The BoE’s Troubles Really Are All About The Leverage

When the Bank of England surprised the market with further open market operations earlier this week, The Blind Spot reached out to Helmholtz, a former sell-side pro with nearly 40 years worth of experience running rates desks at various investment banks. Helmholtz has agreed to give us his thoughts on the condition of pseudonymity, but we can testify to his top-rated expertise.

The commentary relates to both the temporary expanded collateral repo facility and the Bank of England’s additional measures to support market functioning. The BoE announced both on Monday, October 10.

So what exactly is all of this about?

Helmholtz’s immediate thought is not to confuse the repo facility, which is a liquidity facility, for a purchase facility. That means it does not address the main problem facing Liability Driven Investment (LDI), which is leverage.

As he explains:

Let’s start with a simple LDI model for a pension fund (PF). Assume the starting cash in the PF = 100, and liabilities = 300; for the sake of argument let’s assume a gilt exists that exactly matches the duration of the PF liabilities. As the pension fund only has cash equal to one-third of its liabilities, it must enter into a liability-driven investment with 2x leverage – now duration of assets = liabilities (although note cashflows will NOT match as leverage has to be unwound at some point).

If the gilt falls in price by 20 per cent, assuming liabilities move in the same way, then the LDI still matches liabilities. But the pension fund now has to post 2×20 = 40 as extra margin against the falls in the leveraged liability drive investment. And that’s where the spiral starts.

While the Bank of England may have nipped a lot of the pain in the immediate aftermath of the mini-budget, Helmholtz says it did little to prevent the risk of ongoing long-gilt sell-offs and the associated doom loop they bring with them:

Typically haircuts on gilts are small, but the market movements we’ve seen in the last few weeks will overwhelm even those, particularly for the highly convex instruments (long dates + IL) – which are the main assets in PFs…..

That’s why the original purchase facility helped by providing a floor on gilt prices.

Non-gilt-related assets must be factored into this too:

So PF typically hold a mixture of assets – gilts (nominal + IL), corporate bonds (these two classes are the assets which can be matched against liabilities for duration – the matching assets), infrastructure assets and some equities. Note that the non-gilt holdings are often in fund format or with a manager, which reduces the effectiveness for use as collateral.

The good news, according to Helmholtz, is that there’s less need to be worried about the sensitivity of the duration of bonds relative to their yields, something known as convexity:

Provided assets (including LDI) and liabilities are properly matched, duration and convexity will move together. The problem remains the leverage. At the start the PF is leveraged 2x; 100 cash = 300 assets. After the fall it’s 80 cash vs. 280 assets (240 gilts + 40 margin) = 2.5 x. (Another way to think of it is a matched portfolio plus a 40 loss).

But while the BoE tools address a lot of these issues, Helmholtz is not convinced they address them all (our emphasis):

It will allow PFs (or their agents) to post a wider variety of collateral against the falls in value of their LDI, but it will not help with the level at which the market clears. Waking up this morning, the falls in price yesterday rather confirm that analysis. So maybe the new facility will prevent a market failure in terms of collateral posting and similar, but I doubt it solves the underlying problem.

How does this resolve? I would expect that PFs need to liquidate their non-matching assets to fund the LDI losses; which actually makes sense at the higher nominal and real yields. It seems like the BoE is hoping this can be done by providing liquidity rather than price support.

Some related issues on Helmholtz’s mind remain the following:

1. Many pension fund assets are held in fund format. That means it will be difficult for them to be posted as collateral or for the assets to be sold (hence some gatings already). He is not close enough to the PF market to know how much they hold in really illiquid stuff like private equity, but he thinks those will probably face issues of solvency rather than liquidity.
2. Helmholtz says there may be a small silver lining for schemes that have inflation caps (most of them), which will reduce their liabilities for this year.
3. There will be knock-on effects for holders of long-duration assets (mainly in the insurance world.) But solvency II regulation will not have helped matters.
4. Derivative fallout now needs to be watched.

All of this suggests firesales remain the biggest risk for the sector, and that the trouble ain’t over yet.