Some Context for the Gilt Market’s Alleged ‘Contagion’ Risk

Contagion. That is the word many commentators we follow latched onto Monday, when the Bank of England (BoE) announced new support measures for pensions forced to sell Gilts to meet sudden margin calls (lenders’ demands to raise account funds to cover possible losses on positions bought with borrowed money).^[i] When trouble seemingly erupted in late September, a vicious circle of rising interest rates and margin call-driven forced selling prompted the BoE to offer to step in as Gilt buyer of last resort for pensions to stop this technical issue from roiling broader markets.^[ii] That appeared to stem the tide for a while, but another 10-year yield spike Monday to a fresh closing high of 4.55% accompanied more intervention to “reduce risks from contagion,” in the BoE’s words.^[iii] There is a lot more to that sentence, as we will discuss, but much of the coverage we read buried the context, in our view, prompting hunts for looming risks. We aren’t dismissing the UK bond market ructions, but we don’t think this is likely to be a 2008-scale financial crisis in waiting.

For simplicity, we will avoid rehashing the political backdrop other than to say we think yields initially spiked when the market overreacted to the new government’s mini-budget that attempted to offset a stealth tax increase with tiny tax rate cuts. For our purposes today, it simply matters that when long rates jumped, it triggered some problems in a corner of the UK pensions market called Liability-Driven Investing (LDI), which is a tactic pensions will use to match their investments with their future liabilities.^[iv] In practice, this is easier said than done, in our view, as some active pensions have unfunded liabilities, meaning they are still accepting new participants, don’t know what their actual benefits payments will be and must earn a long-term return to fund final-salary payments to all participants. As a result, many funds will invest in stocks and other securities as well as bonds, which can earn the needed return over time but also subjects the pension’s total value to market volatility in the interim. Enter LDI, which uses interest rate swaps (contracts to exchange pay outs depending on interest rate shifts) and other derivatives (financial contracts deriving their value from other assets) to hedge against market movement in order to keep a portfolio’s market value (and funding ratio) more stable when volatility strikes.^[v]

We won’t get into the technicalities of all these contracts, as we think that too is beside the general point. In our view, all most investors need to know is that a lot of funds used LDI to get exposure using borrowed money to fixed income, aiming to increase returns when yields were low. The derivative contracts were designed to rise in value when interest rates fall. But when rates rise and values fall, it requires pensions to post more collateral to back the loan.^[vi]

Usually bond markets have relatively low expected volatility, which gives the funds time to pick and choose how to raise collateral in the least disruptive manner possible.^[vii] But Gilt yields’ sudden rise shortened the clock, leading pensions to take the easiest route possible and sell them.^[viii] That sent yields higher, increasing collateral calls, forcing more selling, which repeated until the BoE stepped in two weeks ago Wednesday, offering to buy up to £65 billion in Gilts from pensions.^[ix]

Initially that seemed to help, as Gilt yields settled back down.^[x] But apparently concerns didn’t fade since the BoE’s support had an expiration date of 14 October, triggering warnings from commentators we follow that the bank had merely kicked the can and delayed the reckoning. 10-year Gilt yields climbed, closing at 4.55% Monday—exceeding their pre-intervention high of 4.32% on 27 September—before falling back to 4.19% today.^[xi]

We think recent Gilt volatility is largely sentiment driven, and it looks to us like BoE adjustments alleviated some of the alarm. Initially, it offered to buy up to £5 billion in bonds daily for a total of £65 billion.^[xii] But of the £40 billion offered as of Monday, funds had taken only £5 billion—ostensibly a sign the industry doesn’t need as much firepower as the BoE has offered.^[xiii] In our view, that is a positive sign, an indication pension funds aren’t reliant on BoE support. But the BoE decided to handle this by upping its daily tender offer to £10 billion in hopes of deploying close to the full allotment before the programme expires.^[xiv] It also introduced a new vehicle, the Temporary Expanded Collateral Repo Facility (TECRF), which will let pension funds use their banks as a proxy to park corporate and government bonds at the BoE for 30 days in exchange for cash.^[xv] They can use this cash to meet immediate collateral requirements to buy more time to raise permanent funding without dumping Gilts at bargain basement prices. That programme runs through 10 November, but the BoE will continue offering banks liquidity through its permanent repo facilities, and its communique is clear that it expects banks will use this to assist pension funds with their liquidity needs.^[xvi] Lastly, on Tuesday, the BoE upped its Gilt purchase allotment by £20 billion and widened the programme to inflation-linked bonds, which have also sold off lately.^[xvii]

Our opinion on this is rather mixed. On the positive side, as we think Gilt yields’ spike is part of a sentiment-fuelled global move and unlikely to last, we don’t think the BoE is putting a small bandage on a gaping wound. It appears to us like the BoE is applying one of the biggest lessons learned in 2008 by extending temporary cash to entities that may lack cash in the very short run, but have longer-term assets that exceed their liabilities. In industry parlance, they are illiquid but solvent, and helping such entities find liquidity can prevent collapses and forced selling, basically bridging the gap until sentiment improves. But theory doesn’t always square with reality. Based on our observations, whenever monetary policy institutions come at these problems with a plethora of temporary measures, it can foment alarm, as it creates the perception of systemic issues requiring emergency support. We think the US Federal Reserve’s laundry list of assistance measures in 2020 had this effect, for example, and the BoE’s latest move appeared to do the same Monday if market volatility is any indication.

Using the word “contagion” probably didn’t help, in our view. In the post-2008 period, financial commentators commonly used this term to mean a problem that threatened the stability of multiple banks and, hence, the financial system. But here is what the BoE actually said: “Against the backdrop of an unprecedented repricing in UK assets, the Bank announced a temporary and targeted intervention on Wednesday 28 September to restore market functioning in long-dated government bonds and reduce risks from contagion to credit conditions for UK households and businesses.”^[xviii] Read carefully, we don’t think this is contagion as in preventing the mass sell-off in Gilts from infecting other asset classes and causing a run on the entire financial system in echoes of September and October 2008. Rather, we think the bank is saying it wants to reduce the likelihood that spiking Gilt yields translate to much higher consumer and business borrowing costs, especially for people on variable-rate mortgages. In our view, it is trying to avoid temporarily driving up the cost of borrowing, in the same way mortgage rates’ spiking parallel to Gilt yields two weeks ago seemingly contributed to the first intervention. We aren’t passing judgment on this as a goal, but from our perspective, what the bank is trying to stave off isn’t traditional “contagion” as many commentators we follow seem to view it post-2008.

In our view, calling this “contagion” isn’t the best word choice. Perhaps the BoE has also realised this, as Tuesday’s announcement opted for seemingly more obtuse language, saying “the prospect of self-reinforcing ‘fire-sale’ dynamics pose material risk to UK financial stability.”^[xix] That still seems overstated to us, but it is apparently a more accurate reference to the acute problem of forced selling and, based on Gilt yields’ subsequent movements thus far, didn’t amplify alarm.^[xx] Regardless, we don’t think the UK bond market’s current travails change much longer term. Market plumbing problems aren’t new, in our view; they have always existed, in one form or another, and occasionally they bubble up to the surface in a way that contributes to volatility. But our historical research shows these instances are usually temporary, appearing at times of heightened volatility before fading into the background as more normal market movement returns.

Note, we aren’t calling the system perfect, nor are we weighing in on LDI overall. Pension regulators are reportedly planning to look into this situation and may recommend some changes, which might be sensible if they can identify the most beneficial solutions and minimise the risk of unintended side effects.^[xxi] Considering hedging strategies are typically meant to reduce volatility rather than exacerbate it, in our experience, some of the relevant fund managers could very well elect to make changes on their own and limit their use of these instruments going forward.

But for investors today, we think what matters is that LDI triggering forced Gilt sales probably has more bark than bite as a lasting driver for broad financial markets. In our view, it is largely a sharp after-effect of this year’s bond market decline, which doesn’t look likely to last—not with inflation expectations coming down throughout the developed world, which should help mitigate upward pressure on long rates globally.^[xxii] Whilst our research shows bond market moves tend to be global, we think the fading furore over the mini-budget should also help improve sentiment toward Gilts, rendering the LDI fear a fleeting curiosity over time.