LDI – the full unexpergated story

This is a text of an after-dinner speech I gave to the Pensions Network on Thursday November 24th.

Our story begins in 2004 when the UK adopted an accounting standard that discounted pension scheme liabilities to the AA Corporate bond rate.

In 2011, Professor Iain Clacher of Leeds University wrote

At a macroeconomic level, mandating the use of AA rated bond yields to discount the present value of pension liabilities has resulted in schemes purchasing greater amounts of financial investments which better match the estimated present value of their pension accounting liabilities. Pension schemes have been divesting real asset investments, such as equities, and investing in financial investments, such as long‐dated index linked gilts and corporate bonds, that are a better match for the estimated accounting liability. The systemic effect of this has been to exacerbate the bubble in long‐dated bonds and, in particular, index linked gilts.


If that sounds hard work for an after-dinner speech – relax. Things will get easier.


The Bank of England invests its staff pension scheme almost entirely in gilts and bonds, this costs the taxpayer just over 50% of the pensionable salaries, an amount beyond the means of most private sector employers. A better way had to be found to match investments to liabilities and the problem became of interest to Merrill Lynch and to two young bankers keen to explore secondary markets for pension products. They worked out that by swapping long term and short- term gilts using derivatives and by exploiting the Repo market (effectively a means of borrowing), pension schemes could create synthetic holdings of gilt several times the amount actually held. This was known as leveraging and the bankers, Rob Gardner and Dawid Konolu-Atulu set about selling the idea to large DB pension schemes. They found ready customers in Friends Provident and Royal Mail and this led to the establishment of Redington, an investment consultant named after a famous actuary but regulated not by actuaries but the FCA, powered by bankers using banking products to solve pension problems.


This version of LDI was very different from the vision of John Ralfe, who had – with brilliant timing – engineered the Boots pension scheme into gilts at the start of the century. It meant that schemes using this approach could have their cake and eat it, by borrowing gilts they could show they fully covered their liabilities, but through the leverage, they had money left over to buy real assets which could provide growth for the scheme and ensure that the contribution rate was kept well below 50% and at acceptable levels to the sponsoring employer.

Unsurprisingly, this approach soon became very popular, Redington prospered, actuarial practices soon started copying and today 60% of the 5800 DB pension schemes in the UK use this form of leveraged LDI. Not only was this concept adopted but it was maintained. From the beginning to the end of quantitative easing, low interest rates meant that bond prices remained high. In 2019, index linked gilts were the highest performing asset class. While liabilities were expensive because the discount rate was on the floor, the value of the LDI portfolio saved the day. The number of schemes into the PPF was smaller than estimated, despite dire predictions to the contrary, corporate DB plans stayed afloat and this was largely attributed to LDI. It should be noted that the last decade was also kind to other asset classes and while LDI invested schemes did well, so did others – who eschewed borrowing – schemes such as LGPS and Rail Pen – plus of course all DC schemes, became increasingly well-funded , especially when accounted for on a “best estimate basis” where the discounting of liabilities was linked to the return of scheme assets rather than AA bonds (or gilts).

But all good things come to an end and so did the low interest rates encouraged by quantitative easing and maintained by a relatively orderly economic climate. By 2022, inflationary pressures meant that interest rates started rising and the yield on gilts started rising fast. This meant that schemes were becoming very solvent very quickly which pleased everyone. But it also meant that the value of the gilts borrowed through LDI fell sharply. It was like having negative equity on your mortgage but worse as the banks were able to call for the margin between what the gilts were borrowed at and what they were worth. Trustees started having demands on them for increased collateral which, if not met, would mean that the borrowing would be taken away, denuding the scheme of all those synthetic gilts which made the scheme solvent and now super funded.

To keep the synthetic gilts in place, schemes would either deeds  get a line of credit from the sponsoring employer or start selling growth assets. The trouble was that much of the growth assets were illiquid (we know what that means) so all that could be liquidated were gilts, which were rather less valuable than they had been.

In July, Mercer issued its clients with a warning note, telling Trustees to buckle up for the ride and make sure they had provisions in place to meet cash-calls to come. The mechanism to convert assets to cash was known as a “liquidity waterfall” and schemes started to worry about whether the stress-testing that they had been assured had been done, had gone far enough.

It hadn’t.


On 22nd September, the Bank of England announced it would start buying back gilts as part of “quantitative tightening”

On the next day, the mini-budget announced a program of tax-cuts and spending that the market assumed would require £100bn of borrowing.

Over the weekend, Liz Truss and her Chancellor made it clear that the mini budget was only the beginning and more could be expected in November and the new year.

The uptick in gilt yields on the 22nd became a torrent by the following Monday and the collateral calls started arriving thick and fast. LDI had originally been conceived of as bespoke to each scheme – that’s the Redington way. But the demand for this wonder drug spread to smaller schemes who needed packaged solutions. Legal & General, BlackRock and Insight were first to pioneer pooled LDI funds where each scheme had effectively a protected cell in which their liabilities were matched.

But with pooled funds housing hundreds of these cells, the human resource needed to get each to cough up collateral was immense, especially as smaller schemes had less governance and took longer to get things done. Many of the smaller schemes could not buy extra units in their funds (the way collateral was posted) and so their assets were sold to meet margin calls. These assets were gilts, and this was how the “doom loop” happened. Selling gilts was like expecting a buyer to catch a falling knife and the pooled fund managers were having trouble giving them away.

By this point , the pooled fund managers had the ear of the Bank of England who now understood that this flood of unpurchasable gilts was putting the integrity of the gilt market in jeopardy. The Bank saw the potential for serous financial instability and that was when it decided to intervene.

By 11am on September 28th, the 30 year gilt yield had already risen 100 basis points on the day, this was as much as the Pensions Regulator had stress-tested the system for and things were in danger of getting out of control. The Bank announced it would buy up to £65 billion of gilts over the next fortnight with a hard close on October 14th. The Bank left pension schemes in no doubt that they had better have their house in order by the 14th. Despite this, the pension industry sort an extension which they hoped to be announced on the morning of 12th October

At 9. 30 that morning, I found myself on stage at the PLSA being asked by Fiona Bruce in front of 1500 people “Henry, do you support an extension of the support for long dated gilts by the Bank of England”. My answer was no – it felt like I’d farted in a crowded lift.

The Bank kept to its deadline and Kwasi Kartheng was fired on the 14th. The expected spike in gilt yields the following Monday did not happen and by the 20th the Prime Minister had resigned. The arrival of Jeremy Hunt and then Rishi Sunak had calmed the markets and rates returned to pre mini-budget rates.

Bud damage had been done. Con Keating and Iain Clacher estimate that so far this year, UK DB pension schemes have seen £500bn pounds “go missing” as a result of schemes matching investments to liabilities using gilts as the investment yardstick. £300bn of this is thought to have been lost since the mini-budget.

Many small schemes are thought to have been locked out of their LDI programs and have little to show for their investment in pooled funds. They are likened to the swimmer who when the tide goes out is found to be wearing no clothes. Many other schemes were forced to sell assets in a fire-sale, often to hedge fund managers buying on the yellow label. Employers are being asked to re-stock the shelves as scheme liabilities become more expensive again, as yields receive. This means yet more cash injections, money that is not going into DC pensions or helping staff meet the rising cost of living.

The pension industry’s reaction to this is to call what happened a black-swan which like all other financial disasters is considered a once in a thousand year event. Those of us like, I suspect the majority of you, and certainly me, are left asking some important questions.


  1. What if, instead of valuing pension scheme liabilities against a gilt-based discount rate, the actuaries had agreed a market return based on scheme assets? Would that really have reduced member security or would it simply have stopped the need for LDI?
  2. What if TPR or the FCA or PRA had intervened and deemed the use of Swaps and Repos “borrowing” – deeming leverage LDI illegal. What commentators such as Baroness Bowles to be the case.
  3. What if the Pensions Regulator had recognised that with 60% of pension schemes piling into LDI, there was systemic risk if things went wrong and stopped its spread?
  4. What if the stress testing by TPR , BOE and FCA had gone further and broken the LDI model, would this have required schemes to reduce or do away with leverage?
  5. And what if the people who read the blogs by Con Keating, Iain Clacher and others on my website, had taken notice of what was being said and unwound the leverage for themselves as some – (notably the PPF), did.

This is of course in the realm of speculation, but I do believe that the problems of 2022 were foreseeable for many years, and I do so because they have been predicted on my blog.

It seems to me, that the pension industry has been the architect of its own demise. It has forced billions to be diverted from productive use into DB pensions and much of that money has been frittered away meeting unnecessary collateral calls resulting from ill-advised if not illegal investment in banking instruments. Much of the investment has been made by people ill-equipped to understand the risks and – certainly in the case of pooled funds – the apparatus for ensuring collateral was properly posted, broke down.

The industry is now saying it will learn lessons, reduce leverage, widen the assets that can be posted as collateral and improve advice and governance, so people don’t find themselves naked on the beach in future.

But I question whether this is simply a patch on a broken system. We have new DB funding regulations under consultation, perhaps now is the time to go back to basics and ask how we invest our DB pension funds so that we are not forced into these artificial arrangements which look an accident waiting to happen.

About henry tapper

Founder of the Pension PlayPen,, partner of Stella, father of Olly . I am the Pension Plowman
This entry was posted in pensions and tagged , , . Bookmark the permalink.

5 Responses to LDI – the full unexpergated story

  1. John Mather says:

    Will the estimate of the cost ever be replaced by the true cost?

  2. Byron McKeeby says:

    Alliance Boots said in 2012 that only one in five of its 75,000 British workers was in its main final salary scheme, 5,500 were in the existing defined contribution scheme and the vast majority — almost 55,000 — were in no scheme at all.

    Of the two main DB schemes, the smaller Alliance Unichem scheme closed to new members in 2002 and both schemes had closed to future accrual by June 2010.

    Boots’s new private equity owners, Kohlberg Kravis Roberts, had agreed in 2007 to pay £418m in instalments over 10 years to plug what was then described by commentators as “the retail and pharmaceutical group’s pension hole”, just six years after the gilts switch.

    Please bear these details in mind when claiming the Boots gilts switch in 2001 was such a success.

  3. Allan Martin says:

    Firstly, please also bear in mind the history of DB funding and pension scheme solvency. In the 1970’s the leaving DB service entitlement moved from a refund of historic employee contributions to an inflation ravaged deferred pension, 1978- GMP revaluation helped, 1980- anti-franking helped, 1986-91 non-GMP revaluation helped, 1997 statutory pension increases helped and all increased the total pension fund benefits provided. Over time the non-guaranteed investment return assumption (and associated risk) had to recognise scheme cash flow and solvency coverage. Company reporting was necessary.

    Secondly, repaying deferred pay pension promises are uniquely different to bank borrowing or house purchase mortgages. Pension provision involves assumed returns higher than bank or mortgage rates (BoE excepted). That risk:return balance needs prudence, measurement, monitoring, management and regulation. LDI is just management of some of the risks.

    • Byron McKeeby says:

      But at the W&P select committee hearing this week I heard that interest rate hedging, taking on fixed rates versus short-term variable rates, was said by one participant (an MP, I think) to be what you would/should do with house mortgages (or, by inference, bank borrowing), suggesting pensions were not “uniquely different” at all.

  4. Jnamdoc says:

    Well said Byron. Can we stop presenting the likes of the Boots story as a success. Scheme closure, 55,000 workers without a scheme, and half a billion £ plug is not a good outcome. I’m sure proponents of the Boots / LDI world will be saying it was for the good of the schemes, but killing with kindness is still killing, and that is what the industry has done to DB.
    Collective schemes for workers that provide an income for the non-working elderly after a life-time of work and savings should be aligned to a shared growth in the economy – that is what a DB system properly managed would deliver and it is the only moral solution. If its ok for the MPs and civil servants, why not for the Many? Passing DB over to the actuaries and bankers who treat it as a “product” has been the downfall – their priority has been to minimise the risk to the product, not on promoting a coherent integrated system to deliver a fair privately funded living wage to the elderly.

Leave a Reply