The recent panic in the LDI market has received a lot of coverage, some of it breathless and designed to promote a short narrative, but much of it calm and authoritative – and generally pointing out the broader issues. As with previous commentary on things like Covid, we are not setting ourselves out to be experts on LDI investing, rather we are reading what the experts are saying, and adapting our narrative accordingly, rather than trying to shoe-horn the events into our pre-existing viewpoint.
We thus continue to recommend the articles in the FT on markets (it’s just their opinion writers we tend to disagree with), even though they are behind a paywall. This article for example, provides a very neat summary of how the GBP1 trillion + industry evolved in recent years and the key point of course is in plain sight; the name. Liability Driven Investing started out as a bank product but evolved into an asset management product where Defined Benefit Pension schemes were able to avoid the regulatory driven demands to maintain solvency in response to changes in the calculations of their liability as discount rates moved. A relatively plain vanilla product, in essence it involved having a lot of gilts on the balance sheet that went up in value as interest rates fell, the rise in the assets thus offsetting the equal rise in the liabilities (lower discount rate increases pension liability). However, as this article explains, what happened next was, what always happens; leverage and illiquidity enters the system and is used to ‘enhance returns’, without duly acknowledging the higher risk. (Edit, a nice piece here from Terry Smith, sums it up nicely).
Yet another unintended consequence of having interest rates too low for too long.
In his excellent new book, the Price of Time, Edward Chancellor, makes many fascinating points, but central to his thesis is that, holding the rate of interest below its natural rate always produces problems, as capital is misallocated, speculation explodes to the upside and economic growth is suppressed by poor credit allocation, zombie companies and financialisation crippling productivity growth. It also leads to pension fund problems as yield chasing emerges and assets and liabilities are mismatched. This time of course was no different and the LDI problem is just the latest piece of the (predictable) jigsaw.
Thus the taking on of leverage by the LDI funds was compounded by the foray into illiquid assets – the pool of exciting ‘alternatives’ being promoted by the fund managers and the consultants in recent years. The attraction of not having to mark-to-market’ was obvious to the funds and the lack of measured volatility allowed the risk management industry (who seem to care about volatility above all else) to declare these as ‘lower risk’. Thus rather than simply matching assets and liabilities, pension funds started to use leveraged derivative structures to hedge their liabilities, that allowed them to not just own bonds – in this case gilts – but also a higher yielding basket of illiquid alternatives, such as Private Equity, Private credit and infrastructure bonds. Once again, returns were enhanced, but so was risk. In contrast to the headline, risk was not reduced, it was simply moved and in this case largely just hidden.
This has left LDI as part of the great unwind; as rates have increased, the counterparties to the derivatives need more collateral, and that could only be raised by selling their liquid assets, in this case gilts. Thus the system was threatened with a ‘doom loop’ of selling triggering yet more selling. As such, the role of the Bank of England was to step in and act as buyer (rather than lender) of last resort and thus act as a circuit breaker. This apparent flip from QT back to QE was presented by the new government’s many opponents as a disaster, when in fact it was merely sensible financial management.
The fall in the price of gilts triggered a call for increased collateral from LDI managers, which could only be met by the underlying pension fund selling more gilts.
Systematic risk appears because of a bad system
The BoE needed to intervene because, as we have seen time and again, the financial sector has got itself in a mess by mis-matching its assets and its liabilities; either by borrowing short and lending long, or by promising/requiring daily liquidity when the underlying investments are illiquid. Term structure or liquidity premium exist for a reason and both represent risk, but too often the industry focusses almost exclusively on volatility as its risk metric. This means that policies that suppress volatility lead to other, more hidden risk being taken, but not acknowledged. A case of ‘teaching to the test’.
As such, when volatility eventually does spike, as it has in bonds this year, it can trigger forced selling in assets as the dread Value at Risk (VaR) models designed for leveraged traders are nevertheless imposed on long term investors. As former BoE Governor Mervyn King opined in his book with John Kay, Radical Uncertainty, part of the problem is the false sense of certainty that comes with all these model based ‘solutions’ and risk models that work, right up until they don’t (usually when they meet the real world). This in turn exposes the other hidden sources of return/risk that the regulators have once again apparently missed and results in the whole structure having to be unwound, usually by the only entity not subject to the rules forcing the unwind, i.e. the central bank.
While we can applaud the markets team at the Bank for stepping in quickly, it does not absolve the policy makers from helping create the problem in the first place, with interest rates at the wrong price and not enough monitoring of the risks being taken as a consequence.
Addendum. This is not 2008 redux.
Many of the more excitable commentators are making comparisons with 2008, not least because the markets seem to be following a 2008 analogue. Many are placing Credit Suisse as ‘the new Lehman’, which it isn’t. It is nevertheless perhaps worthwhile for a brief reminder as to what we believe to be the real reason that a financial crisis became an economic one, ie why Lehman wasn’t like LTCM and ‘contained’. It was, once again, the fixed income markets at fault (it almost always is, since they are where the leverage and the illiquidity usually sit). We know about the asset liability mismatch in the CDS market (brilliantly covered in The Big Short) but less is talked about the Commercial Paper (CP) market, which was the real vector through which the financial markets hit the real economy. Back in 2008, yield chasing investors (them again) held cash in money market funds (MMFs) rather than in banks, as they had a bigger return and were regulated as securities rather than a counterparty risk with banks. This was because they invested heavily in 30 day commercial paper and were thus a form of shadow banking system. Corporates meanwhile used the CP market for managing their working capital requirements and the totals ran into the hundreds of billions/trillions. When Lehman went down, the regulators required all bonds referenced to Lehman marked to zero – including the CP, meaning that several of the big MMFs ‘broke the buck’, ie went below par. Understandably, this triggered he equivalent of a bank run on several of these shadow banks and more importantly meant that, when it came to roll the 30 day paper into October 2008, the MMF industry said no; they had just been confronted with a potential duration mismatch. This then triggered a working capital crisis for corporates, forcing destocking of inventory and a suspension of new orders as they struggled for liquidity. This in turn appeared as a sharp drop in GDP and a fall in prices (discounts to shift inventory) that the hysterical commentariat instantly labelled ‘The Great Recession’ – mostly on account of them having ‘predicted’ lower house prices.
The reason for revisiting all this is that we only realised what was happening with the CP market when the Fed announced it was going to step in and buy commercial paper. Then, as now, when regulators create distressed sellers and deny access to other buyers, they have to ‘make the market’. That, we believe is what is happening now with LDI, most of us only realised it was a problem when it was in the middle of being ‘solved’. It’s not good, and it’s not pretty, but it was necessary.
