This is the fourth in our series of GT Alerts reporting on the House of Lords Financial Services Regulatory Committee (the Committee) inquiry into the non-bank financial institutions (NBFI) sector.
Our first three GT Alerts have focussed, in some detail, on whether banks and NBFIs should be treated differently from a prudential regulation perspective. The evidence provided to the Committee by a number of witnesses has identified that banks, as a result of being authorised to take deposits from the public and by providing access to the money transmission systems, are in a different position from NBFIs and, to that extent, arguably have less systemic significance. Investors in NBFIs, debt funds in particular, have access to information and the ability to undertake due diligence on the debt fund advisors and negotiate the investment parameters.
In this GT Alert, we consider in more detail whether, even though different from banks, NBFIs nonetheless pose systemic risks. This is a matter that the Committee has focussed on from the outset. However, before considering the evidence presented to the Committee, we will examine what happened to the once-famed hedge fund, Long Term Capital Management or LTCM, which we referred to in our previous GT Alert.
The LTCM Affair
LTCM was a hedge fund founded by former Salomon Brothers Vice Chairman and bond trader John Merriweather in 1994. One of its trading strategies involved profiting from temporary price differences between related securities, described in industry jargon as “convergence trades.” While not doing this full justice, let us assume that two bonds (Bond A and Bond B) are exposed to exactly the same or substantially similar underlying credit risk. The probability of Bond A defaulting should be the same or substantially similar to Bond B defaulting and so, all other things being equal, Bond A and Bond B should trade at the same, or materially similar, price. LTCM’s thesis was that divergence in the market price between Bond A and Bond B was an anomaly and that the prices of the two bonds would, before too long, converge – hence the name convergence trades. If Bond A had the lower market price, LTCM would take a long position in Bond A (in the expectation that the price of Bond A would rise) and if Bond B had the higher market price, LTCM would take a short position in Bond B (in the expectation that the price of Bond B would fall). If LTCM’s thesis was correct, as the market prices of Bond A and Bond B converged, LTCM would profit, potentially on both its positions but at least on one of them, as the mispricing was eliminated. To implement this thesis, LTCM used sophisticated quantitative models. It had a superlative intellectual pedigree: among its partners were two winners of the Nobel Prize in Economics, Robert Merton and Myron Scholes, as well as individuals with significant market experience, including David Mullins, a former Vice Chairman of the Federal Reserve and Eric Rosenfeld, another star bond trader. It was confident, with some justification, in its thesis. For the first three years of being established, LTCM’s returns were eye-catching, justifying the confidence its partners had in their thesis and their models.
The divergence in market price between Bond A and Bond B was typically small and, in order to generate a sufficient return for its investors, LTCM had to use leverage. In this context, leverage meant that it would enter into derivatives contracts to buy and sell bonds that had a notional value that was a multiple of its capital (in this context, the funds provided to it by its investors and its retained profits). Reports vary, but LTCM’s capital at the time of its failure was between $4 billion and $5 billion, but the notional value of the derivatives it entered into exceeded $1 trillion. If its convergence thesis was correct, the value of its derivatives would rise. If not, they would fall, but the size of its profit or loss, in either case, would be magnified by its use of leverage.
All was going well. Then, in 1998 a black swan event occurred: the Russian sovereign debt default caused market prices of bonds to behave in a way that was different to what LTCM’s models predicted. When black swan events occur, panic usually replaces logic at least for a period and investors seek safe havens to protect themselves. To follow through on the hypothetical example above, the market price of Bond B (the bond in respect of which LTCM had a short position) may have fallen, but the market price of Bond A (the bond in which LTCM had a long position) may have fallen rather than risen, as expected, to an extent that the loss on the long position exceeded the gain on the short position, resulting in a net loss for LTCM.
As users of leverage have found on numerous occasions, as LTCM’s trading positions declined in value, its trading counterparties, mainly regulated financial institutions, required LTCM to post margin. Margin calls involve real money – the party that is required to post margin must transfer cash to its counterparty. The greater the notional amount of the derivatives, the greater the amount of the cash required. It took relatively little time for LTCM to run out of liquidity.
Had this situation meant that LTCM’s partners and its investors suffered significant losses themselves, that would not necessarily have had a systemic impact. However, if LTCM could not settle the debts it owed to its counterparties, they too would suffer losses. Suddenly, the losses at a single hedge fund assumed a systemic significance, as they contaminated its bank counterparties. In the end, the New York Federal Reserve had to arrange a bail out of LTCM, involving a number of leading regulated financial institutions, which together injected $3.6 billion into LTCM to cover its liquidity needs.
What Did We Learn from LTCM?
LTCM evidences that there comes a point where the activities of a single hedge fund required government intervention in order to mitigate the risks to the financial system. However, before taking LTCM as conclusive evidence that NBFI’s pose a significant systemic risk that warrants additional regulation, it may be worth considering the unusual features of LTCM in slightly more detail:
- LTCM was not making loans to borrowers in the expectation of being paid interest and repaid principal, collateralised by real assets or otherwise. It was investing in order to exploit differences in market prices of related securities, which its models indicated should not exist. Its losses were related to the behaviour of market prices of bonds, which was driven, in turn, by investor behaviour following a black swan event rather than a payment default by a borrower. The black swan event that occurred was sufficient to make investors behave in a way that was irrational. Such irrationality may be magnified through feedback loops: as some investors sell in a flight to safety, other investors do the same;
- its use of leverage was extreme, based on the confidence it had in the correctness of its models; and
- it models had proved valuable until the occurrence of a black swan event. Black swan events cannot be predicted and all investors may be impacted by them. However, investors that are conservative may generally have a “margin of safety” or have the resources to weather the storm until the impact of the event subsides. The consequence of black swan events on highly leveraged investors might be more severe, all other things being equal.
While there may be hedge funds active today that use strategies similar to LTCM, the NBFIs that are the main subject of the Inquiry do not fall into this category. Providers of private credit are, like banks, lenders. They advance credit to borrowers who operate in the real economy, having assessed the ability of each borrower to service the loan that it wishes to borrow, as well as the value of the collateral that it is able to provide. They also assess how they would achieve recovery from a borrower in the event of a payment default. This is driven by appropriate due diligence, the structuring of the loan, the legal rights that it bargains for, and obtaining an understanding of the legal environment in which those rights will be enforced. While none of these matters are necessarily simple, they do not involve taking the risk on the market price of financial instruments, which are, in the event of a black swan event, subject to rapid and violent volatility, without the benefit of a margin of safety, guarantees, security, or contractual rights. This is a distinction that is important in assessing the systemic risks of NBFIs.
Nonetheless, LTCM illustrates the systemic risks posed by using leverage by NBFIs and the interconnectedness between NBFIs and the regulated financial sector.
What the Witnesses Say
In considering the systemic risk that NBFIs who act as lenders to entities in the real economy pose, the Committee has heard from academics, regulators, and NBFIs themselves. We examine certain of these in this GT Alert.
In one of the first oral evidence sessions, in July 2025, the Committee heard from two academics, one with an economics background, Professor Ludovic Phalippou of the University of Oxford and one with a legal background, Dr. Nerine Lalafaryan of the University of Cambridge.
The Committee asked both experts for their view on whether the increase in NBFIs activities posed systemic risks.
Professor Phalippou’s view was that he did not see NBFIs posing a systemic risk. While not doing it full justice, his evidence was that investors in NBFIs were provided with information that enabled them to assess the risks that NBFIs were taking, and in addition, there were other publicly available sources of information that could be used to assess risk. Dr. Lalafaryan, in her evidence, agreed with Professor Phalippou in terms of whether NBFIs posed a systemic risk, but added an additional dimension in her response: the interconnectedness between private debt markets and public debt markets or bank markets, but also the interconnectedness between different participants themselves and between different types of capital.
Dr. Lalafaryan stated:
….I agree [with Professor Phalippou] that, from a systemic risk point of view, there is less of a concern, but from a regulator’s point of view—I appreciate that the Bank of England has already done quite a lot of work on this—it might be worth looking into this in more detail and understanding not just the interconnectedness of public markets or the bank market and private markets but the interconnectedness between different types of investors and the blurring of capital. We are seeing interconnection and blurring between not just the markets but different types of investors. For example, post Covid, we are increasingly seeing this trend of hybrid funds, which are funds that invest in both equity and debt. Before Covid, there was a clear distinction between equity funds and debt funds, and with Covid we saw an exogenous risk, which was not really factored in in the rationale of the fund-raising of the funds. Now, some of the more sophisticated funds, the bigger players, are in a better position to come up with structures such as hybrid funds, and hybrid funds cause blurring of investors, so investors are investing in both equity and debt. There is not really that clear distinction that there used to be before between equity and debt investors, but we are also seeing an increased blurring in capital, with equity capital and debt capital. With these different interconnections, we are seeing convergence of markets, investors and capital, which is relatively new; it did not exist before Covid; it was not really there after the global financial crisis….
Dr. Lalafaryan’s concern may be illustrated by a hypothetical example. Suppose that a private equity fund (the Sponsor) seeks to purchase a hotel group with 10 hotels. The hotels are all well established, have an operating history, competent management, and the potential for further growth. From an equity perspective, the investment makes sense – it is confident that with some investment, the operation of the hotels can be improved further. The private equity fund approaches a private credit fund established by an insurance company (Lender A) for senior debt. It seeks to borrow 75% of the purchase price of the hotels. This is the maximum leverage that Lender A is permitted to provide. Lender A’s pricing and other terms are attractive. However, it is not able to provide more than 60% of the purchase price. This is the first interconnection between the private equity sector and the private credit sector.
Because of Lender A’s decision to lend less than is ideal, the Sponsor approaches a hedge fund to provide mezzanine financing to make up the difference (Lender B). Lender B is willing to provide this loan and so another interconnection is established between the Sponsor, a participant in the private equity sector, and Lender B, who for these purposes, may also be treated as a participant in the private credit sector, albeit one of a different nature to Lender A. This is the second interconnection.
Now let us suppose that Lender A decides to syndicate part of the senior loan to a bank (Lender C). Lender A and Lender C have the same seniority and so will have the same risk of loss. This is the third interconnection, this time between the private equity sector (the Sponsor) and the bank sector (Lender C).
Let us now suppose that Lender B wishes to finance part of the mezzanine loan using a loan-on-loan facility provided to it by another bank (Lender D). This creates a fourth interconnection. However, this interconnection has two dimensions. First, Lender D is exposed to the risk of Lender B, this being an interconnection between the private credit sector and the bank sector. There is also an interconnection between the Sponsor, representing the private equity sector and the bank sector. Thus, a single transaction has given rise to a series of inter-connections.
Now let us assume that a single sovereign wealth fund (Investor E) has invested in the Sponsor (the provider of the equity), as well as in Lender A (the senior lender), Lender B (the mezzanine lender), Lender C (Lender A’s co-lender pursuant to the syndication) and Lender D (Lender B’s loan-on-loan lender). Let us suppose, just to illustrate the impact of interconnectedness, that there is a black swan event that reduces the value of the hotels to zero. Investor E will, by virtue of the interconnectedness, have sustained a loss on each of its investment exposures. Black swan events rarely have idiosyncratic impacts, and so it is unlikely that only this particular hotel portfolio will be impacted.
This hypothetical example allows us to make two observations. While the NBFIs in this example, being Lender A and Lender B, may not pose a systemic risks themselves, because of the interconnectedness, Investor E may be adversely impacted. The systemic risk of the NBFIs’ activities is not just a function of their activities or their leverage. It is a function of the interconnectedness, leading ultimately to the exposure of Investor E. At the same time, the interconnectedness is opaque – it will not be easy for regulators to assess the inter-connections.
Dr Lalafaryan’s comments led one of the members of the Committee, Lord Eatwell, to make the following observation:
….I want to pick up the issue of systemic risk. I was rather surprised by the confidence that you both display that there is not a systemic risk issue. First of all, as you have both made clear, you do not know—you do not know what the connections are. The data are incredibly opaque. In those circumstances, we literally do not know what the risks are…..
Lord Eatwell’s observation may be worthy of consideration, particularly because his own subject matter expertise is as impressive as the witnesses – he was professor of financial policy at the Cambridge University Judge School of Business between 2002 and 2012.
LTCM Lives
At the end of November this year, Pablo Hernandez de Cos, the general manager of the Bank for International Settlements, gave a lecture at the London School of Economics entitled “Fiscal threats in a changing global financial system.”
The topic of the lecture was the role that NBFIs, in particular hedge funds, are playing in the sovereign debt market, using a strategy called a “relative value trade.” This is not a million miles away from the convergence trades LTCM used. In a relative value trade, a hedge fund would identify a difference between the spot price of a sovereign bond and the forward price of the same sovereign bond.
Let us suppose that a UK sovereign bond with a five-year maturity (Bond C) is trading in the spot market at par (say £100). Let us further suppose that a one year forward contract in respect of Bond C trades at just above the spot price (say £100.50). If the hedge fund purchases Bond C in the spot market for £100 and then enters into a forward contract to sell Bond C for £100.50 for delivery in one year’s time, it should have locked in a small but certain profit, assuming that the counterparty to the forward contract does not default. However, the hedge fund has to finance its acquisition and holding of Bond C for a period of one year.
Mr. Hernandez de Cos identified that the larger hedge funds may find banks that were willing to provide it with £100 (i.e. the spot price of Bond C) without any haircut, using a repurchase or “repo” agreement. Under this arrangement, the hedge fund would sell Bond C to the bank for £100 and agree to buy it back in one year’s time for (say) £100.15. As a result, in a year’s time, the hedge fund would pay £100.15 to the bank and buy back Bond A, before selling it to purchaser under the forward contract for £100.50. In other words, assuming the purchaser under the forward contract performed, the profit to the hedge fund would be 35 pence. Not huge but certain. However, if the trade was for £1 million of Bond C, the hedge fund’s risk-free profit would be £35,000, £350,000 if it was for £10 million, and so on.
So, the trade only makes sense if the bank is willing to provide the hedge fund with a repo facility in a sufficiently significant amount. Thus, the bank would be exposed to the risk both of the hedge fund defaulting on its obligation to buy back Bond C under the repo arrangement as well as the risk that the market price of Bond C will have fallen in one year’s time. If it has not hedged this risk, the risk-free trade for the hedge fund might result in loss for the bank. The bank is, of course, entitled to whatever interest payments are received in respect of Bond C for the one-year period, though this would not mitigate the bank’s exposure to counterparty risk and market risk if it needed to sell its holding of Bond C in the open market, particularly if a black swan event had occurred. Thus, leverage in the hedge fund world may, if sufficiently significant, pose a risk to the bank world.
This led Mr. Hernandez de Cos to make the following comment in his lecture:
…The guiding principle should be to pursue “congruent regulation” when the vulnerabilities are similar across different types of financial institutions (including banks), while properly accounting for differences in business models and potential financial amplification risks. The regulatory framework needs to be sufficiently granular to adequately recognise differences among the very heterogeneous set of NBFIs and the extent to which they may contribute to systemic risks….
In other words, when financial institutions, be they NBFIs or a banks, are exposed to similar risks, they should be regulated in a similar way – we might think of this as “equivalent regulation.” However, the regulation should recognise not just the risk exposure but also the differences between different types of financial institution and their business models – the heterogeneity of NBFIs - with equivalent regulation giving way to congruent regulation.
Conclusion
We started this Alert with a consideration of LTCM and how its trading losses came close to impacting on the regulated financial markets. We recognised, however, that LTCMs near systemic impact was a result of its trading strategy, which relied on the movement of the market price of bonds as predicted by its models and that NBFIs that lend money to real world enterprises are not necessarily subject to the same risks, though black swan events and their impact are never predictable. It is, however, the inter-connectedness that Dr. Lalafaryan identified, between “markets, investors and capital” and the opacity of information on the interconnectedness that Lord Eatwell identified that are relatively new phenomena that may warrant regulatory intervention. Nor, as Mr. Hernandez de Cos explained, should all NBFIs be subject to the same regulatory requirements, the regulatory approach varying in accordance with the business model of each NBFI and not just the risks to which it is exposed.
The hearings of the Committee continue. GT will continue to monitor the developments.