that returns will exceed RPI by three point five percent. 127 00:15:20,550 --> 00:15:30,150 It really shouldn't come as a surprise. That members and employers are now required to pay much more for a less good defined benefit pension. 128 00:15:30,150 --> 00:15:39,650 The question I want to ask now is. Is there a sound justification for this huge lowering of the discount rate? 129 00:15:39,650 --> 00:15:46,820 Now, to answer this question, we have to ask the following question. So how do we set the discount rate? 130 00:15:46,820 --> 00:15:50,750 What's the appropriate way in which to discount the pensions payment? 131 00:15:50,750 --> 00:15:59,870 Pensions promises many years in the future in order to convert those liabilities into a single present value? 132 00:15:59,870 --> 00:16:06,740 Well, let me just turn to the. Kurt. 133 00:16:06,740 --> 00:16:14,720 A lot of practise of the pensions regulator. It gives us two different methods of selling the discount rate, 134 00:16:14,720 --> 00:16:19,700 which correspond with two very different philosophies about how to set the discount rate. 135 00:16:19,700 --> 00:16:23,900 So the code says that the discount rate must be chosen prudently. 136 00:16:23,900 --> 00:16:28,830 That's the downward adjustment I talked about. Take me an account. 137 00:16:28,830 --> 00:16:30,150 Either the following two things, 138 00:16:30,150 --> 00:16:38,790 either the yield on assets held by the scheme to fund future benefits and the anticipated future investment returns and or. 139 00:16:38,790 --> 00:16:48,150 The market redemption yield on government or high quality, which is to say high quality corporate bonds. 140 00:16:48,150 --> 00:16:57,510 Now, in the 1996 valuation, which I described earlier, you assess fall, the first of the two related approaches say for one crucial difference. 141 00:16:57,510 --> 00:17:04,520 Back then, there was no requirement, as I mentioned, to discount the returns prudently. 142 00:17:04,520 --> 00:17:13,950 Now, accordingly, the USS Discount on the basis of their best estimates of the long term average rate of returns on the various assets in the scheme. 143 00:17:13,950 --> 00:17:20,250 Without this pessimistic adjustment, now, I'm going to argue that US's approach back then was actually the right one, 144 00:17:20,250 --> 00:17:24,960 save for the fact that they didn't make this prudent downward adjustment. 145 00:17:24,960 --> 00:17:30,330 Just add this prudence and and basically USS was following the right approach. 146 00:17:30,330 --> 00:17:36,060 Before I defend. You assess his approach back then. 147 00:17:36,060 --> 00:17:46,540 What I want to do is turn to an influential case grounded in a theory of value from financial economics for adoption of the second approach. 148 00:17:46,540 --> 00:17:55,760 And you can see from my red X, I've not been convinced by the financial economics case for the second approach. 149 00:17:55,760 --> 00:18:00,420 Now, here's the financial economics case. 150 00:18:00,420 --> 00:18:12,950 According this argument, the discounting of pensions liabilities has nothing to do with the return on the assets held in the schemes pension fund. 151 00:18:12,950 --> 00:18:18,440 Rather, irrespective of whether these assets are actually held by the scheme, 152 00:18:18,440 --> 00:18:25,160 what they believe is that the sound measure of the value of the liabilities is return on those financial assets that, quote, 153 00:18:25,160 --> 00:18:34,850 match the liabilities by providing streams of cash flows that match or approximate those of a promise pension 154 00:18:34,850 --> 00:18:43,680 in their magnitude and their duration and the level of certainty that these cash flows will be delivered. 155 00:18:43,680 --> 00:18:52,570 Well, here's a theory behind such a discount rate. Irrespective of how the pension funds invested, whether invested in equities or bonds, 156 00:18:52,570 --> 00:19:03,900 these liability matching securities provide an accurate measure of the value of the liability, which is the promised pension in the following respect. 157 00:19:03,900 --> 00:19:12,780 Anybody should be willing to exchange the pensions payments they had been promised for a portfolio portfolio of financial assets, 158 00:19:12,780 --> 00:19:18,910 which are guaranteed to as higher degree. To produce the promised cash flows. 159 00:19:18,910 --> 00:19:24,440 And these will be bonds of the right temporal duration, long dated bonds. 160 00:19:24,440 --> 00:19:32,510 Bonds also with inflation protection. And also bonds that reflect the credit risk of the employer who's making the pension 161 00:19:32,510 --> 00:19:38,460 promise the risk that that employer may actually not be able to deliver on that promise. 162 00:19:38,460 --> 00:19:43,610 I think the financial economics approach makes a lot of sense. 163 00:19:43,610 --> 00:19:47,870 For the valuation of a pensions promise, understood as the price. 164 00:19:47,870 --> 00:19:52,300 It makes sense to pay to purchase this promise. 165 00:19:52,300 --> 00:19:59,650 From someone, so the financial economics valuation is actually akin to the sorts of valuations with which we're familiar, 166 00:19:59,650 --> 00:20:05,050 such as the valuation of a house where we're trying to ascertain how much money that 167 00:20:05,050 --> 00:20:11,390 property would exchange for if placed on the market or what it would exchange for. 168 00:20:11,390 --> 00:20:20,510 I think it can be shown, however. That a so-called actuarial valuation of a pensions liability is actually very 169 00:20:20,510 --> 00:20:25,940 different from a determination of the market exchange value of that liability. 170 00:20:25,940 --> 00:20:32,990 And in fact, an actuarial valuation, that thing that used to come around every three years now tends to come around over a year, 171 00:20:32,990 --> 00:20:41,540 too, with you assess an actuarial valuation, doesn't actually involve evaluation of the liabilities. 172 00:20:41,540 --> 00:20:47,620 Rather, the purpose of such evaluation is as follows, 173 00:20:47,620 --> 00:20:55,190 it's to determine whether the pensions contributions placed into the scheme to date have been invested in a matter that is, 174 00:20:55,190 --> 00:21:00,920 quote, sufficient and appropriate to make provision for the scheme's liabilities. 175 00:21:00,920 --> 00:21:09,240 And this purpose is captured by a term known as the technical provisions. Now, as it stated in the UK Pensions Act 2004. 176 00:21:09,240 --> 00:21:12,880 I mean, you can find it in other countries as well. 177 00:21:12,880 --> 00:21:20,950 Every scheme is subject to a requirement that it must have sufficient and appropriate assets to cover its technical provisions, 178 00:21:20,950 --> 00:21:31,910 and the scheme's technical origins means the amount required on an actuarial calculation to make provision for the scheme's liabilities. 179 00:21:31,910 --> 00:21:40,800 Stressed the amount required to provide for the liabilities is not necessarily the value market or otherwise of these liabilities. 180 00:21:40,800 --> 00:21:50,010 What I want to turn to you now are two arguments for why the yield of a bond portfolio that matches 181 00:21:50,010 --> 00:21:55,650 the liabilities does not necessarily provide the discount rate for their technical provisions, 182 00:21:55,650 --> 00:22:00,070 as we've just defined it. Here's my first argument. 183 00:22:00,070 --> 00:22:08,320 Now, first, our report begins with the observation that existing and sound regulatory practises typically require that a weaker, 184 00:22:08,320 --> 00:22:13,750 less credit worthy sponsoring employer of a pension scheme needs to put in more 185 00:22:13,750 --> 00:22:19,900 contributions in order to make provision for a given pension promise than a stronger, 186 00:22:19,900 --> 00:22:24,340 more credit worthy sponsoring employer needs to put into the pension fund. 187 00:22:24,340 --> 00:22:32,830 The rationale is pretty obvious. One can allow more credit worthiness bonds to stay. 188 00:22:32,830 --> 00:22:39,730 Or other and stronger and pensions terminology. To put in less by way of contributions now, 189 00:22:39,730 --> 00:22:47,570 because we know that they are more able to put in more contributions later if returns on the assets in which these contributions have invested. 190 00:22:47,570 --> 00:22:49,510 Turnout lower than expected. 191 00:22:49,510 --> 00:22:55,780 But alas, less credit worthy sponsors, pensions, farmers will be worth less to a worker than a more credit worthy sponsors. 192 00:22:55,780 --> 00:23:00,250 Promise of the same pension payments for the simple reason that a worker will be rightly 193 00:23:00,250 --> 00:23:05,020 less certain that he'll end up receiving the full amount of the pension he's been promised. 194 00:23:05,020 --> 00:23:08,800 So a defined benefit pension that Trinity College, 195 00:23:08,800 --> 00:23:19,080 Cambridge promises will be worth worth more than a pension promise of a single employer scheme whose sole sponsor is on the edge of insolvency. 196 00:23:19,080 --> 00:23:25,870 It doesn't follow, however, that Trinity College must be required to put more money into its pension fund than an 197 00:23:25,870 --> 00:23:31,180 employer on the brink of insolvency in order to make sufficient provision for its pensions. 198 00:23:31,180 --> 00:23:41,290 Given the primary objective, the pensions regulations to ensure that the benefits promised to members are paid, the very opposite is true. 199 00:23:41,290 --> 00:23:45,310 Setting the technical revisions as the market exchange value, I mean, 200 00:23:45,310 --> 00:23:50,800 saying how much you have to provide for selling that as a market exchange value of a pensions promise would have the perverse 201 00:23:50,800 --> 00:23:58,110 consequence of requiring Trinity College to put in more money to secure its pensions promise than an employer on the edge of insolvency. 202 00:23:58,110 --> 00:24:07,130 And that's absurd. Nine turned my second argument for why the bond yield of a portfolio that matches the liabilities 203 00:24:07,130 --> 00:24:13,970 doesn't necessarily provide the discount rate for the technical provisions as defined. 204 00:24:13,970 --> 00:24:20,270 Remember, the technical provisions tells us the extent to which one must make provision for pensions 205 00:24:20,270 --> 00:24:28,970 promises by funding them in advance rather than ours and when the promised payments come due. 206 00:24:28,970 --> 00:24:35,350 So the technical provisions is therefore essentially a collateral requirement. 207 00:24:35,350 --> 00:24:42,430 And as my earlier discussion has revealed, it follows the nature of a pensions promise that bonds will provide a closer match 208 00:24:42,430 --> 00:24:49,360 to a better substitute for a pension than other financial assets such as equities. 209 00:24:49,360 --> 00:24:50,710 But as I shall now demonstrate. 210 00:24:50,710 --> 00:25:01,630 It is, however, a fallacy to maintain that the bond like nature of a pension liability gives rise to any requirement to fund it out of matching bonds. 211 00:25:01,630 --> 00:25:10,270 Or, if not that, at least funding it to a level sufficient to purchase these matching bonds. 212 00:25:10,270 --> 00:25:11,730 So here's the moment, a regulated, 213 00:25:11,730 --> 00:25:21,270 defined benefit pension is essentially a non tradable bond issued by means of the trustee, by a corporation to an employee. 214 00:25:21,270 --> 00:25:27,390 The employee issues a bond to you in exchange for contributions. 215 00:25:27,390 --> 00:25:29,800 A corporation may, however, 216 00:25:29,800 --> 00:25:39,130 issue unsecured corporate bounds bonds to their employees that pretty much replicate most of the characteristics of a defined benefit pension. 217 00:25:39,130 --> 00:25:47,910 Moreover, regulatory requirements for unsecured corporate bonds don't extend much beyond provisions to guard against fraud. 218 00:25:47,910 --> 00:25:57,420 Now, since these corporate bonds are unsecured, then any funding or collateral requirements now, 219 00:25:57,420 --> 00:26:03,870 the relative lack of regulation of such corporate bonds renders it difficult to sustain the view that defined benefit 220 00:26:03,870 --> 00:26:11,670 pensions must be funded up to a level sufficient to allow for the purchase of matching bonds to the liabilities. 221 00:26:11,670 --> 00:26:22,960 Because the mere fact that a defined benefit pension involves a bond like promise does not establish that it must be funded to this level. 222 00:26:22,960 --> 00:26:25,680 If it did, then by parity of reasoning, 223 00:26:25,680 --> 00:26:34,030 one would need to insist that a corporation may not issue a bond unless it secures it with assets that match the liability. 224 00:26:34,030 --> 00:26:43,110 They would, however, defeat the point of issuing corporate bonds if one were required to secure them in this manner from the outset. 225 00:26:43,110 --> 00:26:51,140 Since anyone would hardly be able to make use of the revenue from the sale of bonds to invest in anything other than the securing of that very bond. 226 00:26:51,140 --> 00:26:55,710 We could no longer issue a bond to raise money to spend on other things. 227 00:26:55,710 --> 00:27:01,740 Rather, one would have to use up the money from the sale of the bond monies issued to purchase a comparable bond to serve as collateral. 228 00:27:01,740 --> 00:27:11,700 That's absurd. It was therefore a mistake for financial economists to claim that the bond like nature of a pensions promise gives rise to a 229 00:27:11,700 --> 00:27:22,470 requirement that it be secured by means of funding to the level of the value of a liability portfolio of liability matching bonds. 230 00:27:22,470 --> 00:27:33,930 I've shown that financial economics doesn't provide a justification for insisting that the pensions promise must be secured by contributions, 231 00:27:33,930 --> 00:27:42,840 ring fenced in, the fun guarded by a trustee whose level should be set so high that it is possible to 232 00:27:42,840 --> 00:27:48,090 purchase a bond portfolio that comes very close to matching the defined benefit liability. 233 00:27:48,090 --> 00:27:55,940 Question I want to turn to now is the following. Is there, however, a sound regulatory justification? 234 00:27:55,940 --> 00:28:00,420 This is really what the question is about. What about? 235 00:28:00,420 --> 00:28:03,740 Financial economics and their theory of value. 236 00:28:03,740 --> 00:28:10,980 But is there a regulatory justification grounded in sound public policy considerations rather than the theory of value? 237 00:28:10,980 --> 00:28:17,060 For saying that you have to provide for pensions promised up to love. 238 00:28:17,060 --> 00:28:27,090 Well, imagine. And I now turn to this question that we this reframed question, properly framed question about setting the technical divisions. 239 00:28:27,090 --> 00:28:32,080 What I want to do now is present what I take to be the strongest argument. 240 00:28:32,080 --> 00:28:39,160 For the relevance of the yield on the liability matching bond portfolio to the setting of the tactical provisions to the setting the discount rate 241 00:28:39,160 --> 00:28:48,760 for the purpose of determining whether the assets held by the scheme are sufficient to provide for the pensions that have been promised to members. 242 00:28:48,760 --> 00:28:57,510 Now, recall that I mentioned earlier. That strong movement sponsoring employer hire, 243 00:28:57,510 --> 00:29:04,440 the assumed rate of return on the assets by which the regulator will allow them to discount the liabilities. 244 00:29:04,440 --> 00:29:08,220 It's based on a theory that a financially strong employer is more capable and a weaker 245 00:29:08,220 --> 00:29:13,610 employer of underwriting a portfolio of growth assets such as equities stocks and bonds. 246 00:29:13,610 --> 00:29:16,360 That's on account of. 247 00:29:16,360 --> 00:29:25,450 What's called the equity risk premium, that a growth portfolio will have a higher expected return than a government bond portfolio, 248 00:29:25,450 --> 00:29:34,300 where the promise rate of return of a bond portfolio is from a reliable government is regarded as risk free. 249 00:29:34,300 --> 00:29:44,430 The higher expected return. On equities is regarded as the compensation necessary to induce individuals to purchase equities, 250 00:29:44,430 --> 00:29:51,170 given the greater risks involved in holding stocks in comparison with bonds. 251 00:29:51,170 --> 00:29:56,780 No regulator will, however, insensibly. If someone says I want to invest in equities. 252 00:29:56,780 --> 00:30:04,950 Require the sponsors to underwrite this risk. The sponsor will need to demonstrate that they would be able to put more assets into the 253 00:30:04,950 --> 00:30:10,390 pension fund in the event that the returns on growth assets are less good than expected. 254 00:30:10,390 --> 00:30:16,860 Now, a week sponsor who is less able to demonstrate the ability to make good large funding shortfalls, 255 00:30:16,860 --> 00:30:22,320 will actually be required to invest expensively at the outset in financial assets 256 00:30:22,320 --> 00:30:27,980 that run a much less great risk of falling short of covering the pensions promises. 257 00:30:27,980 --> 00:30:33,840 A sponsoring employer who is so weak that they can't demonstrate any ability to make good any shortfall, 258 00:30:33,840 --> 00:30:37,420 even a year or two in the future will be required. Honestly, 259 00:30:37,420 --> 00:30:45,450 it's almost entirely prefund the pension promise at the very outset when they make the promise by investing at the 260 00:30:45,450 --> 00:30:53,290 outset in assets that come close to perfectly matching the pensions liability in their duration and magnitude, 261 00:30:53,290 --> 00:30:58,640 at least putting enough money in the fund at the outset that you can purchase such a portfolio. 262 00:30:58,640 --> 00:31:08,160 A long dated, inflation linked government bonds come closest amongst financial assets. 263 00:31:08,160 --> 00:31:12,780 PARFUM insurance company annuities to perfectly match their pensions liability. 264 00:31:12,780 --> 00:31:21,300 Such a portfolio will approximate the composition of a portfolio that actually matches in insurance companies annuity and therefore a very 265 00:31:21,300 --> 00:31:30,940 weak sponsor will be required at great expense to discount pensions liabilities at something close to the risk free government bond yield. 266 00:31:30,940 --> 00:31:37,940 Now, stronger response is not required to invest in a portfolio bonds that match the liabilities. 267 00:31:37,940 --> 00:31:42,200 They'll be permitted to invest at far less expense in a growth portfolio. 268 00:31:42,200 --> 00:31:49,070 Who's higher returns are expected to cover the promised pensions payments in future years. 269 00:31:49,070 --> 00:31:55,700 They'll be allowed to do so so long as they can demonstrate an ability to underwrite the risks by showing how they'll 270 00:31:55,700 --> 00:32:02,390 deliver on their pensions promises in the event that the returns on these risky assets are less good than expected. 271 00:32:02,390 --> 00:32:09,830 So how much they demonstrate this? Well, here's one way USSS proposed so-called Safe Harbour. 272 00:32:09,830 --> 00:32:16,130 One way of underwriting the risk is showing that in the event that the downside risks of the growth portfolio materialise, 273 00:32:16,130 --> 00:32:24,270 one, we'll still have the wherewithal to purchase a liability matching bond portfolio that has this very high chance. 274 00:32:24,270 --> 00:32:29,530 Of all the promised. That's for this reason, amongst others, 275 00:32:29,530 --> 00:32:37,270 that the regulator pays close attention to the extent to which the assumed rate of return on growth assets held by the scheme, 276 00:32:37,270 --> 00:32:43,500 which is to say the discount rate on growth assets exceeds the risk free rate of government bonds. 277 00:32:43,500 --> 00:32:51,030 A sponsor who holds fewer growth assets in the scheme. 278 00:32:51,030 --> 00:33:00,000 ServiceMaster, whoo, whoo hoo hoo! Has growth assets worth less in the scheme than liability matching assets is 279 00:33:00,000 --> 00:33:03,840 going to need to generate the following that he has resources to make good. 280 00:33:03,840 --> 00:33:09,450 The gap between the lower price of a growth portfolio with higher expected 281 00:33:09,450 --> 00:33:15,720 returns and the higher price of a bond portfolio with a lower expected returns. 282 00:33:15,720 --> 00:33:25,710 Even in the event that this gap grows even larger on account of either a fall in the stock market or a decline in the yield of bonds. 283 00:33:25,710 --> 00:33:32,370 Now, the useful metaphor of USSS chief risk officer is the following. 284 00:33:32,370 --> 00:33:43,190 He's described liability matching bond portfolio as a safe harbour to lock down the assets that must be within affordable reach of the sponsor. 285 00:33:43,190 --> 00:33:47,540 Can't show you a graph. Now. 286 00:33:47,540 --> 00:33:58,650 This graph, two graphs provide a key to understand, maybe the key to understand the difficulty that defined benefit pension schemes. 287 00:33:58,650 --> 00:34:05,100 Are in now, given the focus on the significance of bond yields. 288 00:34:05,100 --> 00:34:09,650 Now, we can see from this graph that back in the mid 90s, 289 00:34:09,650 --> 00:34:18,680 the risk free rate on long dated government bonds was actually a good deal higher than the risk free rate is now, in fact, back in the mid 90s. 290 00:34:18,680 --> 00:34:23,720 The risk free rate on government bonds that match pensions liabilities was so high that even the USS 291 00:34:23,720 --> 00:34:29,540 was heavily invested in equities and assuming a very high eight and a half percent rate of return. 292 00:34:29,540 --> 00:34:38,220 So putting very few assets in at the outset and assuming that it would mostly come through through returns on investment. 293 00:34:38,220 --> 00:34:51,060 USS, nevertheless, had more than enough assets in the fund in 1996 to cover the purchase of liability matching bonds back in 1996, USS. 294 00:34:51,060 --> 00:34:58,620 Was one hundred percent once. One hundred and twenty five percent funded on a solvency basis, 295 00:34:58,620 --> 00:35:02,280 which was pegged to the cost of purchasing a mixture of index-linked and fixed interest 296 00:35:02,280 --> 00:35:07,260 sterling securities that replicated the portfolio of an insurance company's annuity. 297 00:35:07,260 --> 00:35:18,090 So back in 1996, there was plenty of money to spare in the scheme, if you want to move all the way over to a buyout by an insurance company. 298 00:35:18,090 --> 00:35:26,540 Well, at its most recent 2008 valuation. Instead of being one hundred and twenty five percent funded on this buyout solvency basis, 299 00:35:26,540 --> 00:35:34,730 USSR was only 56 percent funded on a solvency basis, which is to say that the assets held in the scheme. 300 00:35:34,730 --> 00:35:43,570 At the 2008 evaluation, they haven't yet figured out the solvency evaluation for 2020 for the 2018 valuation. 301 00:35:43,570 --> 00:35:52,120 It was worth only 30 percent, six percent of the cost of the purchase of this liability matching bond portfolio. 302 00:35:52,120 --> 00:35:56,380 So employers would have to come up with an extra 50 billion. 303 00:35:56,380 --> 00:36:05,670 Mean talk about deficits. 50 billion pounds in 2018 in order to raise enough money to purchase such a liability 304 00:36:05,670 --> 00:36:11,630 matching bond portfolio that replicates the insurance companies annuity portfolio. 305 00:36:11,630 --> 00:36:16,390 Therefore. Unlike in the 1990s and the early 2000s, 306 00:36:16,390 --> 00:36:28,220 when the safe harbour of a liability matching bond portfolio could be purchased in exchange for the assets under the scheme with money to spare. 307 00:36:28,220 --> 00:36:37,640 Today, such a bond portfolio is far beyond the reach of what employers could afford to move to in the short term. 308 00:36:37,640 --> 00:36:46,800 Now, USF doesn't actually require that the employer must to immediately to move to a liability matching bond portfolio. 309 00:36:46,800 --> 00:36:52,250 USSS current requirement is that a safe harbour must be within affordable reach. 310 00:36:52,250 --> 00:36:54,610 In the following respect. 311 00:36:54,610 --> 00:37:04,270 It must be possible to raise sufficient revenues by means that extra contributions of 10 percent of payroll over the next three decades. 312 00:37:04,270 --> 00:37:12,470 In order to enhance a market value of the assets in the portfolio. To a level which makes it possible to purchase a liability matching bond portfolio. 313 00:37:12,470 --> 00:37:17,090 Actually not quite a stringent requirement of the liability matching portfolios. 314 00:37:17,090 --> 00:37:20,990 The bio portfolio. Why are they chosen? 315 00:37:20,990 --> 00:37:25,100 Three decades of an extra 10 percent in contributions. 316 00:37:25,100 --> 00:37:32,420 But the time period of three decades has been chosen because it's been determined that the higher education sector will at least 317 00:37:32,420 --> 00:37:40,250 collectively be in sufficiently strong financial condition to pay such extra contributions for at least the next three decades. 318 00:37:40,250 --> 00:37:48,360 The covenant is strong for at least that long. But bond yields, as you can see from the graph. 319 00:37:48,360 --> 00:37:49,390 Now, 320 00:37:49,390 --> 00:38:01,200 the little the Michael the action portfolio is so expensive that USS would now fall short of being able to purchase a liability matching portfolio, 321 00:38:01,200 --> 00:38:06,490 even with 30 years of Cygnet significant hands contributions. So the question now is, Will? 322 00:38:06,490 --> 00:38:17,950 Does a defined benefit scheme really need, as you assess, insists it does, to be within affordable reach of a liability matching bond portfolio? 323 00:38:17,950 --> 00:38:23,260 Or my a defined benefit scheme prudently fund pensions promises out of a portfolio of 324 00:38:23,260 --> 00:38:31,210 growth assets whose value falls so far short of that of a liability matching portfolio. 325 00:38:31,210 --> 00:38:39,910 That one couldn't afford to purchase such a portfolio even after three decades of an extra 10 percent pensions contributions. 326 00:38:39,910 --> 00:38:44,600 Well, I think the answer is. No. 327 00:38:44,600 --> 00:38:51,310 We don't need the safe harbour. And here's why such a safe harbour is unnecessary. 328 00:38:51,310 --> 00:38:58,980 So long as one builds in a prudent margin. By which to returns on growth outfits. 329 00:38:58,980 --> 00:39:06,490 Might fall short of a best estimate while still covering the costs of promised pensions. 330 00:39:06,490 --> 00:39:10,630 The risks of remaining largely invest in growth assets are minimal for an open, 331 00:39:10,630 --> 00:39:18,430 ongoing cash flow positive scheme with a long investment horizon such as USS. 332 00:39:18,430 --> 00:39:28,820 That's recently been calculated that returns as modest as CPI plus one point two percent, which translates now into RPI plus five tenths of a percent. 333 00:39:28,820 --> 00:39:36,540 That returns only that low. Are all of us needed to fully fund the current U.S. as defined benefit pension promise? 334 00:39:36,540 --> 00:39:45,280 Out of contributions as low as 26 percent of payroll, which is back what we were paying couple of years ago. 335 00:39:45,280 --> 00:39:51,480 You assess his best estimate of returns on his current portfolio, which is now 65 percent weighted towards growth. 336 00:39:51,480 --> 00:39:59,820 Assets are going to be moving towards bonds in the future. You assess as best estimates of returns on its current 65 percent equities and property 337 00:39:59,820 --> 00:40:05,310 portfolio is about CPI plus three point two percent or RPI plus two point five percent. 338 00:40:05,310 --> 00:40:09,540 That's their best guess as to what returns they'll get per annum. 339 00:40:09,540 --> 00:40:14,970 And this is comfortably two percent above the needed return of CPI, 340 00:40:14,970 --> 00:40:22,530 plus one point two percent to fund our current promise at twenty eight percent contributions. 341 00:40:22,530 --> 00:40:32,520 Seems like quite a margin of prudence now. There is, of course, a risk that growth assets will fall. 342 00:40:32,520 --> 00:40:40,560 More than two percent below their expected return and therefore fail to cover the promised pensions. 343 00:40:40,560 --> 00:40:46,820 The question is, Will, how great is this risk? Of falling below the margin of prudence. 344 00:40:46,820 --> 00:40:55,100 Well. Stochastic modelling that you assessors engage in shows the higher risk investments tend to 345 00:40:55,100 --> 00:40:59,840 perform better over the long term and that the probability of loss decreases over time. 346 00:40:59,840 --> 00:41:06,340 And this is represented by many other people. Seems relatively safe. 347 00:41:06,340 --> 00:41:09,960 So long as you can hold. Growth outside of the long term. 348 00:41:09,960 --> 00:41:17,210 Unfortunately, however. The following significant downside risk remains. 349 00:41:17,210 --> 00:41:28,000 As USSR assault also points out, and this is replicate, did another stochastic modelling the punch, the potential size of any loss that occurs? 350 00:41:28,000 --> 00:41:35,700 Increases with time as well. And in particular, the magnitude of the worst possible outcome increases with time. 351 00:41:35,700 --> 00:41:42,030 And moreover, it appears to be a concern regarding such extreme downside risk, 352 00:41:42,030 --> 00:41:53,220 which actually explains the existence of demand for long dated bonds issued by governments and other institutions with excellent credit ratings. 353 00:41:53,220 --> 00:42:06,980 So as. Exley made Smith explain it as being pencil piece, defending the financial economics approach to devaluating pensions liabilities. 354 00:42:06,980 --> 00:42:13,220 The belief, the true belief that it's highly likely in the long term that equities will generate sufficient returns to cover one's liabilities, 355 00:42:13,220 --> 00:42:25,590 quote, manifestly fails to explain why anyone would be prepared to purchase long dated bonds at the market, writes. 356 00:42:25,590 --> 00:42:33,720 If they be sympathised for so much less, I'll have a portfolio of equities as is being proposed. 357 00:42:33,720 --> 00:42:41,010 The key must lie in the unlikely event where the equities fail to be adequate to explain why people purchase bonds. 358 00:42:41,010 --> 00:42:47,040 Although the probability of this event may be small, it bears heavily on the minds of investors as a whole. 359 00:42:47,040 --> 00:42:51,580 Those who buy bonds. And I said, we can see why. 360 00:42:51,580 --> 00:42:58,940 A promise which holds fast when everything else is collapse all around is potentially very valuable. 361 00:42:58,940 --> 00:43:07,550 Moreover. A funded, defined benefit promise is just the sort of thing that one is meant to be able to rely on, 362 00:43:07,550 --> 00:43:12,680 even in adverse circumstances in which everything else is collapsed all around. 363 00:43:12,680 --> 00:43:21,630 So we can see that in the midst of a global pandemic, just giving rise to a record breaking recession and unemployment, 364 00:43:21,630 --> 00:43:32,220 defined benefit pensions promises are still pay on time and in full by private funded pension schemes. 365 00:43:32,220 --> 00:43:40,910 But he's actually made a Smith note. If anyone could synthesise equities into assets that provide the same cash flows, bonds come what may. 366 00:43:40,910 --> 00:43:49,370 More cheaply then nobody would ever purchase long dated bonds. There is, however, clearly high demand for such bonds. 367 00:43:49,370 --> 00:43:57,080 So how do we respond to this challenge? Well. To respond, this sounds when you do the fine we need. 368 00:43:57,080 --> 00:44:02,780 To show. But at least some defined benefit pension schemes. 369 00:44:02,780 --> 00:44:09,670 Are actually in a better position than others to manage the downside risk of equities. 370 00:44:09,670 --> 00:44:20,050 Even though others find it rational to purchase bonds because the downside risks of equities are too great for them to manage. 371 00:44:20,050 --> 00:44:24,700 The explanation has to be that well defined benefit schemes have certain opportunities to manage the 372 00:44:24,700 --> 00:44:29,560 risk of equities that are not widely available and that other people have to resort to bonds to manage. 373 00:44:29,560 --> 00:44:39,960 Well, how might we? So this. Well, here, I think, is how a defined benefit pension scheme can navigate the open seas of equities. 374 00:44:39,960 --> 00:44:51,300 Even when too far from the safe harbour of a liability matching bond, pour in a huge if necessary. 375 00:44:51,300 --> 00:44:57,930 So here's how a scheme can hunker down and ride out the vortex of even a large drop in equities. 376 00:44:57,930 --> 00:45:04,890 So long as the scheme can be foreseen to be cash flow positive for the next several decades, as in the case of USS. 377 00:45:04,890 --> 00:45:13,970 There's actually no need to quickly repair our deficit and funding in order to ensure that pensions are paid as they fall due. 378 00:45:13,970 --> 00:45:24,990 Because these tensions involve many small and predictable payments that stretch several decades in the future. 379 00:45:24,990 --> 00:45:32,530 An ongoing enduring open. Cash flow positive scheme can therefore safely spread out DFC. 380 00:45:32,530 --> 00:45:35,740 Deficit recovery payments over a long period of time. 381 00:45:35,740 --> 00:45:41,920 For example, the three decades or which USSS covenant has been rated strong in spread out deficit, 382 00:45:41,920 --> 00:45:50,670 covering repayments to to make good a funding shortfall over three decades in order to recover from even very large losses without, 383 00:45:50,670 --> 00:45:59,670 in so doing, running the risk of defaulting on the promises of the pensions they promise to pay as they fall to. 384 00:45:59,670 --> 00:46:12,610 Now, it's here that. The multiemployer nature of certain defined benefit schemes such as USS plays a crucial role. 385 00:46:12,610 --> 00:46:20,300 And USS. Hundreds and the institution. 386 00:46:20,300 --> 00:46:26,300 Even if not every single employer can be counted on to be around and financially strong 387 00:46:26,300 --> 00:46:30,710 enough to make deficit recovery payments that are spread out over the next three decades. 388 00:46:30,710 --> 00:46:40,590 The UK higher education sector as a collective can be counted on to endure and remain sufficiently financially solvent for at least that long. 389 00:46:40,590 --> 00:46:51,600 Now, the so-called. Last man standing. Mutuality of USS ensures that there will be this enduring corporate body capable of 390 00:46:51,600 --> 00:47:00,090 making good on debts whose repayments have been amortised over several decades. 391 00:47:00,090 --> 00:47:06,750 Now on the last man standing arrangement, if any one employer in the scheme becomes insolvent. 392 00:47:06,750 --> 00:47:13,560 Responsibility to make good any underfunding, other promises is automatically shared out by the other employers, 393 00:47:13,560 --> 00:47:18,930 spread out amongst them through small increases in their contributions. 394 00:47:18,930 --> 00:47:27,900 Now, this last man standing structure is a form of insurance on the principle spreading of losses in this case from insolvency. 395 00:47:27,900 --> 00:47:35,400 Even the intently amongst the large number of remaining solvent universities. 396 00:47:35,400 --> 00:47:45,520 And this is the insurance of a mutual association which predates the rise of sophisticated financial instruments. 397 00:47:45,520 --> 00:47:49,540 This is a form of insurance through mutualisation that doesn't rely on a transaction 398 00:47:49,540 --> 00:47:53,170 with an outside insurance company or government agency to secure protection, 399 00:47:53,170 --> 00:47:59,940 which is expensive these days, expensively underwritten by a bond portfolio. 400 00:47:59,940 --> 00:48:04,620 But because allows for long term investment in equities and the way of explained the insurance of a 401 00:48:04,620 --> 00:48:10,860 mutual association is now more cost effective than insurance that requires the backing of bonds, 402 00:48:10,860 --> 00:48:19,650 given how low the bond yield is now. The mutuality of the one hundred and forty or so UK universities and Oxbridge colleges 403 00:48:19,650 --> 00:48:25,020 that form the membership of USS provides insurance against the systemic risks. 404 00:48:25,020 --> 00:48:34,430 The fall in the value of their assets as a result of recession, global financial crisis, pandemic or the like. 405 00:48:34,430 --> 00:48:40,310 Now we saw that the insolvency risk. 406 00:48:40,310 --> 00:48:45,120 And I'm seeing some comments in the chat box, things all right? 407 00:48:45,120 --> 00:48:51,610 I mean, I'm seeing a bit of flickering. It's just somebody that's putting a question that I'll call on when we get to the Q&A sessions. 408 00:48:51,610 --> 00:48:56,200 So nothing for you to worry about. OK, thanks, great. 409 00:48:56,200 --> 00:49:06,580 So we saw the insolvency risk and last man standing scheme is spread out geographically in space across all the sponsoring employers. 410 00:49:06,580 --> 00:49:12,490 By contrast, systemic investment risk. I mean investment risk that say systematic. 411 00:49:12,490 --> 00:49:17,510 I mean, when the stock market crashes, it affects everyone. 412 00:49:17,510 --> 00:49:22,940 Can also be spread out. But in this case, rather than is spread out geographically, you can't do that. 413 00:49:22,940 --> 00:49:28,850 I mean, the the stock market affects everyone. We spread out a systemic form, 414 00:49:28,850 --> 00:49:34,850 the stock market across future time slices of the multigenerational collective body that 415 00:49:34,850 --> 00:49:42,270 constitutes a multiemployer scheme to spread out the risks over time rather than over space. 416 00:49:42,270 --> 00:49:47,510 So pensions promises are secured by the ability to make good shortfalls on the 417 00:49:47,510 --> 00:49:53,170 returns on equities to extra contributions spread out over multiple decades. 418 00:49:53,170 --> 00:49:58,480 This method of underwriting investment risk provides an alternative to the requirement to remaining 419 00:49:58,480 --> 00:50:07,510 within this affordable 30 year journey to purchase of a liability matching bond portfolio. 420 00:50:07,510 --> 00:50:11,590 Now, with the insurance underpinning of Matt last man standing, 421 00:50:11,590 --> 00:50:19,450 which Welday financial assets are freed up for constant and long term investment in a manner that seeks higher returns, 422 00:50:19,450 --> 00:50:28,010 rather than being tied down to the need to provide collateral in the form of liability matching bonds to secure the pensions. 423 00:50:28,010 --> 00:50:40,150 Now, this discussion underscores the importance of keeping defined benefit schemes open and cash flow positive for as long as possible. 424 00:50:40,150 --> 00:50:44,950 And the most effective means of keeping defined benefit schemes open for as long 425 00:50:44,950 --> 00:50:50,610 as possible is through multiemployer arrangements that are wide and inclusive. 426 00:50:50,610 --> 00:50:59,880 So although individual employers, even industries will come and go in economy of producers of goods and services is here to stay. 427 00:50:59,880 --> 00:51:05,400 And there is therefore a public policy justification for bringing producers together in large 428 00:51:05,400 --> 00:51:13,790 networks of multiemployer schemes like you assess where they will be stronger together. 429 00:51:13,790 --> 00:51:16,670 To achieve this, this is a regulatory public policy issue, 430 00:51:16,670 --> 00:51:22,460 not a matter of financial economics valuate not a matter of valuation theory to achieve this. 431 00:51:22,460 --> 00:51:25,990 The state should extend tax relief. 432 00:51:25,990 --> 00:51:33,310 Especially, or maybe even only to those occupational pension schemes that enter into salts multiemployer arrangements, 433 00:51:33,310 --> 00:51:39,370 which they would also also facilitate to make sure that everyone can enter into a multiemployer arrangement. 434 00:51:39,370 --> 00:51:48,680 The state should therefore provide an incentive. For firms to engage in such risk pooling for mutual advantage that gives rise to such an enduring 435 00:51:48,680 --> 00:51:59,250 party that it can spread the risks of drops in equity safely over long periods of time. 436 00:51:59,250 --> 00:52:07,270 Even if we. Provide such tax incentives to join together. 437 00:52:07,270 --> 00:52:11,710 Four firms joined together in a multiemployer last man standing arrangements. 438 00:52:11,710 --> 00:52:17,080 There will be some employers who fear that they'll actually end up the last man standing. 439 00:52:17,080 --> 00:52:21,950 And those employers may choose to go it alone, even incorporate cost. 440 00:52:21,950 --> 00:52:25,800 The last man standing arrangement. Because ultimately, 441 00:52:25,800 --> 00:52:32,370 the last solvent employer bears responsibility to deliver all the pensions of the members 442 00:52:32,370 --> 00:52:39,900 of the scheme in the event of the insolvencies of all the other employers in the scheme. 443 00:52:39,900 --> 00:52:47,700 And Trinity College, Cambridge famously did just this when they recently withdrew from USS fairly great expense. 444 00:52:47,700 --> 00:52:56,600 They bought out their liabilities at the very high expense of an insurance company. 445 00:52:56,600 --> 00:53:07,810 I want to draw this lecture to close. By noting a striking feature of the actuarial advice on which Trinity College chose to withdraw. 446 00:53:07,810 --> 00:53:16,890 This is vice identified a move to a liability matching portfolio as actually the major source of risk to the sector as a whole. 447 00:53:16,890 --> 00:53:22,960 The scenario illustrating the existential threat that US as last man standing arrangement might 448 00:53:22,960 --> 00:53:31,240 pose to Trinity involve following the risk of high deficit recovery contributions caused by, 449 00:53:31,240 --> 00:53:39,070 well, a valuation like this one we're seeing now and then and then leading to the closure of defined benefit and then the subsequent 450 00:53:39,070 --> 00:53:50,990 shift of the portfolio towards an expensive liability matching bond portfolio to secure the funding of past pension promises. 451 00:53:50,990 --> 00:53:57,660 The increasing cost of funding this portfolio, Bede's under the scenario of the insolvency of universities, 452 00:53:57,660 --> 00:54:01,500 which then drives up the deficit contributions required of the remaining universities, 453 00:54:01,500 --> 00:54:05,490 which then leads to further insolvencies to the point where Trinity College, 454 00:54:05,490 --> 00:54:14,370 perhaps with also's and maybe a couple other Oxford and Cambridge colleges are left, the last solvent employer standing. 455 00:54:14,370 --> 00:54:20,720 So it's noteworthy that what prompted Trinity College leave the scheme was not, therefore the risk. 456 00:54:20,720 --> 00:54:30,270 The DBE scheme posed by remaining open, well invested primarily in equities and other growth assets. 457 00:54:30,270 --> 00:54:40,780 In fact, the chief financial officer of Cambridge has been making the case for being able to fund the scheme prudently at 26 percent contributions. 458 00:54:40,780 --> 00:54:42,330 Sixty five percent growth assets. 459 00:54:42,330 --> 00:54:51,960 So it wasn't the risk that was the scenario that prompted Trinity College to withdraw of the US s scheme remaining open. 460 00:54:51,960 --> 00:54:53,850 An ongoing funded out of equities, rather, 461 00:54:53,850 --> 00:55:03,950 was a risk of the insolvency of multiple employers caused by the high cost of moving to this liability matching bond portfolio. 462 00:55:03,950 --> 00:55:12,900 Now in their latest valuation document. USS actually confirms Trinity College's fears when they sketch extreme downside 463 00:55:12,900 --> 00:55:19,140 scenarios in which they might call on all employers available income and assets. 464 00:55:19,140 --> 00:55:28,890 These involve circumstances in which USS intervenes to force the sector at great expense to purchase this liability matching bond portfolio. 465 00:55:28,890 --> 00:55:33,510 They force them to do so. I fear that if we don't act now, 466 00:55:33,510 --> 00:55:40,980 they'll reach a point at which even all the available income and assets in the higher education sector are insufficient to cover the great cost. 467 00:55:40,980 --> 00:55:46,440 Such bump of. So you see on the screen, your source writes the following in their latest valuation document. 468 00:55:46,440 --> 00:55:51,390 These scenarios are sufficiently extreme that there's likely to have been institutional 469 00:55:51,390 --> 00:55:57,870 failure of the craft substantial and at the higher occasion they're somewhat chillingly, 470 00:55:57,870 --> 00:56:04,980 they say. Moreover, in these scenarios, the trustees, namely you assess this call on employers assets in these scenarios, 471 00:56:04,980 --> 00:56:13,770 would not necessarily be restrained by the possibility of inflicting additional harm on the higher education sector. 472 00:56:13,770 --> 00:56:21,900 So this passage implies that USS is prepared to destroy the entire UK higher education sector 473 00:56:21,900 --> 00:56:28,280 to ensure that defined benefit pensions promises laid down in the past are paid in full. 474 00:56:28,280 --> 00:56:31,760 Well, let's grant, if only for the sake of argument, 475 00:56:31,760 --> 00:56:39,080 that expensive investment in a liability matching bond popular is the most effective means of securing pensions, 476 00:56:39,080 --> 00:56:43,880 promises have been laid down in past years. 477 00:56:43,880 --> 00:56:52,400 We can also accept as a matter of current legal fact, that's a primary duty of the regulator to secure such past promises. 478 00:56:52,400 --> 00:56:58,910 Nevertheless, the justification of such a duty involves normative considerations of public policy. 479 00:56:58,910 --> 00:57:07,740 Such a duty does not follow from any truths. In financial economics regarding the proper method of evaluating a defined benefit pension liability. 480 00:57:07,740 --> 00:57:15,150 Moreover, there cannot be a sound public policy, just a. 481 00:57:15,150 --> 00:57:20,730 So much weight on the securing of past pensions promises even at such high cost, 482 00:57:20,730 --> 00:57:25,300 that destroys the ability of those sponsors to lay down any future pensions promises, 483 00:57:25,300 --> 00:57:36,330 including to the very same workers they promised pensions in the past while also threatening the very solvency of such employers. 484 00:57:36,330 --> 00:57:45,240 For a large multi employer scheme with a strong covenant whose horizons is long, there's no sound case. 485 00:57:45,240 --> 00:57:53,070 The requirement to fund pensions promises against the benchmark of a liability matching bond portfolio. 486 00:57:53,070 --> 00:58:02,030 Such a scheme can instead be funded prudently on the basis of the higher returns on equities that are expected. 487 00:58:02,030 --> 00:58:09,170 It's only through the enactment of regulations and tax incentives, which will give rise to more such multiemployer schemes, 488 00:58:09,170 --> 00:58:16,640 that we'll be able to provide affordable, secure, funded pensions for people in their retirement. 489 00:58:16,640 --> 00:58:22,050 So schemes such as USS ought to be viewed as a model. 490 00:58:22,050 --> 00:58:28,500 Rather than a cautionary tale, it ought to be regarded as the solution rather than a problem. 491 00:58:28,500 --> 00:58:29,289 Thank you very much.