Lessons of UK pension crisis

Back in the 1990s, Long-Term Capital Management, a powerhouse hedge fund littered with Nobel laureates and finance geniuses, had a great convergence trade. It roughly arbitraged away the price spread between two otherwise identical securities. In order to magnify the fairly small arbitrage profits available, leverage had to be installed. 

Soon, many other hedge funds and proprietary traders began replicating these leveraged convergence trades. All was well until the Russian debt crisis of 1998 blew things out of the water and caused the price spread to go in the other direction. In other words, what was overpriced or “rich” – and was shorted by these hedge funds – became even more expensive (or desirable to hold), as a result of the crisis, while the cheaper security that was on the long side of the trade remained the same or became cheaper. 

The widening spread as a result of leverage, margin accounts and margin calls temporarily depleted these hedge funds’ collateral accounts. This was despite the trade being known to be profitable in the long run, since convergence had to happen at some point. We know this has to be the case as the renowned and savvy deep-value investor Warren Buffet was one of the first to offer a rescue package to Long-Term Capital Management.

The 14 banks that were cobbled together by the Federal Reserve Bank of New York to bail out the beleaguered hedge fund were eventually made whole. One report claimed the banks even made good profits when all the positions were unwound between late 1999 and early 2000.

Risk and returns

Liability-driven investing (LDI) is the process of building a liability-driven investment portfolio. It models the assets and liabilities of a pension fund, insurance business or bank portfolios under various scenarios using various assumptions and investment strategies. 

LDI is a returns-based risk management framework for implementing an alternative to pure asset-liability management (ALM), which is normally referred to as balance sheet management. LDI incorporates strategic asset allocation, risk budgeting, portfolio construction and ongoing risk monitoring and reporting. The main difference between ALM and LDI is the risk-adjusted return-seeking component in LDI, and the discount rates used.

Over the past few decades, sponsors of private defined-benefit plans have shifted their focus from asset price volatility to “funded status” volatility. This means that they design their asset allocation programme within an LDI framework to maintain funded status within an acceptable confidence interval. 

Pension plans have two principal sources of funded status volatility. One is the risk associated with return-seeking risky assets, and the other is the interest rate risk that is embedded in the liabilities. The rationale for using LDI is to reduce the contribution of interest rate risk while allowing for exposure to return-seeking assets in a risk-adjusted manner. This approach is meant to lead to a reduction in funded status volatility.

The types of LDI strategies that could be used to reduce interest rate risk without reducing the expected returns are: hedging with long-duration bonds, treasury “strips” or zero coupon sovereign bonds, and interest rate derivatives.

Pensions and LDI

Pension funds have two main economic goals. First, to provide sufficient benefits to their members or participants and, second, to maintain healthy funding or funded status ratios. 

In order to achieve these objectives, pension fund managers have shifted their attention towards the funding status or funding volatility by using a more risk-focused strategy, such as LDI. This has resulted in pension funds marking and segregating the risks associated with their liabilities, and then constructing more transparent returns-seeking portfolios that embody clearer objectives and effective risk management.

An LDI strategy is primarily about the management of interest rate and inflation risks related to the liabilities. Its purpose is to hedge the liabilities against these types of so-called segregated risks. There are other types of risks, such as longevity and credit spread risks, which are not considered here.

The liabilities are calculated using actuarial assumptions, and the value is the projected benefit obligations (PBO). Therein lies some of the fault. Academic literature has always argued that capital market-based assumptions as opposed to actuarial are always more prudent in asset management functions. In any case, a change in interest rates directly impacts the value of the liabilities and hence, the funding ratio. 

Additionally, the amount of leverage which is installed to magnify returns depends on how much of the assets are used to hedge the liabilities, and the balance sheet assets invested in the returns-seeking assets. If, for example, liabilities are 100% hedged, then any change in interest rates will have the same change in asset values. Leverage in the context of LDI is the duration of the LDI portfolio divided by the duration of the liabilities.

Caveats 

The use of LDI helps exploit a wider range of investment choices, adds flexibility in investment policies, and provides further support to de-risking, apart from freeing up resources for investing in returns-generating strategies. For example, the plan sponsor could reduce interest rate sensitivity by using long-duration bonds or by using leverage. The use of leverage is primarily to reduce the gap in interest rate sensitivity between the assets and liabilities, and to magnify the returns-seeking component. 

The bottom line is, how much of the interest rate risk the plan sponsor is willing to tolerate. Plan sponsors need to explicitly state the target levels of key risks that drive the funding ratio volatility. 

The risk management problem that is less manageable is the scenario where interest rates rise and the equity markets crash. This will result in marked-to-market losses, which need to be collateralised in the brokerage margin account on a daily basis. This is not as manageable an outcome for plan sponsors, especially unsophisticated ones, from a derivatives risk management perspective. 

The moral of the story is that the plan sponsor needs to be fully cognisant of what the funding ratio is before it gets into a leveraged LDI solution, especially in managing derivatives-related liquidity risks from margin calls in adverse, black swan scenarios that brought down the once great Long-Term Capital Management. 

Risk parity strategies

The risk parity approach entails a well-diversified portfolio where portfolio construction, combined with leverage, ensures all asset classes have the same marginal contribution to the total risk of the portfolio. By construction, risk parity is not based on the reduction of surplus volatility of the entire portfolio – i.e., its assets versus liabilities – but purely based on the total risk of the portfolio.

In risk parity, the total risk is defined or measured by the volatility of the rates of return of the portfolio. In this scenario, risk parity portfolios make relatively large allocations to lower-risk asset classes, such as government bonds, through leverage. Leverage is installed with the expectation that the portfolio will outperform a typical portfolio without leverage for the same level of risk.

Risk parity can also help in the asset/liability context. Within this framework, the optimal investment strategy on the asset side will consist of a liability-matching portfolio and a risky asset allocation portfolio. The liability-matching portfolio will be dependent on the plan’s liabilities and the risky asset allocation portfolio will be based on risk parity. This approach will lead to lower surplus risk. 

The idea of leveraging depends on the funding ratio. If the plan is underfunded, then leveraging may help in good times as it reduces the surplus risk. However, there are financing costs that need to be considered, and the risk of a black swan drawdown event that can lead to margin calls. At its core, the risk parity strategy is an approach to allocate risk efficiently and to deliver stable returns in normal times.

What happened in the UK?

The issue in the case of the UK pension funds last year was due to their use of leveraged liability-driven funds in their plans. The goal was noble: to help the pension funds hedge their long-dated liabilities. In general, the strategy would be to use long-dated bonds on the asset side to hedge the liabilities. Unfortunately, there is a shortage of liquid long-dated bonds in the UK, which resulted in plan sponsors utilising leverage to help reduce the plans’ surplus risk.

Hence, when yields rose, leveraged LDI strategies generated losses, which in turn triggered margin calls from brokers for additional collateral. This required pension funds to infuse cash, which they couldn’t afford to do. 

In particular, pension plans in the UK have to invest some portion of their assets in pooled LDI vehicles. The cash infusion from these vehicles was slow to come, resulting in what academics call a “funding liquidity” problem. As solvency worsened, LDI funds had to deleverage, thus putting further upward pressure on yields due to the selloff in bonds. 

Again, it is what academics call the “liquidity spiral” problem, which comprises a “loss spiral”, i.e., downward pressure on bond prices, that in turn leads to a “margin spiral”, i.e., margin calls for more collateral.¹

Some of the key issues that need to be considered are: leverage, which raises the risk of forced sales during a meltdown; lack of diversification; and the market size of assets being sold and the pooled LDI funds’ reliance on these assets, which in general are slow to raise liquidity.

These factors, which lead to market dysfunction, can put additional pressure on central banks to intervene as the liquidity provider of last resort.

Plan sponsors in various other jurisdictions, especially the US, seldom use leverage. This is primarily due to the low duration of their liabilities. US plans also have a larger proportion in risk-based assets, while their treasury holdings are much smaller than in the UK, where the value of pension liabilities to that of the bond market is quite high at around 60% to 70%. Moreover, US plans have sufficient allocations to liquid assets to draw upon when prices of leveraged positions fluctuate.

Clear understanding is key

The issues in the UK related to LDI have led many to think about the implications related to policy making. It also raises questions related to asset allocation, macro policy and the broader investment environment. 

UK regulators encouraged defined-benefit plans to de-risk by taking derivatives positions to hedge liabilities. This led to the replacement of interest rate risk with liquidity risk.

Pension de-risking need not be an all-or-nothing decision. The risk profiles of pension plans can be changed through asset allocation and LDI techniques. Plan sponsors eager to de-risk need to determine how to do it. If for example the plan is closed to new members, as is the case for many plans in the UK, the liabilities are given and it might make sense to focus on a buyout, i.e., purchase of annuities. 

In the US, plans use a pure corporate discount curve to discount the liabilities. In such a case, the plan must adjust to target a more economic measurement of the liability, i.e., moving to a higher quality discount curve. This will eliminate investment and longevity risk. 

Those intending to use LDI to de-risk pension plans need to consider doing stress tests to factor in the various market shocks that may occur, as happened in the UK. This will lead to plans maintaining a specific level of liquidity buffer along with a reduced risk profile.

Another point to consider is to de-leverage LDI pooled funds. Collateral can also be widened to include other asset classes outside of cash or gilts. Access to external liquidity or working capital might also help by providing a temporary source of liquidity to bridge short-term market turbulence.

LDI may not be the answer to all pension plans. Each one has its own unique set of challenges and circumstances, so LDI strategies should be designed to be flexible and adaptable in the face of internal and external factors. When implementing an LDI strategy, there must be a clear understanding of – and a plan for – managing liquidity risk. 

*Joseph Cherian is practice professor of finance at the Asia School of Business in Kuala Lumpur and Cornell University (visiting) in New York. Yogi Thambiah is chief executive officer and founder of Princeton Analytics & Consulting in New York.

¹ Market Liquidity and Funding Liquidity, Review of Financial Studies, 2009, Vol. 22, by Markus Brunnermeier and Lasse Heje Pedersen