LDI is just the start

Calm has by and large returned to the pension fund sector after the recent LDI (or “Liability Driven Investment”) induced volatility. And to a large extent the Bank of England should be given the credit for this – they acted quickly and decisively, thus restoring order to markets, and by very clearly limiting their action and putting a hard deadline on their intervention, they concentrated minds both at the Treasury and in the pension funds themselves, and forced others to take responsibility for ensuring the wider issues were addressed.

Although order has to a large extent been restored, it would be a mistake to dismiss the episode as “a minor problem with no follow-on consequences and no wider significance”. Instead I think that the LDI episode should be seen as a canary in the coal mine, and the first warning signs of two much larger issues which could flare up and bite the authorities repeatedly in coming years if they are not careful. The first of these is that after years of extreme monetary policy, markets are fragile and far from well understood, and the second of these is that much of the approach by the regulatory authorities in the last 10 years or so has been shown to be inadequate.

I will explore and discuss both of these later in this note. But first, for those of my readers who are not fully au fait with LDI (and until a couple of months ago that would have encompassed the great majority of the financial community, let alone the wider general public), a few words of explanation as to what LDI is and how it went wrong.

What is LDI and how did it cause a market crash?

With the exception of a few public sector schemes which have a direct claim on current taxation, most defined benefit pension schemes consist of a portfolio of accumulated assets, topped up by regular contributions from the employer and scheme members, and out of which their obligations (ie the pensions for the beneficiaries) are paid. This gives a pension scheme a fairly straightforward balance sheet – current assets, future liabilities – and by calculating a net present value for the liabilities, a scheme’s trustees can assess whether the assets are sufficient to meet their future obligations and whether the level of contributions are correct or need changing.

The main way this calculation of the present value of a scheme’s future liabilities is done is by discounting them with some form of interest rate. This is not the place for a discussion of the details of which rates are used or whether they are the correct rates – the important point is that as interest rates change, the present value of a fund’s future obligations also changes. And in particular, as interest rates go down, the present value of all the future pension obligations rises.

This method has been known by and used by the actuarial and pension industries for many years, but really began to impact pension scheme management when interest rates plummeted in the aftermath of the 2008 financial crisis. Suddenly the liability side of the balance sheet became much larger than before, and scheme trustees faced both increasingly negative net positions for their funds (ie, the assets did not cover the net present value of the liabilities), and also increased volatility for this net position. This started to call the solvency of many schemes into question.

Under pressure from The Pensions Regulator (TPR), trustees were encouraged to improve their funds’ resilience by closing this liability gap or coverage ratio and then keeping the net position under tight control. As a result, the valuation of the liability side of pension fund balance sheets moved from being a minor part of the management task to a dominant one. In effect, instead of most investment decisions being driven by a desire to maximise the return on the assets, scheme trustees began to manage the assets to match the liabilities. In short, investment strategies were driven by the liabilities not the assets – hence the name “Liability Driven Investment”.

The main way in which pension funds sought to reduce the volatility of the net position was by making the assets as similar to the liabilities as possible. And since the liabilities resembled a bond-type payment stream, increasingly pension funds invested in bonds. But this created a problem: the return on bonds was so low that by putting scheme assets into bonds, trustees were forced to accept very poor rates of returns on their assets. It seemed that pension funds were only able to reduce the volatility of their net position by locking in insolvency, or at the very least very high costs.

Here the asset management industry proposed a neat solution: if the bonds were not held outright but through derivative positions, perhaps several times leveraged, then the funds could maintain the bond market exposure that they desired while investing the cash thus released in assets yielding the higher returns they needed. And pension funds, seeing a seemingly easy way to solve their dilemma, rushed to do so.

In short, pension fund portfolios changed from being boring and safe portfolios of directly held assets (“long-only” in the jargon of the market, meaning everything in the fund was owned outright), to portfolios full of derivatives, incorporating both long and short positions and in some cases considerable leverage. This is classic hedge fund activity – borrowing through the derivative markets to maintain leveraged positions that are not owned outright – and it was far from clear that all pension scheme trustees really understood what their funds were doing. (Or even that such positions were allowed: there are in theory strong rules to prevent pension funds from borrowing). But TPR, the regulator, appeared not only to be happy with such activity, but also actively to encourage it, with the result that the pension industry collectively engaged in this activity in considerable size.

And then interest rates rose sharply on the back of Kwasi Kwarteng’s disastrous financial statement. And pension funds who owned fixed income exposure through derivatives found that the values of those positions were tumbling and that they faced heavy margin calls on them to make good the resulting losses. It is true that the value of their liabilities was also tumbling, so that the LDI strategy was delivering what it promised: the net positions of funds engaging in LDI, and so their coverage ratios, were unchanged and so protected. But this protection of their solvency had come at the cost of heavy calls on their liquidity, and in order to meet these calls funds had to sell assets, usually fixed income assets, which sent the market down further and resulted in yet more margin calls.

In other words, markets were caught in a classic death spiral, and it took the Bank stepping in as buyer of last resort to restore order. Which they did, and at the cost of “only” about £20 billion of gilts bought. These days, that really isn’t very much compared to the other amounts of public money being thrown around; the Bank’s decisive action not only worked but kept the cost surprisingly low. But it still represents £20 billion the authorities did not want to have to use – a costly episode indeed to add to all the other costs of the Truss/Kwarteng interlude.

What do we learn from this?

For the regulators, I would suggest there are three learning points:

Firstly, an approach which concentrates too hard on making one financial sector secure will often have the result that risk migrates to other parts of the financial eco-system. After the 2008 financial crisis the FSA and PRA put a considerable amount of effort into making the banking system secure; they were successful (and the banks are now much stronger than they were and much better capitalised), but this was achieved mainly by forcing them to reduce their holdings of risk. As a result risk ended up moving away from the heavily-regulated banks and into less well regulated and less strongly capitalised areas of the financial system, which were less able to handle the risks they were taking on.

Who would have thought a few years ago that the next market crisis would emerge from the pension fund sector, a sector more usually known for being low-risk, long term asset holders? But this is what risk migration results in.

It is probably inevitable that without a single regulator overseeing all aspects of all markets, each regulator will concentrate most on their own bailiwick and be less aware of the consequences on other sectors, and that risk will sometimes therefore migrate from where it should be held to where it is less appropriate. Moving the PRA back into the Bank of England was meant to address this. But TPR is defensive of its independence, and for some people that seems to matter more that the overall efficiency and effectiveness of regulation. Maybe the financial sector needs further consolidation of its regulatory structure.

The second learning point is that what the regulators choose to measure matters, as people will manage the measure, not use it to measure the management. By asking pension funds not only to measure their coverage ratio (assets vs liabilities) but to manage it on a quarterly basis, TPR changed the coverage ratio from an indicator of where the fund was and made it a target – and in the process it lost its usefulness as an indicator. In other words, the measure ceased to be a window through which management saw what was really going on and became merely a mirror in which they saw only what they themselves had done.

In switching their emphasis to close management of the coverage ratio, pension scheme trustees lost sight of the single best way to ensure that they could afford their pension obligations, which is to grow the assets as strongly as they can.

And the last learning point for regulators – and I would suggest the most important – is that what a financial system needs above all for systemic resilience is diversity of business models. A system with many diverse business models will be more resilient than a system with one uniform business model, even if the one business model adopted in the latter is superior to many (perhaps even all) the models in the former. The problem with LDI is not that it makes any one pension fund safer or less safe, but that if everyone employs it, it makes the market overall less safe, because when it moves, it all moves in the same direction.

Regulators need to remember that their task is not to make every single financial institution safe, but to make the market as a whole resilient to shocks. And a monoculture in which all pension funds look similar because they are all following the official advice to employ LDI does not achieve that.

Wider lessons

The LDI episode is an example of the unexpected consequences that policy decisions can sometimes generate. It is extremely unlikely that when central banks collectively decided to adopt the very aggressive monetary policy stances they deemed necessary after the 2008 financial crisis, they foresaw that one result would be that the hitherto boring but safe world of pension funds would prove to be a weak link and one of the first sectors to show distress.

In part this is because, if truth be told, central bank analysis of the long term consequences of ultra-low interest rates and ultra-active central bank balance sheets was fairly minimal full stop. Both were so far outside anyone’s experience that central banks were literally in uncharted territory, and they had their hands so full keeping the show on the road that they had little time for worrying too much about the future.

But the result is that even today, we are still far from complete in our understanding of the aggressive monetary policy of the 2010s. We do not fully understand the long term effect on capital markets as those providing funds (including pension funds) were not rewarded adequately for do so. We do not know the long-term effect on the economy of “zombie companies”, companies that are kept alive only by ultra-low interest rates, absorbing resources others could use more productively in the complete opposite of Schumpeterian creative destruction. We do not know how economies will respond to “orthodox” monetary policy in a world with unorthodox levels of government debt. We do not even fully understand how the move of central banks from being Lender of Last Resort to being Funder of First Resort has affected the way money markets work.

But what we can almost certainly say is that the markets have changed, that new fragilities will continue to emerge, and that the LDI crisis will be far from the last time the markets spring a nasty surprise on those tasked with keeping our financial system safe.