I (Buckner) will be presenting on the subject of LDI and the ‘bigger than it looks’ problem on Nov 3, 2022 04:30 PM London, by Zoom.
If you would like to attend, please contact me here for the flyer, and the Zoom link.
“The September 28th collapse in the gilts market led the Bank of England to announce a £65bn programme to stabilise long-term interest rates and save UK pension schemes from defaulting on their Liability Driven Investment (LDI) positions by the end of that day. Schemes had experienced liquidity problems on their Interest Rate Swap positions, losses on which had triggered margin calls requiring them to sell assets.
“The affair has highlighted the danger of hedging illiquid positions with liquid ones. I shall argue, based on a model of the underlying interest rate transformation involved in LDI, that there is more to the issue than has yet been appreciated.”
Another mystery. This report from PwC says that the total pension deficit has reduced as a result of rise in yields.
They say that their source is data from the Pension Protection Fund. But see the chart above which I have overlaid with my estimate of the price of a 2030 gilt. You see that the 2030 gilt value precisely tracks the PPF liability estimates.
Here is the mystery: why so, given that LDI exists, and that LDI immunises schemes from changes in the long gilt rate?
Note also that the PPF say “”Conventional and index-linked gilt yields are used to value liabilities. *We do not hold sufficient data* to capture the impact of any structural changes to asset allocations nor to accurately capture changes in any leveraged LDI portfolios.”
The chart above simulates the price history of a 2% coupon 30 year gilt purchased at the beginning of January 2017 and held until now (18 October). As you clearly see, the price rose throughout 2019, 2020 and most of 2021, then abruptly collapses, with the decline accelerating towards the third quarter of 2022. As journalists live to say “gilt prices fall as yields rise”. Quite.
No comment for now. I will discuss the odd but significant blip in March 2020 in the next post.
The coverage of the Liability Driven Investments affair has been highly polarized. On the one hand, there are those (mostly consultants) suggesting it is a storm in a tea cup, LDI working ‘as expected’, nothing to see here, move on.
On the other hand, actuaries like Richard Lund, speaking at the First Actuarial Conference about severely disrupted core gilt market functioning, excessive and sudden tightening of financing conditions for the real economy etc etc.
Someone should at least be answering questions. E.g. if there have been losses, how much are they? See the chart above. A long-dated bond (here 18 years) could easily have halved in value with a move in rates from 2% to 5%. Any scheme that sold at those levels will have incurred a massive loss, or so you would think. Clearers like Northern Trust reported being overwhelmed by a slew of margin calls during bond market turmoil, so we know that selling happened.
So who was selling? And how much was lost? “We should be told”.
See BBC News at Ten, 12 October 2022, for easily the worst media explanation of LDI yet. The fun starts 7:10.
Presenter “So let’s turn to our business editor, Simon Jack, who’s the expert on all these things.
Jack:
8:00 “Investors want more interest to justify the extra risk”. Not really. A gilt is risk free because the nominal payment (coupon and principal) is risk free. What investors are demanding is compensation for the extra inflation implied by the market.
8:05 “The cost of borrowing shot up”. Gilts are an asset, not a liability, and don’t involve borrowing. Of course, arbitrage implies the cost of long-dated borrowing will go up commensurately, so perhaps we could forgive him.
8:20 “Some pension managers have used them as security to get ready cash now”. Argh. Pension schemes have bought them using borrowed money, using them as security for the borrowing. No cash, apart from the haircut (‘margin’).
8:25 “To be able to pay out on those pension promises” Argh again. The scheme assets, typically risk bonds or equities, enable the payout. And again, there is no cash apart from the margin (and margin doesn’t have to be cash, necessarily).
Jack’s Wikipedia entry says “Before entering journalism, Jack worked for a decade as a corporate and investment banker in London, New York City and Bermuda. He has said that he neither liked the work, nor showed much ability at it.” Fair enough.
John Cochrane has an amusing post at the Grumpy Economist on the recent shenanigans operations at the Bank. See the diagram above taken from Sir Jon Cunliffe’s letter.
Apologies again for the absence. Eumaeus (and Ulysses) were far away from their homeland for many weeks, and missed the market hell of the past few weeks. More on that later.
Meanwhile, the chart above shows the total return on 10 year gilts against total return on the FTSE, which I have been following for a long time, and which I last discussed in January 2021 here.
Source: WTW, Analysis of Proposed Solvency II Reforms, 21 July 2022
The Association of British Insurers came out yesterday with a firm statement on the proposals to increase the capital that insurers ‘hold’, arguing, based on analysis by Willis Towers Watson that the proposals would in fact on average lower the capital ‘held’.
This results conflicts somewhat with UK government claims that the proposals would free up capital to support investment in wind farms and other worthy things.
Figure 4.1 from Willis is copied above. It had Eumaeus scratching his head for a long while, even with help from his dogs. The block colour represents the range in which ‘all’ the sample of 16 firms sit, except for the maximum and minimum outlier values, so the block colour does not in fact represent ‘all’ firms, but never mind.
But the results are intriguing. Under the PRA scenario A, for example, the average reduction in own funds (Solvency II ‘capital’) leads to an average reduction in capital of about 5%. But that is for all firms. For all annuity providers – who are the main customers for MA – the average reduction is 20%. And worse, there is an outlier firm that loses 30% of its capital. Yikes!
Admittedly, scenario A is severe, and the PRA always signalled this, but none of the other scenarios look particularly encouraging given the purpose of the government proposals. For the proposals were to free up capital for long term infrastructure projects, which could only come from firms with long term liabilities, i.e. annuity providers. But it is annuity providers who are most severely affected under all the scenarios!
It’s all a mess and will lead to more friction between the government and the Bank of England.
That is not to say the reforms are a bad thing. Far from it: in Eumaeus’ view they do not go nearly far enough. But the mess is an indication of the broken nature of life insurance, and of the failure of regulation to deal with the problems in a timely way in the aftermath of the Great Financial Crisis.
The Eumaeus mailbox was buzzing late last week after the publication of a speech by Deputy Governor Sam Woods 1 on proposed reforms to the Insurance capital regime. The general mood was surprise.
