Happy LDImageddon anniversary

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Roughly a year ago, a lettuce outlasted a UK prime minister, after an obscure strategy in a dull industry upended a boring market and almost triggered a financial crisis. What a time it was to be alive.

Barclays’ gilts guru Moyeen Islam has taken a timely look at what UK pension plans have done since the liability driven investment-triggered near-implosion of the long-end UK government bond market in the autumn of 2022.

And lo, he unsurprisingly found “an extended period of retrenchment” for LDI plans as they deleveraged and built up liquidity. But they now look in great shape thanks to a transformed funding situation.

The fundamental idea of LDI — better matching the duration of assets with the long-term nature of the liabilities of defined benefit pension plans — remains sound. And while the rise in gilt yields tripped up some providers last year, the overall effect has been to radically improve their overall funding picture for UK DB schemes.

The crux of LDImageddon was that pension plans with LDI strategies used long-end gilts both to hedge themselves against falling yields and as their liquidity backstop. That meant that when yields spiked they got hit by margin calls that had to be met by dumping gilts, sending yields higher and triggering more margin calls.

As Moyeen puts it (our emphasis below):

All financial crises ultimately end up as a crisis of liquidity, whatever their genesis. The LDI crisis was no different. Having used gilt repo to generate leverage, pension funds were left holding assets (corporate credit, illiquids, etc) that could not be used to realise liquidity when margin calls increased due to the volatility seen in gilts. One key lesson for schemes was that assets that are used to hedge liabilities cannot and should not be simultaneously used for liquidity purposes.

In response to the liquidity crisis, the Bank recommended significant increases in the liquidity buffers that schemes are required to hold in order to be resilient to minimum moves in gilt yields of 250bp: this breaks down into 170bp ( termed “systemic resilience”) and a further 80bp (“baseline resilience”). The systemic resilience buffer is designed so schemes can withstand a severe move in yields without having to sell assets –that might otherwise lead to a negative feedback loop of a firesale that would increase systemic risk from the sector affecting other parts of the financial universe and ultimately lead to an unwarranted tightening in financial conditions that would impact the wider economy. The Bank’s staff estimate that this would by equivalent to the moves seen in the linker market in Q422 and equate to a 1-in-100 year 5-day event. The “baseline resilience” level of 80bp reflects the idiosyncratic risks of LDI funds. The calculation is equivalent to the “lookback” that the clearing house would use in its setting of initial margins (IM) for centrally cleared risk, in this case a 10yr look back window for 99.8 percentile of 5-day moves — for 30yr linkers this is estimated to be around 80bp.

You can see the deleveraging in various data, such as repo volumes for UK pension plans:

After the dash for cash at the peak of the mayhem last autumn, liquidity has normalised at a higher level. Given that leverage has been ratcheted back, that means the “underlying liquidity of schemes has improved and that, from a macro perspective, access to liquidity is not a binding constraint on funds’ activity,” Barclays argues.

As you’d expect in a higher-yield environment, pension plans have continued to tilt towards fixed income to de-risk themselves.

Mooted revisions to Solvency II might increase the share of illiquid assets though, which are currently at about 12 per cent.

However, the most important, lasting consequence of LDImageddon might be that the Bank of England has grudgingly accepted that it may occasionally have to backstop the non-bank bits of the financial system — what Dan Davies last year dubbed “Lombard Street 2.0”.

Here’s Islam, with Alphaville’s emphasis in bold:

In the medium term, perhaps the most important structural change has been flagged by BOE Executive Director for Markets Andrew Hauser. In his most recent speech, Hauser outlined the need for fresh central bank tools that will offer lending to non-bank financial institutions along the lines recommended by the IMF.

Central bank lending to investment vehicles in the non-banking sector would be a significant policy departure as their traditional counterparties lie in the banking sector. However, it would reflect the new reality that the broader non-banking sector is now a key source of liquidity for the corporate sector, and so can be a source of both financial stability and real economy risk. Any new mechanism would be a backstop rather than a routine channel of lending.

Given that the BOE does not regulate parts of the NBFI universe, it does raise issues of adverse selection (ie, how the Bank knows that it is lending to a viable entity). But the key point is that the Bank likely sees itself backing away from significant market interventions, and it would rather have lending facilities in place that can automatically be triggered by its (expanded) set of market counterparties.

In terms of collateral demand, Hauser’s speech specifically cites “gilts — at a minimum” which would support on-going demand for gilt collateral from schemes. Important issues over maturity and haircuts and pricing versus existing facilities are to be settled over time but the direction of travel seems clear: the Bank has little desire or appetite to intervene directly in the gilt market in the way it was forced to in Q4 22 in anything but the most extreme of circumstances

Further reading
— No, current gilt yields don’t vindicate Liz Truss, please don’t make us explain this again
— How much did Truss’s 49 days in office actually cost UK pension funds?