When it comes to LDI collateral, is credit a dirty word?

Shortly after the panic of the gilts crisis last September had subsided, there followed an inevitable witch hunt for what, or who, was to blame.

One alleged factor pointed to was overly-strict minimum requirements for high quality collateral assets. The argument went something along the lines of “if schemes could have posted corporate bonds as collateral, instead of only gilts and cash, they wouldn’t have been forced sellers and wouldn’t have lost their hedges”.

This instantly jarred with me. It felt much like blaming seat belts for road deaths. Seat belts are there to protect you, even if in a minority of cases they cause injury themselves. The same could be said of requiring high quality collateral for derivative transactions. Had everyone forgotten that these robust regulatory requirements were put in place to protect investors following the Global Financial Crisis (GFC) of 2008? 

It felt much like blaming seat belts for road deaths. Seat belts are there to protect you, even if in a minority of cases they cause injury themselves.

Following the GFC I witnessed first-hand the efforts of regulators, fund managers, consultants and trustees to fix the industry’s inadequate practices. This was a collaborative effort, and took years of regulatory reform, time, money, and a ready supply of cold towels. But the resulting high standards of collateral quality led to better transparency, greater trading flexibility and improved security for all participants of derivatives markets.

However, following the 2022 gilts crisis, it became apparent that some insurance companies had been using corporate bonds as collateral once more. This caught on and some Liability Driven Investment (LDI) managers have since also started doing this for pension schemes in their segregated portfolios.

I was initially sceptical but was I being too narrow minded? After all, one fifth of our clients are now running a reduced hedge compared to a year ago as a direct result of the stricter liquidity requirements post the mini-budget debacle. Whilst reducing collateral quality initially seemed like a backward step, if it allows schemes to run more hedging this is worth investigating.

To really understand whether this is a good or bad development, it’s important to consider what was so wrong with what used to be done, to understand if those mistakes can be avoided second time around. Let’s start with a reminder of how collateralisation works.

How derivative collateralisation works

When you enter into a swap, on day one no-one owes anyone anything. Both parties are square. But from that point onwards one party will go into profit at the expense of the other based on the market factor the price of the derivative is linked to.

If you are winning you will be posted collateral. If you are losing, you will need to post collateral.

At the outset the two parties agree which assets can be transferred to provide security in the event that one party owes money to the other; these days typically gilts or cash. This protects against the scenario where one party defaults before the contract expires. The detail is set out in a document agreed between the pension scheme and the bank called the credit support annex, or CSA.

Derivatives pre-2008

Now let’s dial back the clock to before the GFC. Back then, all manner of assets were routinely permitted as collateral, under so called “dirty CSAs”. Cash of different currencies, corporate and government bonds, equities, or indeed nothing. Many contracts were not collateralised at all. 

LDI managers tended to be at the better end of market practice, and even following the default of Lehman Brothers in 2008 I was not aware of any pension scheme materially affected by loss of value in their LDI portfolio - thanks to collateralisation. However, it wasn’t just the risk of losing money upon default that was problematic; derivatives with low quality collateral were difficult to value and even harder to trade.

Back in the day if a bank owed money under the swap, and it had a dirty CSA, this was a great opportunity to post some lower quality assets. It could then save any cash or gilts on its balance sheet for other higher purposes.  The longer the swap the better as the collateral could stay there as if it was on long term loan. The flip side is that the party who received that collateral was potentially stuck with it for decades. This wasn’t quite as bad as it sounds as it was just there as security in case of default, except that is, if you wanted to close the contract out early.

Trying to close out a 20-year interest rate swap, where you are in profit and have received corporate bonds as collateral, was inconvenient for the bank on the other side. They would much prefer to retain the current arrangement than settle up for cash. This was a problem as you were either stuck with your swap, or you had to take a write down in the value to close it out early. 

Cleaning up

Following the calamity of the financial crisis, additional regulation for banks tightened everything up.

The changes in regulation in the EU meant everyone had to move to “clean CSAs”, typically permitting only gilts and cash as collateral, sometimes only cash. This was a tiresome and complicated exercise renegotiating terms with each bank for each scheme at the time. Moreover, pension schemes in the main lost value as they were typically in profit at the time on their long-dated interest rate swaps, following the yield falls of 2011. This meant that moving to higher quality collateral arrangements generally involved schemes having to take write downs in the value of their swaps, effectively paying for the privilege of improving their collateral terms.

However, when all was done, the world was a better place. Contracts could be valued more simply, substantially reducing the cost of closing out positions, and everyone benefitted from greater security.

Back to the present day

Coming back to the present day - given this history, can it be a good thing to reintroduce credit collateral into the mix? To my mind, dirty CSAs for 20-year interest rates swaps remain as undesirable as ever, the short comings remain. However, there are other approaches being used to achieve a similar, but superior solution.

One method is using credit repo.  This is like pawning your corporate bonds for cash, with a commitment to buy the corporate bond back in the future. The cash can be posted as collateral in the meantime, and you retain economic ownership of your corporate bonds.

Crucially, credit repo contracts have much shorter terms than interest rate swaps, meaning you aren’t tied in for 20 years; more like 1 year. The repo can be kept separate to the hedging derivatives used, meaning they don’t complicate your hedge and you retain flexibility to reshape it as required. You also know exactly what is happening - you are borrowing cash in exchange for specified corporate bonds, rather than entering into an open-ended agreement where you may receive or post any amount of pretty much anything.

What’s the catch?

A cash loan backed by corporate bonds via a repo is a more expensive way to finance than borrowing cash via a gilt repo. This is of the order of cash interest rate + 0.5% p.a. so this isn’t an ideal source of long term financing but is still much cheaper than a typical bank loan. So it can be a tool to have in your back pocket for a rainy day. It also involves increasing leverage which needs to be fully understood.

Bringing it all together

So all in all, when it comes to collateralisation, using credit as part of a carefully managed strategy can be cleaner than initial impressions. Furthermore, if using credit within the management of collateral creates scope for more schemes to better pursue their objectives with greater flexibility and resilience, it is surely a step in the right direction. Bringing it back to motoring safety, if high quality collateral is the seat belt, credit repo is the airbag.