LDI: the better mousetrap that almost broke the UK

A pension meltdown forced the Bank of England to intervene in gilt markets on Wednesday. Executives told the Financial Times that markets barely dodged a Lehman-Brothers-like collapse – but this time with your mum’s pension at the centre of the drama.

Problems with “pension plumbing” are what caused the mess. The culprit is said to be a popular pension strategy called liability-driven investing, or LDI.

Leverage is a key element of many LDI strategies, and are basically a way pension funds can look like they’re an annuity without making the full capital commitment of becoming one. As the drama unfolded we talked to some analysts and pensions managers who weren’t directly involved with the UK meltdown, but are experts in LDI and liability-matching practices. Here’s a summary of what they told us:

What on earth happened this week?

In short, gilt yields soared. That forced UK pensions to sell gilts. That pushed gilt prices even lower and yields even higher, which forced UK pensions to sell more gilts (and other liquid assets), and so on. The Bank of England intervened to stop the spiral. See the chart below:

That looks bad. And now I have more questions than I started with. Why did UK gilt yields soar in the first place?

It was indeed bad. The benchmark 30-year gilt yield spiked 1.2 percentage points in just three days, a massive move.

Yields have been climbing for a while, but the latest extreme jump happened because chancellor Kwasi Kwarteng’s mini-budget was an homage to Harry Styles’s Don’t Worry Darling performance — a foreseeable disaster with a poorly hidden British accent.

That doesn’t explain why UK pensions were forced to sell.

Yes, that’s the key question here!

The dumbest explanation is that pensions had to sell gilts because gilts are easy to sell. Pensions also sold other liquid assets, like mortgages and high-grade corporate credit, for the same reason. They needed to raise cash.

What did they need the cash for?

Well, the cruel irony is that pensions needed collateral for margin calls on leveraged trades hedging against big moves in . . . UK government bond yields.

So pensions sold bonds (among other things) to raise that cash, pushing yields up, making hedging trades even more expensive, and requiring even more collateral.

If the BoE hadn’t stepped in to arrest the declines, pensions may have defaulted on those contracts, which would have been very bad. That isn’t the same thing as going bust – it’s not like all the investments disappear overnight – but it does risk tying up the pension in a knotty legal fight over settling the default.

As Toby Nangle pointed out on Wednesday, The Pensions Regulator said in 2019 that 62 per cent of the biggest UK pensions had at least some exposure to interest-rate swaps, a type of derivative . . . 

Derivatives! Didn’t Warren Buffett call those “weapons of mass financial destruction”?

[Heavy sigh] Yes, he did.

But interest-rate hedging with swaps is a pretty vanilla type of derivative use, all things considered. Pensions used everything from equity options to single-name credit default swaps, according to The Pensions Regulator’s 2019 survey. Interest-rate swaps are used widely among the large US and UK investors who want to match their assets with liabilities; think pension funds and insurers.

In fact, the basic concept of “liability-driven investing,” or LDI, just means planning your investments’ cash payouts to your future cash needs. In a simple world, you wouldn’t need derivatives to do this at all. Pensions would simply buy bonds that would pay out what they need to pay pensioners, when they need to pay pensioners. Need a lump-sum payment in the future? Buy a zero-coupon bond that matures on that date! Easy.

We don’t live in a simple world, though.

Exactly, and that’s where things get messy. Until a pension plan is fully funded and frozen (meaning it has no new participants or benefits), there isn’t any way to know for sure what its liabilities will be in the future. That means pension funds need to take extra risk. There is an entire industry built to tell them how to take this risk, and there are many different options, like private equity, hedge-fund strategies, and notoriously, the rare kickback scheme.

Generally, when a pension starts putting money into liability-driven investment strategies, it means that it isn’t interested in continuing to watch its pension’s funding ratio swing around with the market.

Pensions generally don’t put all their money into LDI strategies at once. Beyond the uncertainty around what they will owe, many are underfunded, meaning they need to put cash into riskier assets to earn a return and make up for that difference.

But even when they need to put cash into stocks or other risky markets, they can use leverage (swaps, repo, etc) to match their entire investment portfolio’s duration with the duration of their liabilities.

In this scenario, with well-functioning swap hedges in place, the current value/cost of all of a pension’s liabilities declines in a bond-market sell-off. At risk of oversimplification, that means that equities can fall, but if they decline less than the bond market, the pension’s funding status can actually improve.

That is probably a large part of the reason that pensions of the 100 biggest US companies were actually more than fully funded as of August 31, according to consultancy Millman, which estimated that assets totalled 106 per cent of liabilities.

Hold on. LDI is based on a quant-y accounting gimmick?

We can’t say the concept itself is a gimmick — again, the entire insurance business is pretty much built to match assets with expected liabilities. Barring widespread defaults, bonds do have a predictable stream of cash flows.

So in a perfect quantworld, a pension would own a risk-free government bond that pays what it owes a retiree, on the day it owes it to them. Then who cares if the price of the bond goes down today? They’re holding it to maturity. And even here on earth, a company that freezes its defined-benefit plan can then offload its (matched) liabilities to a life insurer who can perform insurance magick (maths) to ensure the payouts happen (for a fee). Works well, right?

But adding interest-rate swaps to the mix sounds like a way to pretend you have that bond portfolio without actually putting up the capital to invest in a bond. (Or forgoing the potential equity returns.)

Yes! It does seem a bit like some quants decided to build a better pension-management mousetrap, doesn’t it? And whenever there is a way to sell questionably structured derivatives to pensions or other “real money” investors, there’s almost always a Wall Street counterparty eager to make the trade.

Like who?

If you think you might know, please do get in contact.

So now they’ve blown up. What does that mean for the rest of us?

Stay on the lookout for more deleveraging outside of the gilts market. In the US, investment-grade corporate debt and agency mortgage-backed securities should be an area of focus, since, like gilts, they are fairly liquid and easy to sell.

For its part, the investment-grade ICE Corporate Bond Index lost 2.2 per cent between Friday Sept. 23 and Tuesday Sept. 27, underperforming benchmark Treasuries; that compares to a 1.5-per-cent loss over the entire prior week, when the Fed raised rates by 75 basis points and took an unexpectedly hawkish stance.

We know LDI strategies are popular in the US too. What’s to stop something like the UK blow-up from happening there?

A couple of experts who help build LDI strategies for US pensions told us that US companies tend to use less leverage than their UK peers.

But it is difficult to find concrete data backing that up, because companies don’t need to report their pension plans’ leverage or swap exposure in their SEC filings. And given the general attitude that is likely to surround LDI for some time, we aren’t especially eager to accept that at face value. Millman loosely gauges LDI participation by pensions’ allocations to the bond market: Its latest annual pension-funding study found that their fixed-income allocation had climbed to 51 per cent by year-end 2021, up from 42 per cent in 2008.

Still, it isn’t clear that a comparable disaster is destined to occur in the US, even if the country’s pensions were similarly levered. Remember, the main cause of the mess was a gilt sell-off that was extremely fast and violent, prompted by a major panic over a mini budget.